biology essay membranes

2 Biological Membranes

Biological membranes, introduction.

Now that you’ve had a chance to better understand microscopes, we now turn to learning about the makeup of the cells themselves. In this chapter, we will start by exploring the composition and function of biological membranes. This will give us the opportunity to review what we know about macromolecules, specifically lipids and proteins and, to a lesser extent, carbohydrates. After that, we will discuss the function of the plasma membrane and how the structure of those macromolecules contributes to the function of the plasma membrane. Finally, we will discuss cell surface–adjacent structures, like the extracellular matrix or plant cell wall, that are found just outside of the plasma membrane.

Topic 2.1: The Chemical Features of Biological Membranes

Learning goals.

• List the four primary features of a biological membrane and explain why they are important for cellular function.
• Explain how the chemical composition of a membrane (including lipids, carbohydrates, and proteins) contributes to its function.
• Explain how thermodynamics and the hydrophobic effect hold membranes together and selectively exclude some molecules but not others.

All membranes in the cell share some common features, despite the fact that they each have a different composition, fluidity, and permeability. In this topic, we will discuss the common features of all membranes and touch briefly on which macromolecules contribute to specific membrane functions.

It’s worth noting that a biological membrane is not the same thing as a phospholipid bilayer , despite the fact that these terms sometimes are used interchangeably. A phospholipid bilayer is made only of phospholipids and nothing else, whereas a biological membrane will have many types of lipids in it (including glycolipids and cholesterol) as well as proteins. We’ll see examples of this later in the topic.

Membranes Are Barriers That Define Compartments

One of the most basic features of cells is that they are separated from the environment by a membrane. Not only do the membranes create a barrier that separates the inside of the cell from the exterior, but they also function in a way that gives the cell a great deal of control over what can enter or exit it. This creates an environment inside the cell that is different from the outside (Figure 02-01).

Molecules are able to move around freely (diffuse) within each cellular compartment. However, it is more challenging for some molecules to cross membranes. This is because biological membranes have specific chemical properties, which determine when and how molecules cross the membrane. Thus, membranes are a key feature that allows cells to maintain compartments with a distinctive chemical composition.

Like everything in science, the idea that biological membranes contain both lipids and proteins took time for cell biologists to accept. Sometime in the 1970s, the current concept of the membrane was developed and was known as the “ fluid mosaic model .” The “fluid” part refers to the lipids being able to move around within the layers of the membrane, and the “mosaic” part refers to the fact that proteins are expected to be scattered across and throughout the membrane. Since then, further research and experimentation have expanded and refined our understanding of how membranes work. In this part of the chapter, we will explore four general features of biological membranes:

• The membrane is a bilayer, made up of lipids and proteins.
• The membrane is selectively permeable.
• The membrane is organized but fluid.
• The membrane is asymmetric.

We will look at each of these in a little more detail. Then in Topics 2.2 and 2.3, we will take a more in-depth look at the relationship between lipids and membrane fluidity as well as the organization and function of membrane proteins.

General Features of Biological Membranes The Membrane Is a Bilayer Made Up of Both Lipids and Proteins

The earliest work on membranes was done on plasma membranes from red blood cells. This is because red blood cells are easy to collect in relatively large quantities. In addition, in mammals, they lack a nucleus and internal membranes. So they contain mostly “pure” plasma membrane without the need to experimentally separate other membrane types during experimental isolation. To isolate these membranes, cells are burst open and the membranes are gently washed to remove cytoplasmic debris. Lipids and proteins are separated by treatment with strong detergents or organic solvents (e.g., ether) and characterized. Next, we’ll take a look at the chemical properties of lipids and proteins separately to discuss how they each contribute to the structure of the biological membrane.

1a. The Lipid Component of Membranes: Formation of Lipid Bilayers

(For a complete review of the basic structure and formation of lipid molecules as well as a description of polar and nonpolar molecules, please see the introduction .)

The main lipid components of membranes are phospholipids , a family of molecules with a polar phosphate head group and two fatty acid tails, each connected to an arm of a glycerol via an alcohol residue (Figure 02-02A). However, they are almost never alone in a biological membrane. Other lipids (glycolipids, sphingolipids, and cholesterol) as well as a great many proteins are found in every biological membrane. The chemical properties of phospholipids allow them to adopt a bilayer shape, a key component to a biological membrane.

In water, fats and oils will form into large droplets of jumbled molecules that do not mix with the water (think salad dressing). However, membrane lipids are able to form sheets because of two very important characteristics:

• They are amphipathic (i.e., there is a polar “head” region, which can freely form hydrogen bonds with water, and a nonpolar “tail” region, which cannot). Some sources refer to polar molecules/regions as hydrophilic , which translates to “water loving.” Conversely, the nonpolar molecules/regions are sometimes referred to as hydrophobic , which translates to “water fearing,” but as these terms are misleading and stray from referencing the chemical properties of the molecules, we don’t find them to be as useful. We prefer the terms polar and nonpolar when describing the head and tail regions of the phospholipids, as they are more scientifically accurate. However, both will be used in this textbook.
• They form a roughly cylindrical shape, which tends to stack together well (much like cans of soda; see Figure 02-02B).

Phospholipids, such as the one shown in Figure 02-02A, come together to form a bilayer—two sheets of phospholipids with their polar head groups oriented outward. Each layer of the bilayer is called a leaflet . Since the nonpolar tails of the phospholipids face inward, away from the water, they are sequestered away and do not interact directly with the water molecules (Figure 02-03). This orientation allows for the maximum freedom of the surrounding water. Because of the stable membrane conformation, if a tear or a hole is created in a lipid bilayer or a biological membrane, it rapidly seals up again.

When phospholipids are mixed with water, they will spontaneously form into spherical bodies called liposomes that have water on the inside as well as on the outside (Figure 02-02B). Liposomes are commonly used for the targeted oral delivery of drugs and other agents in medical treatment. Liposomes are the simplest version of a cell membrane, so they are able to fuse with a real cell membrane to release its contents directly into the cell. This is useful if the compound inside the liposome would not easily pass through a membrane under normal circumstances. A great example of this technology in action is the RNA vaccines developed to target the SARS-CoV-2 virus (the cause of the COVID-19 pandemic). These RNA vaccines use an outer coating called a lipid nanoparticle , which is a combination of phospholipids, cholesterol, and other compounds that are designed to help contain and stabilize the RNA vaccine and ease its entry into the cell.

Thermodynamics Drive the Formation of the Lipid Bilayer: The Hydrophobic Effect

Phospholipids form stable bilayers in an aqueous environment due to thermodynamics. The spontaneous formation of a lipid bilayer from lipids in an aqueous solution, simulated in Video 02-01, shows the power of thermodynamics in action. The laws of thermodynamics explain why it is more energetically favorable for phospholipids to clump together and form a bilayer in water than it is for them to remain dispersed.

The phenomenon where nonpolar molecules in an aqueous environment clump together is called the hydrophobic effect . Clumping together is not favorable because the nonpolar groups attract each other per se, but because, when clumped together, nonpolar molecules do not have to interact as much with water, which is polar. When the nonpolar material clumps together, water and other polar molecules are freer to move and, more importantly, hydrogen bond with each other. This freedom of motion for the water allows for the overall reaction to be spontaneous.

While we don’t want to get too far into the equations of thermodynamics, here is a brief description of how thermodynamics favors the hydrophobic effect. You may want to return to your general chemistry notes for this (or refer back to the information in the introduction ).

Recall from general chemistry that the free energy equation is

[latex]\Delta G= \Delta H-T \Delta S\text{,}[/latex]

where ΔG is the Gibb’s free energy change, ΔH is the change in enthalpy (often described as internal energy or bond energy of the molecule), ΔS is the change in entropy (which is often thought of as motional freedom), and T is temperature.

As a reminder, reactions are energetically favorable when ΔG is negative, which means that energy will be released. This can be accomplished by decreasing the enthalpy or by increasing entropy. In the case of the formation of the lipid bilayer (and any other clustering of nonpolar molecules in water), it is the entropy of the water molecules surrounding the nonpolar groups that is changed when lipids assemble into bilayers. Two points are critical here:

• Nonpolar particles force the surrounding water molecules into an energetically unfavorable configuration that distorts the normal hydrogen-bonded structure of water (Figure 02-04). This cage-like structure restricts the movement of the water surrounding the nonpolar particle. Clumping the nonpolar molecules together reduces the surface area that is required for the nonpolar molecules to interact with the surrounding water.
• In the case of phospholipids, when the nonpolar lipid tails are sequestered away from the surrounding aqueous environment, they don’t come into contact with water at all. Thus, the water molecules are not constrained and are able to freely hydrogen-bond with the polar head groups and/or other polar molecules. This net reduction of the number of “constrained” water molecules provides a large increase in entropy. Based on our equation above, all other things being equal, that increase in entropy can be enough to result in a negative change in free energy. Thus, it requires less overall energy to hold phospholipids in a bilayer formation, making it “energetically favorable.”

1b. The Protein Component of Membranes

The second major component of membranes is proteins, which should come as no surprise. Virtually everything in the cell is either made of proteins or made by proteins, so the importance of understanding how the structure of proteins impacts their function cannot be overstated.

We expect that you have learned about proteins in your general biology classes, so we approach this topic from that perspective. ( If you need a refresher, we suggest looking at the material in the introduction. ) However, even if you have learned about proteins before, there is much more to learn. Proteins are important enough that you should expect to cover them in several courses, with increasing levels of detail and complexity. In this textbook we will be exploring protein structure and function, and how it specifically relates to membrane structure and function, in Topic 2.3.

General Features of Biological Membranes The Membrane Is Selectively Permeable

We say that membranes are selectively permeable because experiments show that they allow some molecules to pass through while excluding others. However, it has also been shown that the properties of different membranes vary…a lot! Some are very permeable to ions and water, while others allow almost nothing through.

The chemical composition of the membrane, specifically the lipids that make up the bilayer, is a large factor in the membrane’s capacity to be selectively permeable. Due to the primarily nonpolar environment inside the core of the bilayer, it is extremely unfavorable for water or other polar molecules to spend any time in there; thus, the interior of the membrane is almost entirely a water-free zone. This does not necessarily mean that small polar molecules (like water) never enter the core of the bilayer, but if they do, they exit again quickly. This is the basis for the reduced permeability of membranes for polar molecules.

To further explore selective permeability, we will look at the simplest example—a synthetic bilayer made entirely of phospholipids. In biological membranes, there are a variety of transport proteins , which allow molecules to pass through the membrane. Looking first at a pure phospholipid bilayer can help us understand why some molecules need transporters, whereas others do not.

• Depending on the properties of the molecule in question (size, polarity), different molecules will have more or less difficulty crossing through that nonpolar portion of the membrane. Small nonpolar molecules (like oxygen, carbon dioxide, and nitric oxide) can cross a lipid bilayer easily, without any help.
• Polar molecules that are very small (like water) are also able to cross bilayers, but not very well. As such, the process is slower, and many membranes have transport proteins to facilitate passage when necessary.
• Larger molecules (amino acids, nucleotides, and glucose) will struggle to cross the bilayer as a result of their size. These molecules can also be polar, which can add to the challenge.
• Charged molecules, regardless of size, will be completely unable to pass through the bilayer without help. This is the basis for many of the electrochemical gradients that the cell uses to drive work in the cell.

Most biologically important molecules are either too big (like glucose) or too charged (like ions) to pass through a membrane spontaneously. Transport proteins help these molecules to go across a membrane even if their chemical properties prevent them from diffusing spontaneously. The selective transport of molecules may or may not require the input of energy, like ATP, or rely on concentration gradients. The Amoeba Sisters have an excellent review video (Video 02-02) that explores the selective permeability of the membrane and how transporters regulate molecules going in and out of the cell.

Having control over which molecules are able to cross a given membrane is a key factor in how cells and organelles function. For example, the mitochondria’s capacity to generate ATP is almost entirely dependent on the formation of a concentration gradient (discussed in Video 02-02, above). Concentration gradients are used extensively by the cell to do work. They are analogous to a hydroelectric dam. A dam holds water in an upstream reservoir and controls the flow of water through turbines to create energy. In this analogy, the membrane is the dam, and specific ions are held at different concentrations on either side. As the ions naturally “flow” from the side of the membrane that has a high concentration to the side that has a low concentration, energy is released. The release of the molecules down the gradient is often used to “power” other cellular processes and thus is a key part in understanding advanced physiology. In this textbook, due to the metabolic pathways that we focus on, we mostly discuss the formation of proton gradients (i.e., H + ions), but other electrochemical gradients are equally important to proper cellular function.

General Features of Biological Membranes The Membrane Is Organized but Fluid

Lipids have a key role to play in the organization and fluidity of a membrane (a key concept). Proteins in the membrane, while also organized, may or may not be fluid, depending on whether they are anchored to internal or external structures. First, we will explore how we define fluidity in this context.

There are two kinds of fluidity in membranes that must be considered:

• lateral motion of lipids within a single leaflet of the bilayer and
• overall “stiffness” of the membrane.

Let’s look at each of these briefly in turn.

Lateral Motion of Lipids within a Single Leaflet Is Far More Favorable Than Movement between Leaflets of the Bilayer

The hydrophobic effect keeps the nonpolar portions of membrane lipids in the center of the membrane and is the reason membranes form spontaneously. As long as the nonpolar portions of the phospholipids stay in the interior of the membrane, their movement is not restricted (Figure 02-05). Within a membrane or bilayer, individual phospholipid molecules can spin rapidly on their axis and/or diffuse laterally within the plane of the bilayer. On the other hand, they very rarely flip-flop across the bilayer, as it would require the polar head group to pass into the nonpolar center of the membrane.

To further illustrate this point, here are some data to convince you:

• the frequency of lateral shifts of a phospholipid within the same layer is roughly 10 6 /sec, but
• the frequency of flips from one leaflet to the other is closer to 10 –5 /sec.

Thus, flips between bilayer leaflets are about 10 11 times less frequent than lateral movements within a phospholipid layer. Based on these numbers, an individual phospholipid would “flip” to the other leaflet about once every 28 hours on average. Thus, when we discuss membrane fluidity, we are generally referring to the lateral mobility of membrane components within a single leaflet of the membrane. It is also why we often refer to membranes as two-dimensional fluids .

Functional biological membranes require some movement of their lipids. The movement of lipids allows for rapid resealing of membranes in response to small holes or tears. However, too much fluidity can be damaging. A membrane that allows too much motion might have trouble keeping everything in its correct location. Think about a boiling pot of water, where it’s boiling so hard that water is spilling out everywhere. Membranes with too much fluidity may become holey or lose vital parts, which will make it challenging for the membrane to maintain its functionality.

Membrane Composition Also Determines How “Stiff” a Membrane Is

Membrane stiffness refers to how pliable or bendable the membrane is. It is a separate but equally important component of membrane fluidity, which is determined by the composition of the membranes. If membrane lipids pack too close together, the membrane will freeze in place, and the result will be a membrane that is too rigid to adapt as the cell moves and changes. On the other hand, a membrane that is too pliable will struggle to maintain the shapes required. Thus, the cell needs to manage the fluidity of its membranes in more ways than one. The cell manages this by controlling the precise lipid composition of the membrane. Remember that within each grouping of membrane lipids (phospholipids, glycolipids, sphingolipids) is a large family of similar molecules. By changing which phospholipids (or sphingolipids, etc.) and how much cholesterol is in the membrane, the cell can maintain its membranes within the correct range. That way they will be fluid enough to allow movement and bending but not so fluid that the cell struggles to control function. In Topic 2.2, we will look more deeply at how this fine balance is maintained by the cell in all kinds of environmental conditions.

General Features of Biological Membranes The Membrane Is Asymmetric

Since membranes are exposed to different compartments and different environments on either side, it stands to reason that one would not expect the membrane, or its composition, to be identical on either side as well. This phenomenon is referred to as membrane asymmetry. There is always an inside that faces the cytoplasm (cytoplasmic side) and an outside facing the interior of an organelle or, in the case of the plasma membrane, the external environment (extracellular side). Both the lipids and the proteins play an important role in membrane asymmetry.

Figure 02-06 shows how the composition of lipids typically differs in the two leaflets of the plasma membrane. Remember that the term phospholipid refers to a whole family of lipids, so even though leaflets are made of phospholipids, the exact composition may not be the same. In addition, there are other kinds of lipids (e.g., sphingolipids and glycolipids) that might be enriched in one side of a membrane versus the other. For example, glycolipids are almost never found on the cytosolic side of the plasma membrane. Glycolipids face the cell exterior, as they play key roles in extracellular signaling and cell identity. Within the cell, organelle identity is also signaled by specific membrane lipids known as phosphatidylinositols, or PIPs for short. PIPs face the cytosolic side of the membrane and are used extensively by signaling and vesicle transport machinery.

The proteins in a membrane are also asymmetric. The parts that are sticking out of each side of the membrane will naturally be different to accommodate their specific functions. For example, a receptor protein will need to have the signal-binding site on the exterior surface of the membrane where the signaling molecule is present. Thus, each membrane protein needs to be inserted into the membrane in a very specific way to preserve proper function. Thus, the two sides of a membrane are not interchangeable, and the cell must keep them properly oriented at all times. We will discuss how this orientation is maintained in specific processes, like vesicle trafficking, in Chapter 4 .

For Many Membranes, Carbohydrates Are Also Major Contributors to Membrane Asymmetry

At this stage, it is essential that we take a moment to point out the role of carbohydrates, the fourth major biological macromolecule. While they sometimes get overlooked in introductory cell biology courses—simply because there are so many other things to cover—they are absolutely essential to proper cellular function. The plasma membrane of animal cells is usually completely covered with carbohydrates—most commonly attached to proteins but also to lipids. This coating (called the glycocalyx ) plays a key role in cellular identity and signaling. Plants and fungi have so many carbohydrates surrounding them that they form a structure known as a cell wall . We will take a moment at the end of this chapter to explore both the extracellular matrix and plant cell wall in a little more detail.

The Origin of Membrane Asymmetry

While it’s easy to understand the need for membrane asymmetry, it is much more complex to understand how asymmetry is created by the cell in the first place. In order to understand that, we must explore how membrane lipids are made by the cell.

All membrane lipids are first synthesized in the cytosol and then inserted into the cytosolic face of the smooth ER. This means that all new lipids are being added to a single leaflet of the ER membrane. As you can well imagine, this causes pressure on the ER membrane, as one leaflet is growing while the other is not. If left this way, the pressure would eventually cause a potentially catastrophic rearrangement as the membrane returns to a more thermodynamically favorable conformation. To avoid this kind of event, the cell uses proteins known as scramblases to rebalance the lipids in each leaflet (Figure 02-07A). This is done nonselectively—the scramblase simply equilibrates the phospholipids on either side of the membrane regardless of type. Later, in other regions of the ER, but more commonly in the Golgi, the membrane is organized more precisely. Enzymes called flippases move specific phospholipids unidirectionally from the exterior leaflet to the interior one so that membrane asymmetry is properly established, and the lipids are placed where they need to be for function. Both enzymes use a similar mechanism—the enzymes are big enough to span the membrane, and the polar head group is slid through a slot in the enzyme so it can cross the nonpolar part of the bilayer (the tails stick out and are left free to the environment; see Figure 02-07). Thus, they can overcome the energetically unfavorable parts of the reaction and flip lipids from one leaflet to another.

Topic 2.2: Maintaining Fluidity in the Membrane

• Describe how the lipid composition can influence membrane fluidity.
• Explain how membrane fluidity is maintained under different environmental conditions, like temperature.
• Explain how fluorescence recovery after photobleaching (FRAP) works, and interpret outputs from these experiments.
• Compare and contrast membrane-adjacent extracellular components in plants (i.e., the plant cell wall) and animals (i.e., the extracellular matrix).

In the first topic of this chapter, we focused on the major characteristics of biological membranes as a whole, and how the chemistry of the membrane can contribute to its properties. Here we’ll go into more depth about how the fluidity of a given membrane is the product of its precise lipid composition. We will also learn about an experimental technique known as fluorescence recovery after photobleaching (FRAP) , which allows us to assess the mobility of molecules within the membrane.

in depth Membranes Are Organized and Fluid

When we think of the term membrane fluidity , the first thought might be that this simply indicates how quickly any given molecule can move from point A to point B within the membrane. However, this is only one aspect of membrane fluidity. As we mentioned in the previous topic, the stiffness of the membrane must also be considered.

Membrane stiffness describes how easily the membrane bends (compared to how easily or quickly the molecules within the membrane can move around). Membranes vary in how stiff they are and how they respond to their shape being distorted. For example, red blood cells are elastic, which means they can be distorted but bounce back to their original shape. On the other hand, white blood cells are much more rigid and don’t deform as easily. As an analogy, a red blood cell is like a plastic bag, and a white blood cell is more like a cardboard box. This is the result in differences in both the lipids and proteins that are associated with each cell’s plasma membrane. The composition of the membrane is key for the cell’s capacity to function in its environment.

Membrane Fluidity Is Heavily Influenced by the Environment and Must Be Maintained at a Constant Level

Organisms live in all kinds of environments, from the hottest regions of the Sahara to the coldest regions of the Arctic and everything in between. They also live in regions of extreme pressure, like the deep sea, which can impact both the freezing and boiling point of water. In all of these extreme environments, cells must control the chemistry of their membranes so that they remain fluid and functional. For example, membranes must not freeze when exposed to low temperatures (or high pressure) or lose their integrity at higher temperatures. Any organism that lacks a mechanism to control the fluidity of its membranes will die if/when environmental conditions change.

There are two major approaches that organisms take to controlling the fluidity of their membranes:

This section naturally focuses on exotherm membranes, whose membrane composition changes more in response to the environment. One good example of membrane composition being used as an environmental adaptation is cold-hardiness in plants. As the temperature lowers in the fall and winter, membrane lipids become increasingly unsaturated. This lowers the melting/freezing point of the membrane, which helps it maintain flexibility, and retains the mobility of the membrane components. This also explains why organisms might be fragile with respect to sudden, unexpected changes in temperature. In the example of the plant, a rapid drop in temperature at the wrong time (like a cold snap in early autumn) could result in the death of the plant, as they wouldn’t have the time required to synthesize new membrane lipids (and other antifreeze proteins, which help avoid freezing).

Membranes That Are Unable to Adapt Chemically Are Susceptible to Failure in Nonoptimal Temperatures

The capacity of a membrane to remain fluid at any given temperature is primarily dependent on its capacity to maintain just the “right” amount of entropy in its phospholipid tails. If they pack too close together, they will freeze, and if they have too much motional freedom, they will fail in other ways.

When a membrane (or anything else) freezes, what really happens is that as the temperature lowers, the kinetic energy of the molecules is also lower, so they become sluggish and do not move around as freely. At some point, the kinetic energy of the phospholipids is so low that they get drawn into the transient induced dipole–induced dipole interactions that continue to happen between neighbors and can no longer break away. Eventually, the phospholipids will get “trapped” by the intermolecular forces being exerted on them by their neighbors, and the lipids will freeze in place. In chemistry terms we say that the membrane is in “gel phase.” You can see the result of this any time you pull a piece of frozen meat from the freezer. It is quite stiff and inflexible compared to its room-temperature counterpart.

At the other extreme, when the temperature is too high for the membrane, the molecules have quite a lot of kinetic energy. If the kinetic energy is too high, then the intermolecular forces that hold the membrane together are too easily broken, and molecules can break free all together. When this occurs, important complexes will come apart, ions and other molecules may be able to slip through the chaotic membrane lipids, and the membrane itself could just fall apart.

Maintaining Fluidity—Membrane Composition

Cells have three different ways that they can modulate in order to influence the fluidity of a given membrane.

Change the Degree of Unsaturation of the Hydrocarbon Chains

Increasing the unsaturation of the lipid fatty acid tails increases fluidity of the membrane. Unsaturated lipids contain one or more double bonds between carbons in the hydrocarbon chain (compared to saturated lipids that have only single bonds in their hydrocarbon chains). The double bonds change the shape of the tail because they create bends, or “kinks,” in the hydrocarbon chain (Figure 02-08). A kinked tail will not be able to pack as tightly with its neighbors. Imagine that on a packed subway everyone stands wide-legged, with their elbows sticking out. The net effect would be that each person would take up more space, and fewer people would be able to get into each car. The net effect of unsaturation is essentially the same—less tight packing, which effectively reduces the number of induced dipole – induced dipole bonds between lipids in the membrane. The result is that the lipids are able to maintain their mobility even at the reduced kinetic energies of lower temperature.

Change the Number of Carbon Atoms in the Fatty Acid Tails

Change the sterol content of the membrane.

Cholesterol is a type of sterol (ringed lipid) that sits in the spaces between fatty acid chains (Figure 02-09). At first glance, based on what we have previously learned about creating space to reduce bonding, one would expect that this would be beneficial at low temperatures, as it would reduce the interactions between the fatty acid tails of the phospholipids. This is exactly true. At lower temperatures, increasing the amount of cholesterol will also increase membrane fluidity. But in this case, the story doesn’t end there.

Paradoxically, cholesterol also has an effect at higher temperatures. However, in this case, increasing cholesterol content will reduce the fluidity of the membrane. It does this by filling the space in between lipids that is created when lipids have high kinetic energy. They also have a stiff ring structure, which also contributes to the stiffness of the membrane. Cholesterol at high temperatures increases the number of intermolecular forces between neighbors, which is important to stabilize the lipids in the membrane.

Think of cholesterol as a fluidity buffer. It helps keep the membrane within a specific range of fluidity so that it doesn’t get too rigid or too fluid. Modulating the cholesterol can help control fluidity at many temperatures, which makes it an important membrane lipid in all environments. ( Note: Cholesterol is unique to animals. However, sterols are an important component of all eukaryotic membranes. Even though plants do not make cholesterol, they do use other phytosterols for the same purpose.)

“Real” Membranes Contain Regions with Varying Fluidity

While it might be tempting to think of the cell membrane as a homogenous thing that is the same across the entire cell, there is significant evidence that this is not true. Both the lipid and protein components of any given membrane will be different in different regions, as different functions will need to occur. For example, lipid rafts are distinct regions of the membrane that contain a higher concentration of sphingolipids with longer fatty acid tails and cholesterol (Figure 02-09). The increased length of the sphingolipid tails in that region helps hold the lipids together and also increases the thickness of the membrane in these regions. Increased cholesterol in the region offsets the tendency toward decreased membrane fluidity that would result from the longer tails. Specific proteins are preferentially drawn to these regions based on their chemical features (e.g., longer membrane-crossing domains). The specific clustering of proteins in lipid rafts can be used to keep proteins together that all work toward the same function (e.g., signaling or vesicle trafficking). A diagram of a lipid raft is shown in Figure 02-10.

Studying Cells: Fluorescence Recovery after Photobleaching (a.k.a. FRAP)

One of the experimental tools that is used to examine how cellular components move is a fluorescence microscopy technique known as FRAP. While it can be used to study any number of cellular functions that involve movement, in this case we will focus on how it can be used to track movement within a membrane.

In this technique, scientists fluorescently label a specific membrane component (usually a lipid or protein). Then the high-powered laser of a confocal microscope is used to photobleach the fluorescent tag, meaning that the fluorescence of the tag is extinguished without damaging the protein or lipid to which it is attached. A small, specific area of the cell is bleached, and then we track how long it takes for unbleached molecules to return to the bleached area. There is an excellent explanation of this in Video 02-03.

The video also shows us the kinds of results we can expect from these experiments. To summarize, graphs are generated from these data, with the fluorescence intensity in your region of interest on the Y-axis and time on the X-axis. Based on the amount that the fluorescence “recovers”—meaning the fluorescent intensity in that region increases again—researchers can compare how mobile a particular lipid or protein is within the membrane. We call this graph a recovery curve , as it shows how the fluorescence “recovers” in the region of interest over time.

Topic 2.3: Structure and Function of Membrane Proteins

• Explain how the primary sequence and the environment of a protein influence its final 3D structure, specifically with respect to the different types of intermolecular forces that it will form with itself and its environment.
• Distinguish between integral and peripheral proteins with regard to their solubility properties, structure, and manner of attachment to membranes.
• Describe some roles of glycoproteins within a biological membrane.
• Describe the structure and use of plasma membrane–adjacent structures like the animal cell extracellular matrix and the plant cell wall.

Earlier in this chapter, we learned that a biological membrane is different from a phospholipid bilayer. A biological membrane also has proteins embedded in it that will change its properties and add functions. In this topic, we will first review the formation of proteins from amino acids, then discuss the characteristics of membrane proteins. We will also examine the unique properties of the plasma membrane as the barrier that separates the cellular contents from the outside world. We end with a brief discussion of what’s on the outside of a cell—namely, the extracellular matrix (animals) or a cell wall (plants, algae, and fungi).

The protein content of a membrane can range from ~25% (in the case of the myelin sheath cells of neurons) to nearly 80% (for the mitochondrial inner membrane), with more typical membranes consisting of protein and lipids in an approximately equal ratio (50:50) by weight.

Because of the associated proteins, the thickness of biological membranes is almost always thicker than that of a simple lipid bilayer. Biological membranes are typically 6.5 to 10 nm thick. A lipid bilayer without proteins is about 5.5 nm thick.

Brief Review of Amino Acids and the Chemistry of Protein Folding

We assume you have preexisting knowledge about amino acids and proteins, so we are approaching this as a review. As always, we encourage you to explore the resources in the introduction if you need a refresher.

There are 20 different types of amino acids that are used to form proteins. Each amino acid contains an amino group (NH 3 + ) and a carboxylic acid group (COO – ). In between is a carbon atom connected to a variable chemical structure called the R group or side chain . To make a protein, the amino group from one amino acid will covalently bond with the carboxylic acid group of another amino acid to create a peptide bond . This creates a repeating N–C–C pattern when the amino acids are strung together. This repeating N–C–C is called the protein backbone .

The sequence of the amino acids in a polypeptide is based on the sequence of codons that are read from the mRNA by the ribosome during translation. We call the order of the amino acids in the polypeptide chain the primary structure or primary sequence.

The R groups are what give each amino acid its distinct chemical properties. Some R groups are large, and some are small; some are negatively charged at physiological pH, and some are positively charged. It’s worth taking some time to look at the chemical structure of the R groups and start investigating how these chemical groups impact protein structure/function.

The order of the amino acids in the primary structure plays a crucial role in determining folding and function of a protein. A given protein will fold in a very specific way depending on the molecular interactions of the amino acids in its primary structure. A variety of intermolecular forces (i.e., hydrogen bonds, ionic bonds, and induced dipole–induced dipole/van der Waal forces) contribute to the final 3D fold of the protein. Proteins will, many times, spontaneously fold into the shape that is the most stable and requires the least amount of energy to maintain. Other times, chaperone proteins will help facilitate the proper folding of a protein. We will learn about chaperones in more detail in Chapter 4 .

When discussing protein folding, we split the different interactions into four categories, which we call the “levels” of protein folding. It’s important to remember that these do not happen sequentially. Instead, they happen more or less all at once in different parts of the protein as it folds into its final 3D shape. These levels are as follows:

• Primary structure forms when amino acids are covalently bound together by peptide bonds. The order of the amino acids is the key feature of the primary structure.
• Secondary structure forms when the backbone interacts with itself via hydrogen bonds. It forms a repeating local pattern, commonly with nearby amino acids. Examples of this structure include alpha helices and beta sheets .
• Tertiary structure forms when R groups get involved. They interact with either other R groups or the backbone, usually via intermolecular forces like the ones mentioned above.
• Quaternary structure forms when different polypeptide chains come together to form a protein (or protein complex). All of the same intermolecular forces get used as we would expect in tertiary structure.

Disulfide bridges are the only other covalent bond used in folding aside from the peptide bond that creates the primary structure. Disulfide bridges form between two cysteine residues. They can play a role in tertiary or quaternary structure but are far less common than the other bonds. To compare, there are usually only a handful of disulfide bridges in a protein, if any, whereas there may be hundreds, or even thousands, of hydrogen bonds. As a result, we tend to discuss disulfide bridges much less often, but that doesn’t mean that they should be forgotten.

The chemical environment (especially pH) also plays a very important role in protein folding. A soluble protein in the cytosol will experience different intermolecular forces than a protein that is embedded in a membrane. In each case, the protein will respond to its folding environment. Unlike the soluble proteins of the cytosol, many membrane proteins must be able to embed themselves in the lipid bilayer, including the nonpolar tail region. This requirement impacts how the protein folds and which amino acids of its primary structure are expected to be on the exterior in each region.

Integral and Peripheral Membrane Proteins

Proteins can be associated with the membrane in multiple ways. At the most fundamental level, proteins may

• exist solely on the surface of a membrane, living peripherally on one side or the other, or
• extend into the nonpolar tail region of the membrane in some way so that they are integrated into the membrane itself.

This may sound like a very simple thing to differentiate: Do they exist on the surface or extend into the membrane itself? However, it isn’t always as easy as it sounds. For example, how do we categorize a protein that sits on the surface but also has a covalently linked lipid tail that extends into the membrane? ( Answer: We consider it to be integral.)

Since proteins in membranes are too small for us to see, the categorization of proteins has historically been based on experimental evidence. So if we include the experimental evidence in our definition, we come up with the following:

• Integral membrane proteins are proteins that cannot be removed from the membrane without destroying the membrane completely . Usually, this requires the use of strong detergents, which disrupt the structure of the membrane so that proteins can be removed. Only proteins that extend into the membrane in some way will require this level of disruption to extract.
• On the other hand, peripheral proteins are much easier to remove. Usually, a simple ionic salt wash is enough to dislodge these proteins from the membrane, as the intermolecular forces are not as strong and can be more easily disrupted.

Figure 02-11 also helps illustrate the different subcategories of membrane proteins:

• Transmembrane proteins (Panel A): These integral membrane proteins cross the entire membrane and stick out on either side into the cytosol. These proteins are held in place primarily through the hydrophobic effect. More on that in the section below.
• Monolayer-associated proteins (Panel B): This is also a type of integral membrane protein. Although they are not considered to be very common, they still occur rarely. In this case, the protein is held in place by an alpha helix that is amphipathic (i.e., nonpolar on one side only). We will not be discussing this type further.
• Peripheral membrane proteins (Panel C): These proteins are most commonly (but not always) attached to the membrane through associations with integral membrane proteins. These linkages are most often via ionic or hydrogen bonding or some combination of several types of intermolecular forces. The association of peripheral proteins with one side of the membrane or the other further contributes to the asymmetry of the membrane.

Transmembrane Proteins Require Specific Secondary Structure to Be Able to Pass through the Membrane

One of the biggest challenges faced when proteins must pass through a membrane is how to deal with their backbones. Unlike the composition of R groups, the backbone of the polypeptide chain is always polar. The H, N, and O atoms that make up the peptide bond are electronegative and capable of forming hydrogen bonds. As such, they face a thermodynamic challenge when required to pass through the hydrophobic center of the lipid bilayer. So how is this addressed in the cell?

The answer lies in their secondary structure. As you may recall, secondary structures are defined as local, repeating structures that are formed via hydrogen bonding of backbone atoms to other backbone atoms. These repeating structures usually have a twofold effect:

• they allow the protein backbone to form hydrogen bonds with itself, which is thermodynamically stable, and
• they usually push the R groups outward, where they are available to interact directly with the environment. This leaves the protein backbone in the center, sequestered away from the nonpolar lipid environment.

The two secondary structures we mentioned earlier are the alpha helix and the beta sheet. Both of these structures are very commonly found in transmembrane proteins. However, the details of how they form are naturally going to be different.

Alpha Helices in Transmembrane Proteins

Many transmembrane proteins have one or more alpha helices in their transmembrane domains. These alpha helices are stretches of about 20+ nonpolar amino acids (depending on the width of the membrane). Remember that in the alpha-helical arrangement, the amino acid R groups extend outward, and the backbone is in the center of the helix. By forming an alpha helix, the nonpolar R groups will shield the polar polypeptide backbone from the nonpolar environment in the center of a lipid bilayer, thus creating thermodynamic stability. Figure 02-12 shows an alpha helix from the side, embedded in a membrane. You can see the backbone structure represented by the purple ribbon winding up through the membrane. Hydrogen bonds between the atoms of the backbone (shown as dashed lines between the ribbon) hold the shape of the helix. The R groups of each amino acid in the polypeptide chain extend outward from the polypeptide backbone to interact with the tails of the phospholipids.

Since transmembrane proteins must be able to exist in both the nonpolar center of the membrane and the aqueous parts of the cell, it stands to reason that these proteins are amphipathic like other components of the biological membranes. The R groups of amino acids within the nonpolar portion of the membrane are nonpolar, whereas in other regions that interact with the aqueous environments, the amino acids will primarily be polar.

It is important to note that a single alpha helix does not form a channel, so it cannot allow anything to pass through it. The molecules of the backbone completely fill the space inside the helix (Figure 02-13). Alpha helices can be used to form the pores and channels that control the entry/exit of many molecules in/out of the cell. Because a single helix cannot act as a channel, several membrane-spanning alpha helices must instead cluster in a roughly circular arrangement through the membrane.

Aquaporin is an excellent example of a transmembrane protein that is made of several alpha helices, which are used to create a central channel (Figure 02-14). Aquaporin is nonselective and allows water and other small solutes to pass through the membrane. Note that the amino acids lining the interior of the channel will need to be polar and possibly charged in order to interact with the water and solutes that are expected to pass through. On the other hand, the exterior of aquaporin will have a large strip that will directly interact with the phospholipid tails of the membrane; thus, those regions will need to be nonpolar.

Beta Barrels in Transmembrane Proteins

While a single alpha helix is stable enough to be used as the sole membrane-spanning structure for a transmembrane protein, a beta sheet is not as well designed for this purpose. The edges of the sheet will have exposed backbone molecules that will not easily interact with the nonpolar portion of the bilayer. However, a beta sheet can circularize itself by hydrogen bonding the ends of the sheet together (Figure 02-15). This means that the backbone’s bonding requirements are met within the nonpolar region of the membrane, and we once again have an ideal scenario for a thermodynamically stable structure that can pass through a membrane. We call these circularized beta sheet structures beta barrels .

Like the alpha helix, beta sheets and beta barrels require a specific arrangement of their R groups in order to form. Unlike the alpha helix, the R groups stick out perpendicular to the face of the beta sheet in an alternating pattern (Figure 02-15). This means that if we were to unfold the beta barrel and examine the primary sequence, we’d likely see a series of amino acids with alternating properties (i.e., nonpolar, polar, nonpolar, etc.) because one amino acid with an R group facing the lipid environment would be next to an amino acid with an R group facing the aqueous pore environment.

We can see this in action by examining the beta barrel structure of a real bacterial protein called outer membrane protein G (Figure 02-16). Once again, we see that this particular secondary structure, the beta barrel, precisely suits the function of the protein as a transporter. This protein is used to take up large carbohydrates. Thus, the outside of the beta barrel must be able to interact with the lipid environment and help hold the protein in the membrane, whereas the inside of the barrel must create enough space to allow specific molecules to pass through the membrane. In addition, the amino acids of the center channel must be polar as well to interact with the carbohydrates that are transported. In summary, the properties of the amino acids must match the chemical environment where they reside and function.

Function of Membrane Proteins

Membrane proteins carry out many different functions in the cell. It is important to remember that membranes in the cell have different sets of proteins in them, as they each must carry different functions. As such, the protein composition of the endoplasmic reticulum (ER) membrane is different from, say, the plasma membrane.

While there are many different functions for membrane proteins, for the most part they’ll fall into one of the following categories (Figure 02-17):

• Structural proteins, such as linkers and anchors— Anchors help attach the membrane to organelles, the extracellular matrix, or even other cells, whereas linkers help connect several proteins in the membrane and can help provide shape. We will see many examples of structural proteins as we discuss cellular function.
• Transporters —These proteins mediate the transport of different types of molecules across the membrane in either direction. We saw examples of these in our earlier discussion of alpha helices and beta barrels. We will look at some examples, such as proton pumps and the ATP synthase.
• Enzymes —Many membrane proteins have enzymatic activity for a whole variety of cellular functions. We will explore many examples, including synthesizing or modifying enzymes, flippases, scramblases (shown earlier, in Figure 02-07), or kinases.
• Receptors —Receptors are key for the cell to be able to sense and respond to its environment. Receptors extend across the membrane and bind to small molecules or other proteins on the outside of the cell and in response initiate a chain of events leading to transmission of a signal inside the cell. Many receptors are also enzymes. We will look at receptors in more detail when we discuss cell signaling.

The Plasma Membrane: An Example of a Real Biological Membrane

The plasma membrane is the membrane that surrounds the cell and is the first point of contact between the cell and its environment. Red blood cells were one of the first cells to have their plasma membranes studied and as such have one of the most well-characterized plasma membranes. In most cell types, the plasma membrane is supported and shaped by a network of proteins. In red blood cells, most of the protein network is inside the cell, just underneath the plasma membrane. However, in other cell types, this network can also be on the outside (via connections to the extracellular matrix or cell wall) or on both sides of the membrane.

Figure 02-18 shows a diagram of the plasma membrane of a red blood cell. In it, we see that the lipid bilayer of the plasma membrane is attached to an internal meshwork of spectrin protein filaments. These filaments provide a framework that supports the membrane and gives it its elasticity. You can also see that lipids and proteins on the extracellular side of the membrane are connected to carbohydrates (as indicated by the glyco- prefix). Since red blood cells move through the bloodstream, spectrin does not make any permanent connections with external structures.

Plasma Membrane Carbohydrate Groups Are Found on the Outside of the Cell

Something that is important to point out in the plasma membrane figures in this chapter (see Figures 02-06, 02-10, and 02-18 as examples) is that the exterior of the plasma membrane is usually covered in a coating of carbohydrates. These carbohydrates are most commonly in the form of glycolipids and glycoproteins that are integrated directly into the plasma membrane. The carbohydrate component of the plasma membrane has a very important function. In biological systems, cells have an identity (and can be recognized) on the basis of the configuration of carbohydrate molecules on the surface of the membrane.

A classic example of this is your blood type. The cell surface polysaccharides carried by your red blood cells are genetically determined so that your body knows which cells belong to you and which are foreign. The ability to identify foreign cells is absolutely vital for your body so that it can recognize pathogens and other invaders. This also influences our ability to carry out blood transfusions when needed. Your “blood type” makes direct reference to the polysaccharides carried on the surface of your cells. The ABO system is the most well known; however, there are other cell surface signals that the body uses to identify which blood belongs to you, such as the Rhesus (Rh) factor. If you receive a transfusion of blood that contains the incorrect polysaccharide markers on the cell surface, your body will produce antibodies to attack the blood and destroy it.

In some cell types (such as bacteria but also many eukaryotic cells), this carbohydrate coating on the plasma membrane is complex enough that it has its own name: the glycocalyx . Most examples of eukaryotic cells that have a glycocalyx are found in animals. This includes many epithelial cells (like the cells lining the gut and our blood vessels). Another really great example of a glycocalyx, which you can actually see with your naked eye, is the slimy coating on fish. The polysaccharide coating is found on virtually all fish, including the sockeye salmon shown in Figure 02-19. It plays multiple roles, including protection, cell-to-cell recognition, and even immune functions.

Plasma Membrane–Adjacent Structures in Animal and Plant Cells

Cells in tissues are surrounded by a matrix of protein, polysaccharide, and fluid. The composition of the extracellular environment varies widely in different tissue types as well as in different organisms from different kingdoms. However, a common theme is that there are quite a lot of carbohydrates. The carbohydrates may or may not be associated with proteins, but collectively they form a three-dimensional network that connects cells together and provides a sort of hydration layer to trap water near the cells. Here we will highlight some of the key differences between this external environment of plant and animal cells.

In both plants and animals, the extracellular macromolecules are synthesized and secreted by the cells that live within them through a process called exocytosis, which we will cover in Chapter 4 .

The Extracellular Matrix of Animal Cells

The extracellular matrix (ECM) of animal cells is made primarily of proteoglycans . These are very similar to glycoproteins, except there is quite a lot more polysaccharide attached to a relatively small protein. The proteins and polysaccharides of the extracellular matrix are all connected to each other in a large 3D network, making it difficult to tell where one molecule ends and another begins. Polysaccharide chains known as hyaluronic acid are a major component of the animal extracellular matrix. Hyaluronic acid is important due to its gel-like properties. It helps trap water, which makes the entire ECM look and feel a bit like Jell-O. Thus, polysaccharides like hyaluronic acid help the extracellular matrix remain hydrated and, as such, resist compression. For this reason, it is a large component of the cartilage in our joints.

The most abundant protein in the mammalian extracellular matrix (not to mention the most abundant protein in the entire human body) is a very large protein called collagen (Figure 02-20). This protein is made of three intertwined polypeptide chains, which makes collagen very strong. Collagen provides structural support and helps with things like wound healing.

The Plant Cell Wall

The plant cell wall is also made mostly of carbohydrates. Interestingly, it contains many fewer proteins and is more rigid than an animal extracellular matrix. There is a cell wall that surrounds every single cell in a plant. So while it can be thought of as a type of extracellular matrix for the plant, its role is more complex than that of the animal extracellular matrix. It provides the structural support for the whole plant, similar to the role of the skeleton in animals.

Like the animal extracellular matrix, the plant cell wall is synthesized and secreted by the cell. Most of the polysaccharide components are synthesized by the Golgi apparatus; however, there is one notable exception. Cellulose is a very strong fiber of crystallized glucose chains that is synthesized in a unique structure called a rosette (Figure 02-21). This rosette is embedded in the plasma membrane, and it moves through the membrane as cellulose is synthesized. Microtubules have a role to play in the synthesis of cellulose and, as such, have a very different organization than we see in animal cells.

While there are a few proteoglycans and glycoproteins in the plant cell wall, the complex, branched polysaccharides are really the key players in both its structure and its function. In addition to cellulose, there are other long polysaccharide chains, such as hemicellulose (used to connect the cellulose together) and pectin. Pectin acts in a similar way to hyaluronic acid in that it helps trap water and create a gel matrix. Incidentally, this is also why we use pectin to make jam.

Topic 2.4: Putting It into Practice Determining Protein Location and Orientation

• hydropathy plots as an example of the bioinformatics approach to studying proteins
• gel electrophoresis and SDS-PAGE as an example of a biochemical approach to studying proteins

Biological research changed quite drastically in 2000, when the human genome was completely sequenced ( see the original publication of the human genome ). Since then, our access to genetic information from a variety of species has increased much faster than we could have possibly imagined. This has fundamentally changed how we do research and has revolutionized how we explore cell biology. We now do much initial work in silico (meaning inside the computer) to first make solid predictions about protein function based on recognizable patterns from their primary sequence. To test the predictions made in in silico experiments, we turn to laboratory techniques that manipulate the DNA and proteins directly. In this section, we will explore one bioinformatic technique and one laboratory technique as examples of the many ways that scientists explore biological questions about membranes and their proteins.

Technique 1: Bioinformatics Approaches and the Hydropathy Plot

The amount of DNA sequence information that is openly accessible on the web continues to increase. The data sets can be quite large and can be difficult to handle without the help of computer programs and algorithms to organize, align, and make predictions about structure, function, and subcellular location of unknown proteins. The field of computer-based analysis of DNA and proteins has become so important that it has its own name—we call this bioinformatics .

Protein folding is subject to the same underlying chemistry and thermodynamics as everything else in the universe. This means that as long as we know the primary sequence of a protein, we can make predictions about structure, which can also tell us important information about protein function. These days we use computer algorithms and modeling software to learn as much as we can about a protein before we start to experiment in the lab. In some cases, lab experimentation on a protein is not possible, so computer predictions are our only option. The number of online apps and tools that are available (many for free) also continues to increase as biological computer scientists continue to build tools to answer questions. There are far too many of these tools to reasonably cover in any detail, so we’ll focus on one of the most fundamental: the hydropathy plot .

A hydropathy plot is a bioinformatic tool that analyzes the sequence of amino acids in a protein and looks for specific patterns. Hydropathy plots can go by a few different names in the wider world (hydropathy index, hydrophilicity or hydrophobicity plots, etc.), but they all search for the same thing.

Specifically, hydropathy plots predict alpha-helical transmembrane domains . To do this, they look for a linear stretch of amino acids that are “hydrophobic enough” to allow them to exist stably within the confines of the nonpolar portion of the lipid bilayer. Usually, this equates to around 18–20 nonpolar amino acids in a row in the primary sequence. This is the approximate number of amino acids required to span the length of a typical membrane bilayer when adopting the alpha-helical arrangement. Video 02-04 explains hydropathy plots (called both hydropathy scales and indexes in this video) in more detail.

The output of the computer analysis of the primary sequence takes the form of a graph (Figure 02-22), where

• the x-axis is the linear sequence of the protein, listed from N-terminus to C-terminus, and
• the y-axis is a number known as the hydropathy index .

The hydropathy index is a measure of the average hydrophobicity of a given amino acid residue ( hydropathy score ) and the average hydropathy score of a certain number of adjacent amino acids on either side combined into one value. The hydropathy score of each amino acid is a constant that has been determined based on computer-generated and experimental data. You can think of it as a number that measures how difficult it would be to immerse that amino acid into water. A higher hydropathy score indicates that it takes more energy to immerse that amino acid in water, which indicates it is more nonpolar than other amino acids.

At each point on the graph that the line is above 0, the amino acid and its neighbors are considered to be nonpolar, and below 0, the amino acid and its neighbors are polar on average. If the peak is high enough and wide enough, then a transmembrane alpha helix becomes a reasonable prediction. In this textbook, the “threshold hydrophobicity” is marked on the graph by a dashed line, and anything that crosses that line is considered to be “hydrophobic enough” to be a potential transmembrane region. Note that only major peaks are counted, not minor ones.

It is worth emphasizing that the hydropathy plot can only make predictions about protein structure. In order to confirm the prediction, you would be expected to follow up your bioinformatic analysis with additional experimentation. It is entirely possible that even though this protein is predicted to be transmembrane, the reality could be quite different. The “transmembrane region” predicted here could actually be a region sequestered inside of a soluble protein, or it may have a completely different function entirely. A good scientist will always follow up computer-based predictions with experiments to prove that the computer predictions are true. We will see examples where hydropathy plot predictions are, in fact, false when we discuss the endomembrane system in Chapter 4 .

Technique 2: Biochemical Approach: Gel Electrophoresis

Despite the many advantages of bioinformatics approaches, there are also disadvantages that must be considered. For one thing, bioinformatics can only make predictions, and those predictions are limited to what we can learn from the amino acid, DNA, or RNA sequence. At some point, the scientist must test their bioinformatic predictions experimentally. One of the most commonly used techniques in cell and molecular biology is known as gel electrophoresis .

Gel electrophoresis is a way to separate macromolecules based on chemical features like size and charge. We can use gel electrophoresis to separate protein, DNA, or RNA (but not usually all at the same time!). In this technique, molecules are inserted into a gel matrix (not dissimilar from Jell-O but made from a different substance) that has pores of particular size. This gel matrix holds the sample and provides a filter to separate the molecules. The term electrophoresis refers to the electric field that is used to separate the molecules. Since the molecules carry a charge, the electric current causes them to move toward one side of the gel (see Video 02-05 below).

There are many different types of gel electrophoresis techniques available. Each one is named slightly differently to represent its unique features. Here’s a short list of the most common types:

• SDS-PAGE ( S odium D odecyl S ulfate P olyacrylamide G el E lectrophoresis) is used to separate proteins based almost exclusively on their molecular weight. This technique involves dissolving and denaturing proteins with the detergent SDS. This detergent adds a strong negative charge to the protein equal to the number of amino acids. This causes it to migrate to the positive pole during electrophoresis based on the size of the protein.
• Agarose gels can be used to separate nucleic acids. Nucleic acids naturally contain a strong negative charge on each residue, so detergents like SDS are not required. This consistent charge/mass ratio allows these molecules to travel to the positive electrode and separate based on size.
• Native gels do not denature molecules in contrast to SDS-PAGE. Thus, this technique exploits the natural charge carried by proteins, so the size, shape, and natural charge of the folded molecule also impact its movement in the gel.
• 2D gels separate molecules based on molecular weight and isoelectric point.

It’s worth noting that gel electrophoresis on its own doesn’t tell us much. It is merely the tool we use to separate the molecules in our sample. It is important to know how the samples are treated before they are separated on the gel. In addition, in many samples, there are lots of proteins mixed together. Thus, often an additional technique is required afterward SDS-PAGE to highlight and/or identify the proteins of interest in the gel (i.e., staining, antibody labeling, mass spectrometry, etc.). We will first show an example where we can use gel electrophoresis to study proteins using SDS-PAGE. Afterward we will briefly touch on how it can be used for studying nucleic acids in preparation for Chapter 3 .

Gel Electrophoresis to Study Proteins: SDS-PAGE

SDS-PAGE is one of the more commonly used types of gel electrophoresis. Its name comes, in part, from the molecules used in the technique. A polyacrylamide gel is a specific type of gel that can be used to separate proteins primarily. SDS is a detergent that helps denature and separate the proteins in your sample. First, a brief video (Video 02-05) shows how the process works.

As you can see, protein samples are first treated with the heat and the detergent SDS (Figure 02-23). The heat helps the proteins to unfold and become linear. The SDS then evenly coats the proteins, which helps them maintain their unfolded state and also adds a uniformly negative charge. Thus, all proteins end up completely unfolded, reducing any variation of travel based on the shape/compactness of their normal folding pattern. This makes it so that proteins are separated by their size and not some other characteristic.

Once the samples are heated and treated with SDS, they are added to the gel matrix using small wells that are set near the negative electrode (i.e., the cathode ). The electric current is then turned on, and the proteins travel toward the positive electrode (i.e., the anode ) at the other end of the gel. As the current runs through the gel, the proteins move slowly toward the positive electrode. Proteins that are larger will take longer to move through the gel matrix, and smaller proteins will travel farther through the gel. This way, the proteins in the sample are separated over time, with the larger proteins staying closer to the site or origin and small proteins moving farther away.

As mentioned at the start, SDS-PAGE on its own doesn’t tell us much. It is merely a way to separate the proteins in a sample by molecular weight. The experiment you want to do will always include treatments of your samples before you do SDS-PAGE. The animation below (Video 02-06) shows an example of an experiment that uses SDS-PAGE to determine the type of membrane protein found in a sample of plasma membrane.

DNA Gel Electrophoresis

While this chapter is not about nucleic acids and DNA, we wanted to add a brief discussion of how gel electrophoresis can be used to study nucleic acids, for two reasons:

• This is, by far, the most common usage of gel electrophoresis in modern times.
• The next chapter of this textbook looks at DNA and genomes in more detail, and you will need to understand how DNA gels work in order to be prepared for the techniques covered in Chapter 3 .

Gel electrophoresis for separating DNA by size follows many of the same principles as SDS-PAGE, but with some differences:

• SDS is not required, as the phosphates in the backbone of DNA create a uniform negative charge on the molecule that is proportional to its size.

Like SDS-PAGE, an electric current is run through the gel once the samples are loaded, and the DNA will travel toward the positive terminal. As a result, just like in SDS-PAGE, the DNA will separate based on the size of the molecule such that bigger molecules will stay closer to the origin and smaller particles will move farther away. Video 02-07 is an excellent explanation of gel electrophoresis using agarose gels for DNA. Compare this technique to what you learned about SDS-PAGE above.

As you explore the techniques in Chapter 3, don’t forget to come back here if you need to remember exactly how gel electrophoresis works.

 Chapter Summary

In this chapter, we’ve explored the structure and function of membranes. To do this, we first needed to explore some of the characteristics of the lipids and proteins that the membrane is made of. We identified four major characteristics of membranes:

• The membrane is a bilayer made up of lipids and proteins.

We also learned that the fluidity of the membrane is dependent on its lipid composition—namely, the length and degree of unsaturation of the phospholipid tails and the cholesterol content. After that we discussed the difference between integral and peripheral proteins and how the secondary structure of membrane proteins contributes to its ability to span the entire membrane, including the hydrophobic portion of the membrane. We rounded out our discussion of membranes by briefly exploring the plasma membrane, which is a unique membrane in the cell, as it is in contact with the external environment. Outside of the plasma membrane is an additional structure known as the extracellular matrix in animal cells and the cell wall in plants and fungi. While these two have some similar roles, they are not the same.

Finally, we explored several experimental techniques throughout this chapter that can help us learn about membranes: fluorescence recovery after photobleaching (FRAP) helps us measure the movement of membrane components; hydropathy plots are a bioinformatic tool that predicts membrane-spanning alpha helices from the primary sequence of a protein, and SDS-PAGE and other kinds of gel electrophoresis can be used to learn more about DNA, RNA, and protein.

Review Questions

Note on usage of these questions: Some of these questions are designed to help you tease out important information within the text. Others are there to help you go beyond the text and begin to practice important skills that are required to be a successful cell biologist. We recommend using them as part of your study routine. We have found them to be especially useful as talking points to work through in group study sessions.

• the polar region (differentiate between charged and uncharged molecules of the polar region) and the nonpolar region
• a region that would be stiff and inflexible
• a glycerol residue (Some molecules have a serine residue instead; does yours?)
• a region that could easily have C=C bonds added (How would that affect the structure of that region?)
• For the molecules you found in Question 1, identify the lipids that would be able to form lipid bilayers on their own. Use the details of the structures that you drew to explain why or why not.
• Define the four levels of protein folding and explain how each one is stabilized.
• Self-assembly of macromolecules is an important concept. What do you think that means?
• Are disulfide bridges covalent or noncovalent interactions? How do they form? Why are they considered to be uncommon in the cytosol?
• The portion of the amino acid that is common to all of them.
• The portion that is unique to each amino acid, known as the R group or side chain.
• The functional groups that will form the peptide bond. Will anything be lost/gained during that reaction? How do these parts relate to the “N” and “C” terminus of a protein?
• Any R groups that you would expect to have acidic properties. Will they gain or lose a proton during that reaction?
• Any R groups that you would expect to have basic properties. Will they gain or lose a proton during that reaction?
• Any R groups that would not be able to form H-bonds with water.
• R groups that would form H-bonds but would not be acidic or basic.
• R groups that could interact ionically with their neighbors.
• Any R groups you would consider to be “big” or “small” relative to the others.

Topic 2.2: The Lipid Bilayer

• What are the major differences between a synthetic phospholipid bilayer and a biological membrane?
• How do cells adjust their membrane composition to maintain fluidity of their lipid bilayers in varying conditions?
• Despite appearances, cholesterol cannot form bilayers on its own. Use the structure of the molecule to explain why.
• Explain how the structure of phospholipids is the basis of the major properties of the bilayers that they form: physical form of the bilayer, self-sealing property, selective permeability, and fluidity of the bilayer.
• Why are lateral movements of phospholipids in a bilayer so much easier than “flips” from one leaflet to the other? Explain why this means that we sometimes call a biological membrane a “two-dimensional fluid.”
• Describe how fluorescence recovery after photobleaching (FRAP) works and list the types of scientific questions that can be answered using this technique.

Topic 2.3: Membrane Proteins

• What is the difference between integral and peripheral membrane proteins? Discuss and compare the different strategies for association of proteins with membranes.
• What is a domain in a protein? How does it relate to structure and/or function?
• Compare the hydrophobic forces that hold a membrane protein in the lipid bilayer to those that help proteins fold into a unique three-dimensional structure.
• How do cells restrict the movement of membrane proteins? Outline the different strategies and provide brief examples.
• What feature of a membrane’s structure allows it to have separate identities and functions on each side?

Topic 2.4: Experimental Techniques in Membrane Biology

• What is a hydropathy plot, and how does it help predict transmembrane regions?
• Explain why hydropathy plots can only make predictions about transmembrane alpha helices and not beta barrels.
• Explain why experimental techniques are needed to verify predictions from hydropathy plots.
• Why is SDS needed for sample preparation in SDS-PAGE experiments?
• What types of questions can be answered using SDS-PAGE?
• How is DNA gel electrophoresis different from SDS-PAGE on proteins?

A selectively permeable barrier that is designed to separate the cell from the external environment, allows communication of activities between cells, and functions to form intracellular compartments. It is composed of lipids, proteins, and carbohydrates.

Composed of two sheets of phospholipids. The lipids are orientated so that the polar heads are facing outward and the nonpolar fatty acid tails are facing inward.

Refers to molecules that have one end of the molecule that is different from the other end. Most often, this refers to an uneven electron distribution, where extreme uneven distribution of electrons results in ionized molecules with a charge. Polar can also refer to shape differences on different ends of a molecule, as is the case with polar cytoskeletal elements.

Refers to molecules that have do not have one end of the molecule that is different from the other end. Most often, this refers to electron distribution. Nonpolar molecules share electrons evenly and never carry a dipole or charge. Nonpolar can also refer to the symmetrical shape of molecules, such as the tetrameric form of intermediate filaments.

“A class of lipids whose molecule has a hydrophilic ‘head’ containing a phosphate group and two hydrophobic ‘tails’ derived from fatty acids, joined by an alcohol residue (usually a glycerol molecule).” Source: Phospholipid. (2023, May 25). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Phospholipid

Generally refers to a lipid bilayer that has two lipids stacked tail to tail with the polar head groups facing the aqueous environment.

A chemical compound or molecule that possesses both a hydrophilic region and a hydrophobic region. Phospholipids that make up the majority of the structure of phospholipid bilayers are an example of an amphipathic molecule.

Latin for “water loving.” Refers to chemical groups that readily associate with polar molecules like water.

Latin for “water fearing.” Refers to chemical groups that do not readily associate with polar molecules like water. Instead, these are generally nonpolar in nature and group with other nonpolar molecules.

A single layer of a lipid bilayer.

“A small artificial vesicle, spherical in shape, having at least one lipid bilayer.” Source: Intermolecular force. (2023, June 22). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Intermolecular_force

“The observed tendency of nonpolar substances to aggregate in an aqueous solution and exclude water molecules.” Source: Hydrophobic effect. (2023, June 13). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Hydrophobic_effect

The Gibbs free energy equation. When the ΔG is negative, this indicates a spontaneous reaction and does not need extra energy input to occur.

A feature of biological membranes that allows only certain molecules and ions to pass through to the inside of the cell. Selective permeability is useful in controlling the composition of the internal cellular environment.

Refers to proteins in a membrane that facilitate the movement of molecules across a biological membrane.

A family of lipids containing a phosphate head, two fatty acid tails, and an inositol head group. These compose a small but functionally relevant component of the cytosolic side of a cell’s membranes.

“A glycoprotein and glycolipid covering that surrounds the cell membranes of bacteria, epithelial cells, and other cells.” Source: Glycocalyx. (2023, July 9). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Glycocalyx

“A structural layer surrounding some types of cells, just outside the cell membrane. It can be tough, flexible, and sometimes rigid. It provides the cell with both structural support and protection, and also acts as a filtering mechanism.” Source: Cell wall. (2023, March 30). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Cell_wall

“A network consisting of extracellular macromolecules and minerals, such as collagen, enzymes, glycoproteins and hydroxyapatite that provide structural and biochemical support to surrounding cells.” Source: Extracellular matrix. (2023, June 2). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Extracellular_matrix

A protein that will move lipids from one monolayer (leaflet) to the other in a lipid bilayer. They are not selective in direction and do not require an energy input.

“Transmembrane lipid transporter proteins located in the membrane that belong to ABC transporter or P4-type ATPase families. They are responsible for aiding the movement of phospholipid molecules between the two leaflets that compose a cell’s membrane (transverse diffusion, also known as a ‘flip-flop’ transition).” Source: Flippase. (n.d.). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Flippase

A method to determine the mobility of molecules within a membrane. A region of the cell is photobleached (the fluorescent molecules are destroyed). The same area is monitored over time to determine if fluorescence returns to the bleached area.

A lipid that has at least one double bonded carbon in its chain.

A lipid that has no double bonds within the hydrocarbon chain.

A molecule in the sterol family that plays a role in the fluidity of membranes.

A distinct membrane region that contains specialized lipids and proteins. Often these regions are a bit thicker than surrounding membrane areas and provide specialized functions like cell signaling.

Refers to fading. Often this is in reference to a fluorescence molecule that is no longer able to be seen.

The graph formed by a FRAP experiment. It measures the fluorescence intensity of a region over time to assess how much fluorescence “recovers” after the initial photobleaching step.

A variable chemical structure off the central carbon of an amino acid. The different chemical features of the R group define each amino acid and give it its overall characteristics (size, shape, charge etc.).

The repeating atomic structure formed when amino acids are joined through a peptide bond. These form N-C-C repeats referring to the central atoms in the amide, R-group, carboxylic acid groups that form each amino acid.

Refers to the sequence of the protein. Amino acids are joined into a polymer. The unique sequence of each polypeptide chain dictates how it will fold.

“The force that mediates interaction between molecules, including the electromagnetic forces of attraction or repulsion which act between atoms and other types of neighboring particles, e.g., atoms or ions. Intermolecular forces are weak relative to intramolecular forces—the forces which hold a molecule together.” Source: Intermolecular force. (2023, June 22). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Intermolecular_force

“Proteins that assist the conformational folding or unfolding of large proteins or macromolecular protein complexes.” Source: Chaperone (protein). (2023, April 2). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Chaperone_(protein)

Refers to local repeated backbone-backbone interactions of the polypeptide chain. There are two common forms of secondary structure: alpha helices and beta sheets.

“A common motif in the secondary structure of proteins and is a right hand helix conformation in which every backbone N−H group hydrogen bonds to the backbone C=O group of the amino acid located four residues earlier along the protein sequence.” Source: Alpha helix. (2023, July 19). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Alpha_helix

“A common motif of the regular protein secondary structure. Beta sheets consist of beta strands (β-strands) connected laterally by at least two or three backbone hydrogen bonds, forming a generally twisted, pleated sheet. A β-strand is a stretch of polypeptide chain typically 3 to 10 amino acids long with backbone in an extended conformation.” Source: Beta sheet. (2022, September 9). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Beta_sheet

All the structure of the protein that is not primary or secondary or quaternary. This defines the 3D shape and encompasses interactions between side chains, side chains–backbone, and backbone-backbone (that are not local or repeated).

Refers to the interactions between two independent polypeptide chains. Not all proteins have quaternary structure. Some will only have quaternary structure transiently.

The covalent linkage of thiol groups between two thiol residues. Commonly, these are found on the R groups of cysteine amino acids.

A type of protein that is embedded into biological membranes. Transmembrane proteins are an example of integral membrane proteins that span the entire biological membrane and are used to transport material from one side to the other.

Membrane proteins that do not embed in the biological membrane. Instead, they bind to other integral proteins or more rarely interact with the lipids themselves.

A protein that has a region that extends through the full membrane. As a result, the protein will have exposure to both sides of the membrane.

A protein that is embedded in a biological membrane but only interacts with one layer of the bilayer.

“A beta sheet composed of tandem repeats that twists and coils to form a closed toroidal structure in which the first strand is bonded to the last strand (hydrogen bond). Beta-strands in many beta barrels are arranged in an antiparallel fashion. Beta barrel structures are named for resemblance to the barrels used to contain liquids.” Source: Beta barrel. (2022, May 5). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Beta_barrel

A biological membrane that separates the cell from the surrounding environment. It is composed of a phospholipid bilayer, proteins, and carbohydrate components.

“Proteins that are heavily glycosylated. The basic proteoglycan unit consists of a ‘core protein’ with one or more covalently attached glycosaminoglycan (GAG) chain(s).” Source: Proteoglycan. (2023, February 2). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Proteoglycan

A carbohydrate polymer often found in the extracellular matrix and often serves a lubrication function.

“The main structural protein in the extracellular matrix found in the body’s various connective tissues.” Source: Collagen. (2023, July 3). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Collagen

A special protein structure where cellulose is made in plant cells destined for use in the cell wall.

Refers to experiments conducted in a computer; computer simulations.

“An interdisciplinary field of science that develops methods and software tools for understanding biological data, especially when the data sets are large and complex.” Source: Bioinformatics. (2023, July 7). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Bioinformatics

A graph used to predict the number of position of transmembrane domains in an amino acid sequence. Each amino acid’s hydropathy score is plotted with the N terminus of the protein starting at position 0 on the x axis. A peak on this plot above the threshold indicates a potential transmembrane domain.

An index (reported on the y axis) that is a measure of the average hydrophobicity of a given amino acid residue (hydropathy score) and the average hydropathy score of a range of amino acids on either side combined into one value.

A set value indicating the hydrophobicity of a given amino acid.

“A method for separation and analysis of biomacromolecules (DNA, RNA, proteins, etc.) and their fragments, based on their size and charge.” Source: Gel electrophoresis. (2023, July 17). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Gel_electrophoresis

A method to separate molecules (primarily proteins but sometimes DNA) by size. For protein separation, SDS (a detergent molecule) helps denature (unfold) the protein and gives it a consistent size-to-charge ratio. Then the proteins are loaded into a gel matrix and induced with an electric charge. The negatively charged proteins move toward the positive terminal. Larger molecules take longer to move through the gel and stay at the top, while smaller molecules can move more easily through the gel and reside at the bottom.

A method of gel electrophoresis used in biochemistry, molecular biology, genetics, and clinical chemistry to separate a mixed population of macromolecules such as DNA or proteins in a matrix of agarose, one of the two main components of agar. Source: Gel electrophoresis. (2023, July 17). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Gel_electrophoresis

“Native gels are run in non-denaturing conditions so that the analyte’s natural structure is maintained. This allows the physical size of the folded or assembled complex to affect the mobility, allowing for analysis of all four levels of the biomolecular structure.” Source: Gel electrophoresis. (2023, July 17). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Gel_electrophoresis

“A form of gel electrophoresis commonly used to analyze proteins. Mixtures of proteins are separated by two properties in two dimensions on 2D gels.” Source: Two-dimensional gel electrophoresis. (2023, March 2). In Wikipedia, The Free Encyclopedia . https://en.wikipedia.org/wiki/Two-dimensional_gel_electrophoresis

Fundamentals of Cell Biology Copyright © 2024 by Lauren Dalton and Robin Young is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License , except where otherwise noted.

• Previous Article
• Next Article

Cover Image

Structure and organization of membranes

Membrane proteins, internal membranes in eukaryotic cells form organelles, sending messages across membranes, membranes in health and disease, closing remarks, recommended reading and key publications, biological membranes.

• Split-Screen
• Article contents
• Figures & tables
• Supplementary Data
• Peer Review
• Open the PDF for in another window
• Cite Icon Cite
• Get Permissions

Helen Watson; Biological membranes. Essays Biochem 15 November 2015; 59 43–69. doi: https://doi.org/10.1042/bse0590043

Download citation file:

• Ris (Zotero)
• Reference Manager

Biological membranes allow life as we know it to exist. They form cells and enable separation between the inside and outside of an organism, controlling by means of their selective permeability which substances enter and leave. By allowing gradients of ions to be created across them, membranes also enable living organisms to generate energy. In addition, they control the flow of messages between cells by sending, receiving and processing information in the form of chemical and electrical signals. This essay summarizes the structure and function of membranes and the proteins within them, and describes their role in trafficking and transport, and their involvement in health and disease. Techniques for studying membranes are also discussed.

Membranes are composed of lipids, proteins and sugars

Biological membranes consist of a double sheet (known as a bilayer) of lipid molecules. This structure is generally referred to as the phospholipid bilayer. In addition to the various types of lipids that occur in biological membranes, membrane proteins and sugars are also key components of the structure. Membrane proteins play a vital role in biological membranes, as they help to maintain the structural integrity, organization and flow of material through membranes. Sugars are found on one side of the bilayer only, and are attached by covalent bonds to some lipids and proteins.

Three types of lipid are found in biological membranes, namely phospholipids, glycolipids and sterols. Phospholipids consist of two fatty acid chains linked to glycerol and a phosphate group. Phospholipids containing glycerol are referred to as glycerophospholipids. An example of a glycerophospholipid that is commonly found in biological membranes is phosphatidylcholine (PC) ( Figure 1 a), which has a choline molecule attached to the phosphate group. Serine and ethanolamine can replace the choline in this position, and these lipids are called phosphatidylserine (PS) and phosphatidylethanolamine (PE), respectively. Phospholipids can also be sphingophospholipids (based on sphingosine), such as sphingomyelin. Glycolipids can contain either glycerol or sphingosine, and always have a sugar such as glucose in place of the phosphate head found in phospholipids ( Figure 1 b). Sterols are absent from most bacterial membranes, but are an important component of animal (typically cholesterol) and plant (mainly stigmasterol) membranes. Cholesterol has a quite different structure to that of the phospholipids and glycolipids. It consists of a hydroxyl group (which is the hydrophilic ‘head’ region), a four-ring steroid structure and a short hydrocarbon side chain ( Figure 1 c).

Schematic representations of three types of membrane lipid.

( a ) Phosphatidylcholine, a glycerophospholipid. ( b ) Glycolipid. ( c ) A sterol.

The sugars attached to lipids and proteins can act as markers due to the structural diversity of sugar chains. For example, antigens composed of sugar chains on the surface of red blood cells determine an individual's blood group. These antigens are recognized by antibodies to cause an immune response, which is why matching blood groups must be used in blood transfusions. Other carbohydrate markers are present in disease (e.g. specific carbohydrates on the surface of cancer cells), and can be used by doctors and researchers to diagnose and treat various conditions.

Amphipathic lipids form bilayers

All membrane lipids are amphipathic—that is, they contain both a hydrophilic (water-loving) region and a hydrophobic (water-hating) region. Thus the most favourable environment for the hydrophilic head is an aqueous one, whereas the hydrophobic tail is more stable in a lipid environment. The amphipathic nature of membrane lipids means that they naturally form bilayers in which the hydrophilic heads point outward towards the aqueous environment and the hydrophobic tails point inward towards each other ( Figure 2 a). When placed in water, membrane lipids will spontaneously form liposomes, which are spheres formed of a bilayer with water inside and outside, resembling a tiny cell ( Figure 2 b). This is the most favourable configuration for these lipids, as it means that all of the hydrophilic heads are in contact with water and all of the hydrophobic tails are in a lipid environment.

A membrane bilayer and liposome.

Spontaneous formation of bilayers by membrane lipids. The hydrophilic heads (pink circles) will always face the aqueous environment in bilayers ( a ) and liposomes ( b ). The hydrophobic tails will face inward away from the water.

Early experiments by E. Gorter and F. Grendel in 1925 were the first to demonstrate that biological membranes are bilayers. These researchers extracted the lipids from red blood cells and found that they occupied a space that was twice the surface area of the cell. Red blood cells contain no internal membranes, so they deduced that the plasma membrane must be composed of two layers of lipids.

Biological membranes and the fluid mosaic model

The fluid mosaic model proposed by Jonathan Singer and Garth Nicolson in 1972 describes the dynamic and fluid nature of biological membranes. Lipids and proteins can diffuse laterally through the membrane. Phospholipids can diffuse relatively quickly in the leaflet of the bilayer in which they are located. A phospholipid can travel around the perimeter of a red blood cell in around 12 s, or move the length of a bacterial cell within 1 s. Phospholipids can also spin around on their head-to-tail axis, and their lipid tails are very flexible. These different types of movements create a dynamic, fluid membrane which surrounds cells and organelles. Membrane proteins can also move laterally in the bilayer, but their rates of movement vary and are generally slower than those of lipids. In some cases, membrane proteins are held in particular areas of the membrane in order to polarize the cell and enable different ends of the cell to have different functions. One example of this is the attachment of a glycosyl-phosphatidylinositol (GPI) anchor to proteins to target them to the apical membrane of epithelial cells and exclude them from the basolateral membrane.

Fluorescence photobleaching is one experimental method that is used by scientists to demonstrate visually the motility of proteins and lipids in a bilayer ( Figure 3 ). A lipid or membrane protein located on the surface of a cell is tagged with a fluorescent marker such as green fluorescent protein (GFP). A beam of laser light is then focused on to a small area of the cell surface using a fluorescence microscope in order to bleach the fluorescent tags in this area so that they no longer emit a fluorescence signal. This small area of membrane is observed over time and gradually the fluorescence increases again, indicating that other tagged proteins or lipids are diffusing into this region from elsewhere in the membrane. This demonstrates that the lipid bilayer surrounding cells is fluid in nature and allows lateral diffusion of both lipids and membrane proteins.

Photobleaching.

Cells expressing a GFP-labelled protein in the endoplasmic reticulum were subjected to photobleaching. ( a ) A cell before bleaching. ( b ) The same cell immediately after bleaching of the square section shown. ( c ) The same cell 5 min after photobleaching. Adapted from Figure 1b from Lippincott-Schwartz, J., Snapp, E. and Kenworthy, A. (2001) Studying protein dynamics in living cells. Nat. Rev. Mol. Cell. Biol. 2 , 444–456.

Despite all this movement of lipids and proteins in the bilayer, vertical movement, or ‘flip-flop’, of lipids and proteins from one leaflet to another occurs at an extremely low rate. This is due to the energetic barrier encountered when forcing the hydrophilic head (in the case of lipids) or hydrophilic regions (in the case of proteins) through the hydrophobic environment of the inside of the membrane. This near absence of vertical movement allows the inner and outer leaflets of the bilayer to maintain different lipid compositions, and enables membrane proteins to be inserted in the correct orientation for them to function. However, some enzymes facilitate the process of lipid flip-flop from one leaflet to another. These flippases, or phospholipid translocators, use ATP to move lipids across the bilayer to the other leaflet. In eukaryotic cells, flippases are located in various organelles, including the endoplasmic reticulum (ER), where they flip-flop newly synthesized lipids.

How membranes are made

Biological membranes are formed by adding to a pre-existing membrane. In prokaryotes this occurs on the inner leaflet of the plasma membrane, facing the cytoplasm. In eukaryotes, membrane synthesis takes place at the ER on the cytoplasmic leaflet of the ER membrane (termed the ‘inside’ of the cell). Lipids then leave the ER and travel through the secretory pathway for distribution to various subcellular compartments or the plasma membrane.

In eukaryotic cells, enzymes that span the ER catalyse the formation of membrane lipids. In the cytoplasmic leaflet of the ER membrane, two fatty acids are bound, one by one, to glycerol phosphate from the cytoplasm. This newly formed diacylglycerol phosphate is anchored in the ER membrane by its fatty acid chains. The phosphate is then replaced by the head group (e.g. phosphate and choline). Flippases in the ER membrane can then move some of these newly formed lipids to the luminal side of the ER membrane. Similarly, flippases in prokaryotes can transfer new lipids from the inner leaflet of the plasma membrane to the outer leaflet. These flippases are responsible for adjusting the lipid composition of each layer of the membrane. In eukaryotes, lipids must then be distributed to the various intracellular membranes. The traffic of vesicles between organelles in combination with signals that direct particular lipids to specific locations is required to create the correct lipid composition in all of the cellular membranes ( Figure 4 ). Vesicles bud from the ER and travel via the ER–Golgi intermediate compartment (ERGIC) to join with the Golgi, where sorting of lipids takes place. The Golgi then sends lipids in vesicles to various destinations, including the plasma membrane and lysosomes. Lipids and proteins are internalized from the plasma membrane into endosomes. Organelles, such as mitochondria, acquire lipids from the ER by a different mechanism. Water-soluble proteins called phospholipid-exchange proteins remove phospholipids from the ER membrane and deposit them in the membranes of the appropriate organelles.

Membrane traffic in eukaryotic cells.

The main compartments of eukaryotic cells are shown. Arrows indicate movement of lipid vesicles between them, with colours at the tail end indicating origin and those at the head end indicating destination.

Distribution of lipids

The inner and outer leaflets of bilayers differ in their lipid composition. In mammalian cells, the outer leaflet of the plasma membrane contains predominantly PC and sphingomyelin, whereas PS and PE are found on the inner leaflet. During programmed cell death (apoptosis), PS is no longer restricted to the inner leaflet of the plasma membrane. It is exposed on the outer leaflet by the action of an enzyme called scramblase which is a type of flippase enzyme. PS is negatively charged, unlike PC, which has no net charge. The movement of PS into the outer leaflet therefore changes the charge of the plasma membrane as viewed from the outside of the cell. This change in surface charge labels the apoptotic cell for phagocytosis by phagocytic cells such as macrophages.

Lipid composition also varies between the organelles within eukaryotic cells. Cholesterol is synthesized in the ER, but the ER membrane has a relatively low cholesterol content, as much of the cholesterol is transported to other cellular membranes. The prevalence of cholesterol in membranes increases through the secretory pathway, with more in the Golgi than in the ER (the trans -Golgi network is richer in cholesterol than the cis -Golgi), and most in the plasma membrane. This increase in cholesterol through the secretory pathway results in slightly thicker membranes in the late Golgi and plasma membrane compared with the ER, and is thought to be a contributing factor to protein sorting through the pathway, as membrane proteins in the plasma membrane generally have longer hydrophobic transmembrane domains than membrane proteins that reside in the ER.

Membrane proteins are the nanomachines that enable membranes to send and receive messages and to transport molecules into and out of cells and compartments. Without membrane proteins the phospholipid membrane would present an impenetrable barrier and cells would be unable to communicate with their neighbours, transport nutrients into the cell or waste products out of it, or respond to external stimuli. Both unicellular and multicellular organisms need membrane proteins in order to live. The membrane proteins that are present in a particular membrane determine the substances to which it will be permeable and what signal molecules it can recognize.

Synthesis of membrane proteins

In eukaryotic cells, the synthesis of membrane proteins destined for the plasma membrane, ER or any other membrane-bound compartment begins on cytosolic ribosomes. After a short segment of protein has been synthesized, the ribosome, mRNA and nascent protein chain associate with the ER, where the rest of the protein is made and simultaneously inserted into the membrane. This phenomenon was first explained by Günter Blobel, David Sabatini and Bernhard Dobberstein in the 1970s. These scientists proposed that there is a ‘binding factor’ which recognizes the emerging protein chain and can dock the ribosome at the ER membrane. We now know that there is an N-terminal signal sequence within membrane proteins. These signal sequences are not identical but share a common motif, namely a hydrophobic stretch of 20–30 amino acids, a basic region at the N-terminus and a polar domain at the C-terminus of the signal. These N-terminal signal sequences are recognized by the signal recognition particle (SRP), which has binding sites for the signal sequence, ribosome and the SRP receptor which is embedded in the ER membrane. Upon binding the SRP, the ribosome pauses protein synthesis. The SRP binds to the SRP receptor, adjacent to a translocon pore in the ER membrane. The translocon is a protein pore through which membrane protein chains can be threaded into the membrane. It has a laterally opening gate to allow newly synthesized proteins into the ER membrane. Once the ribosome is at the translocon, the SRP dissociates and protein synthesis resumes. This process is referred to as co-translational targeting, and the main events are summarized in Figure 5 .

Co-translational ER protein targeting.

The key steps of ER targeting are summarized. Each component is labelled and the ER membrane is represented by double blue lines. The signal sequence (shown in black) becomes the first transmembrane domain of the protein in this example.

Co-translational targeting is the dominant mechanism for protein delivery to the ER in higher eukaryotes, whereas yeast and prokaryotes favour post-translational targeting, whereby proteins are delivered to the ER after completion of synthesis. Post-translational targeting also occurs in higher eukaryotes, often when a membrane protein is so small that the signal sequence does not emerge until the whole protein has been synthesized. Post-translational targeting can be carried out both by SRP-dependent and by SRP-independent mechanisms.

Structure and function of membrane proteins

Membrane-spanning proteins are diverse in structure and function. They can be constructed of α-helices or from β-barrels. The β-barrel membrane proteins often function as pores, with hydrophobic amino acids facing out into the bilayer. In addition, there are other non-spanning proteins which associate with the bilayer, often using a hydrophobic anchor. Here we shall focus on the α-helical membrane proteins. These proteins have at least one α-helical hydrophobic stretch of amino acids, around 20 residues in length, which corresponds to around 30 Å (the thickness of an average phospholipid bilayer). If an α-helical membrane protein spans the membrane more than once, it will have more than one of these hydrophobic sections. For example, the Ca 2+ -ATPase of the ER and sarcoplasmic reticulum (SR) spans the membrane 10 times, so it has 10 hydrophobic stretches of around 20 amino acids each.

Membrane proteins control what enters and leaves the cell

A vital class of membrane proteins are those involved in active or passive transport of materials across the cell membrane or other subcellular membranes surrounding organelles. For a cell or an organism to survive, it is crucial that the right substances enter cells (e.g. nutrients) and the right substances are transported out of them (e.g. toxins).

Passive and active transport

Molecules can cross biological membranes in several different ways depending on their concentration on either side of the membrane, their size and their charge. Some molecules, including water, can simply diffuse through the membrane without assistance. However, large molecules or charged molecules cannot cross membranes by simple diffusion. Charged molecules such as ions can move through channels passively, down electrochemical gradients. This movement is described as ‘downhill’, as the ions or molecules travel from an area of high concentration to an area of low concentration. This requires channel proteins but no energy input. Passive transport can also be mediated by carrier proteins that carry specific molecules such as amino acids down concentration gradients, again without any requirement for energy. Active transport moves species against concentration gradients and requires energy, which is obtained from ATP, from light, or from the downhill movement of a second type of molecule or ion within the same transporter ( Figure 6 ).

Passive and active transport.

The different types of membrane proteins involved in passive and active transport are shown.

Passive transport

Passive transport is the movement of molecules across biological membranes down concentration gradients. This type of transport does not require energy. Channels form water-filled pores and thus create a hydrophilic path that enables ions to travel through the hydrophobic membrane. These channels allow downhill movement of ions, down an electrochemical gradient. Both the size and charge of the channel pore determine its selectivity. Different channels have pores of different diameters to allow the selection of ions on the basis of size. The amino acids that line the pore will be hydrophilic, and their charge will determine whether positive or negative ions travel through it. For example, Ca 2+ is positively charged, so the amino acids lining the pores of Ca 2+ channels are generally basic (i.e. they carry a negative charge).

Channels are not always open. They can be gated by ligands which bind to some part of the protein, either by a change in membrane potential (voltage gated) or by mechanical stress (mechanosensitive). The nicotinic acetylcholine receptor is an example of a ligand-gated ion channel which opens upon binding the neurotransmitter acetylcholine ( Figure 7 ). The nicotinic acetylcholine receptor is a pentameric membrane protein composed of five subunits arranged in a ring, with a pore through the centre. In the closed state, the pore is blocked by large hydrophobic amino acid side chains which rotate out of the way upon acetylcholine binding to make way for smaller hydrophilic side chains, allowing the passage of ions through the pore. Opening of the nicotinic acetylcholine receptor allows rapid movement of Na + ions into the cell and slower movement of K + ions out of the cell, in both cases down the electrochemical gradient of the ion. The difference in gradients between Na + and K + across the membrane means that more Na + enters the cell than K + leaves it. This creates a net movement of positive charges into the cell, resulting in a change in membrane potential. Acetylcholine released by motor neurons at the neuromuscular junction travels across the synapse and binds to nicotinic acetylcholine receptors in the plasma membrane of the muscle cells, causing membrane depolarization. This depolarization of the muscle cells triggers Ca 2+ release and muscle contraction.

The nicotinic acetylcholine receptor.

The pentameric structure of the receptor is shown, with the pore region (P) indicated. Transmembrane helices (M1–M4) are labelled in each subunit. The bilayer is shown in orange. Reproduced from Berridge, M.J. (2012) Cell Signalling Biology; doi:10.1042/csb0001003, with permission.

Carrier proteins are the other class of membrane proteins, apart from channels, which can facilitate passive transport of substances down concentration gradients. Carrier proteins transport molecules much more slowly than channels, as a number of conformational changes in the carrier are required for the transport of the solute across the membrane. A molecule such as a sugar binds to the carrier protein on one side of the membrane where it is present at a high concentration. Upon binding, the carrier changes conformation so that the sugar molecule then faces towards the opposite side of the membrane. The concentration of sugar on this side is lower, so dissociation occurs and the sugar is released. This is downhill movement and, although slower than movement through channels, it requires no energy.

The cystic fibrosis transmembrane conductance regulator (CFTR) is an ATP-dependent chloride ion (Cl − ) channel that has an important role in regulating the viscosity of mucus on the outside of epithelial cells. ATP is used to gate the channel, but the movement of Cl − occurs down its electrochemical gradient, so does not require energy. A heritable change in the CFTR gene which results in a single amino acid deletion in the protein causes cystic fibrosis. This is a serious illness in which thick mucus accumulates in the lungs, causing a significantly lower than average life expectancy in patients who have the disease. Unimpaired ion transport is vital for our survival and health, and conditions such as cystic fibrosis highlight the need for research into these types of proteins.

Active transport

The transport of molecules across a membrane against a concentration gradient requires energy, and is referred to as active transport. This energy can be obtained from ATP hydrolysis (primary active transport), from light (as, for example, in the case of the bacterial proton pump bacteriorhodopsin), or from an electrochemical gradient of an ion such as Na + or H + (secondary active transport).

Calcium ions signal many events, including muscle contraction, neurotransmitter release and cellular motility. However, high cytoplasmic concentrations of Ca 2+ are toxic to the cell. Therefore Ca 2+ must be tightly regulated and removed from the cytoplasm either into internal stores (the ER, and the SR in muscle cells) or into the extracellular space. This Ca 2+ removal is carried out by a family of Ca 2+ -ATPases, including the sarco/endoplasmic reticulum Ca 2+ -ATPase (SERCA), which hydrolyse ATP to move Ca 2+ against its electrochemical gradient into the ER and SR ( Figure 8 ). There are Ca 2+ -ATPases in the ER, Golgi and plasma membrane, and despite their sequence similarity, these proteins are differentially targeted to the appropriate membrane. These Ca 2+ pumps are primary active transporters. SERCA moves two Ca 2+ ions into the ER or SR for every ATP molecule that is hydrolysed. The pump undergoes a cycle of binding ATP and phosphorylation, and undergoes large conformational changes every time it transports a pair of Ca 2+ ions. SERCA is a P-type ATPase (so called because it is phosphorylated during ion transport). There are many P-type ATPases, and they are conserved in evolution across many species. The Na + /K + -ATPase is one of these P-type ATPases, and it works in a similar way to SERCA to pump Na + out of the cell and K + into the cell using energy derived from the hydrolysis of ATP. We have now obtained three-dimensional structures of SERCA in a number of conformational states, which allow scientists to visualize the transport process.

The sarco/endoplasmic reticulum Ca 2+ -ATPase (SERCA).

The crystal structure of SERCA in the ADP- and Ca 2+ -bound state is shown. D351 (in red) is the residue phosphorylated during the movement of Ca 2+ ions into the ER or SR. The three cytoplasmic domains, phosphorylation (P), nucleotide binding (N) and actuator (A) are labelled. ADP is shown in yellow and Ca 2+ ions in green. Protein Data Bank (PDB) code 2ZBD, rendered using PDB Protein Workshop.

Secondary active transport requires an ion electrochemical gradient to drive the uphill transport of another solute. The downhill movement of one species drives the uphill movement of the other. This can be symport (in which both types of molecule or ion travel across the membrane in the same direction) or antiport (in which the two species travel in opposite directions), as shown in Figure 9 .

Symport and antiport.

The two types of co-transport are shown, with examples.

In order to transport glucose into cells, the Na + –glucose symporter uses the electrochemical gradient of Na + across the plasma membrane. The concentration of Na + is much higher outside the cell, and the inside of the cell is negatively charged relative to the outside, so by allowing Na + to travel down its electrochemical gradient, these transporters can move glucose uphill, into the cell and against its concentration gradient. This is referred to as symport, as both Na + and glucose travel in the same direction—in this case into the cell. In order for this symport to be sustainable, the Na + gradient must be maintained. This is done by the Na + /K + -ATPase, which uses ATP to pump the Na + back into the extracellular space, thus maintaining a low intracellular Na + concentration.

Both Na + and Ca 2+ are present at much higher concentrations outside the cell than inside it. Like the Na + –glucose symporter, the Na + –Ca 2+ exchanger uses the electrochemical gradient of Na + across the plasma membrane to move a second species (Ca 2+ ) against its electrochemical gradient. However, in this case the transporter is an antiporter, as it uses the concentration gradient of one substance moving in (Na + ) to move another (Ca 2+ ) out of the cell. This antiporter has an exchange rate of three Na + ions in to two Ca 2+ ions out. It moves Ca 2+ out of the cell faster than the plasma membrane equivalents of SERCA, but has a lower affinity for Ca 2+ than these P-type ATPases. Again this transporter relies on the Na + /K + -ATPase to maintain the low intracellular Na + concentration.

Solving the structure of membrane proteins

In order to understand more fully the mechanisms of action of membrane proteins such as the transporters described here, we can determine their three-dimensional protein structures. As a result of huge advances in structural biology in the last 50 years, we now have access to many thousands of protein structures in online databases. This enables researchers to visualize the structure of their protein of interest, and thus gain insight into its mechanism.

X-ray crystallography

The structure of whale myoglobin was solved in 1958 using X-ray crystallography, earning John C. Kendrew and Max Perutz the Nobel Prize in Chemistry. This was the first protein structure to be solved using this technique, and since then thousands of proteins have been solved using this method. X-ray crystallography works by firing a beam of X-rays at a crystalline structure and measuring the diffraction of the X-rays after they have passed through the structure of interest. This generates an electron density map, showing where different atoms in the structure are located. For regular crystalline solids such as salts this is relatively straightforward, but for large irregular molecules such as proteins it can present many technical challenges. Before a protein is subjected to X-ray beams, it must first be purified and crystallized. In nature, proteins exist in the busy milieu of a cell, surrounded by thousands of other types of proteins, as well as lipids and other molecules. A common method of obtaining enough of the protein of interest involves expressing the relevant gene in a system such as bacteria. The gene is tagged with a small protein tag which can be used to isolate the protein of interest. Bacterial systems allow large amounts of protein to be produced cheaply and quickly. However, if the protein of interest is from a species that is only distantly related to that in which it is normally expressed (e.g. a human protein produced in Escherichia coli ( E. coli )), the lack of correct glycosylation enzymes and the differences in protein folding and assembly may prevent the production of a biologically active protein. In addition, the expression of membrane proteins that make pores or channels can kill the host organism.

A pure protein sample is then crystallized by allowing water to evaporate away, in exactly the same way as a solution of salt will form crystals naturally when left to dry. Optimum conditions for this must be determined, and crystallization conditions are not always straightforward, as they differ from one protein to another. For soluble proteins such as myoglobin this is easier than for insoluble membrane proteins. Membrane proteins have lipid-soluble domains that will not dissolve in an aqueous medium. This significantly decreases the ease with which membrane protein structures can be solved using X-ray diffraction. However, there are ways in which scientists can overcome this difficulty. Generally, membrane proteins are removed from the membrane in which they were made and placed in an environment of lipids and detergents for crystallization. Sometimes the lipids associated with the protein are apparent in the crystal structure.

The number of solved crystal structures of proteins is constantly growing as technology improves and expertise is shared among scientists to help to optimize conditions for crystal production. The Protein Data Bank (PDB) is an online archive of protein structures which can be freely accessed by scientists worldwide. At the time of writing, 88% of the structures in the PDB have been solved by X-ray crystallography, and there are currently just under 70 000 X-ray crystal structures in the database. The number of membrane protein structures in the PDB is increasing rapidly with the refinement of crystallization techniques ( Figure 10 ).

X-ray crystal structures of membrane proteins.

The increase in the number of solved crystal structures of membrane proteins is shown from 1985, when the first such structure was solved. Adapted from White, S.H. (2009) Biophysical dissection of membrane proteins. Nature 459 , 344–346.

Other structural techniques

Nuclear magnetic resonance (NMR) spectroscopy is another valuable technique for elucidating membrane protein structure. Molecules are placed in a magnetic field and the resonance properties of different atomic nuclei are measured, which gives an indication of where in a particular molecule different atoms are located. Generally, NMR is limited to smaller proteins, below around 35 kDa in size. It also offers the potential to visualize proteins in a more physiologically relevant environment (e.g. in lipid bilayers or micelles). Another advantage of NMR is that it does not require the protein to be locked in a crystal lattice—a structure which can distort the natural shape of the protein.

Electron microscopy can also be used to study membrane protein structure. By freezing membrane proteins in their natural lipid environments, it is possible to investigate their structure using high-resolution electron microscopy. This provides a snapshot of the naturally occurring conformation of individual proteins in the bilayer.

Interactions between lipids and proteins in biological membranes

The lipids that surround membrane proteins in biological membranes play an important role in the activity of these proteins. As was mentioned earlier, some membrane protein crystal structures include lipids bound to the outside surface of the transmembrane domains of the proteins. It is thought that these lipids bind tightly to the protein, and have a long-lived interaction with the transmembrane region. In other cases, lipids are thought to interact briefly with membrane proteins, rapidly moving away and being replaced by other membrane lipids. The activity of membrane proteins is considered to be dependent to some extent on the lipids that surround them in the membrane. Certain types of K + channel are thought to bind to negatively charged membrane lipids, as the activity of these channels increases at higher anionic lipid concentrations. These types of interaction can be studied by placing a purified form of the protein of interest in an artificial bilayer and measuring its activity. By altering the types of lipid present in the artificial bilayer, deductions can be made about the lipids that the protein requires in order to be active. Fluorescence spectroscopy and electron spin resonance are two techniques that are used to measure how strongly membrane proteins interact with specific lipids around them.

Molecular dynamics simulations use computer algorithms to work through theoretical problems. These simulated experiments are useful for investigating interactions between membrane proteins and lipids, as in real membranes these interactions are often so fleeting that they are very difficult to measure. Molecular dynamics simulations have predicted that in the case of the nicotinic acetylcholine receptor, the negatively charged lipid, phosphatidic acid, is required for activity. These simulations have also shown that cholesterol stabilizes the receptor and that the phosphatidic acid forms a shell around the protein which is more long-lasting than the interactions with other membrane lipids. Although molecular dynamics simulations are extremely useful, they are limited by the assumptions and approximations on which they are based. As in many areas of biology, a combination of experimental and computational research is required if real progress is to be made in understanding the complexity of biological membranes.

Inside the plasma membrane that surrounds eukaryotic cells lie many other membranes which define the intracellular compartments, or organelles. Each of these organelles has distinct functions and contains specific complements of proteins adapted for these roles. With the exception of a few proteins that are coded for by the mitochondrial genome, synthesis of all of the proteins that are required in these organelles begins on ribosomes in the cytoplasm, and therefore the proteins must be directed to the correct destination. We have seen earlier how this is achieved with membrane proteins, and most organelles have some kind of signal sequence that can be recognized by various receptors and which ensures that the protein arrives at the correct organelle.

Organelles have distinct lipid compositions

Besides the specific protein complement of each organelle, the lipid make-up of the bilayers surrounding organelles varies. Lipids are synthesized in the ER, and flippases move lipid molecules between leaflets of the bilayer. For organelles in the secretory pathway and the plasma membrane, lipid transport into these compartments is mediated by vesicular membrane traffic through the pathway. The cholesterol concentration in membranes increases from the ER through the Golgi to the plasma membrane. Cholesterol makes membranes thicker and more rigid, so the low levels of cholesterol in the ER membrane render it thin and facilitate the insertion of newly synthesized membrane and secretory proteins. PC becomes relatively less abundant through this pathway, with more found in the ER than at the plasma membrane. PS and PE are found throughout the secretory pathway in the cytosolic leaflet of the membranes. This differential lipid composition through the secretory pathway is achieved by targeting specific lipids into transport vesicles. Proteins included in these vesicles act as labels and direct the lipids to the right compartment. Forward-moving (anterograde) vesicles destined for the plasma membrane are rich in cholesterol. Lipids also move backwards through the secretory pathway, from the plasma membrane towards the ER. This is known as retrograde traffic. Retrograde vesicles from the Golgi are enriched in lipids such as PC, which are concentrated in the ER.

The lipid composition of the mitochondria is very different from that of the secretory pathway compartments. Mitochondrial membranes are much richer in PE and cardiolipin than is the ER. Cardiolipin is synthesized in the mitochondria and is predominantly confined to this organelle. As membrane proteins have evolved along with their organelles and surrounding lipids, it follows that different lipid compositions are required in different organelles for the optimum activity of the proteins within their membranes. The structure of the ADP/ATP carrier in mitochondria has been solved and was found to include cardiolipin and PC molecules bound to the protein. The activity of this carrier protein is dependent on the presence of cardiolipin, which is relatively abundant in mitochondrial membranes.

Proteins must be targeted to the correct organelle for cells to function

The targeting of newly synthesized membrane and secretory proteins to the ER has already been briefly discussed. However, there are many different destinations within the cell to which a protein can be sent, and sometimes proteins are located in more than one of these. The signals and protein machinery that are required to target proteins to the correct compartment are many and various, and much of the detail of the exact mechanisms involved has yet to be clarified.

Vesicular transport

Traffic through the secretory pathway is by vesicular transport in both anterograde and retrograde directions. Proteins and lipids can be included and excluded from vesicles by various means in order to selectively determine which molecules move forward or backward through the pathway. Vesicles are coated with proteins that determine their destination. Generally these coat proteins (COPs) are directional—COPII coats anterograde vesicles, and COPI coats retrograde vesicles. Proteins that travel in vesicles (referred to as cargo) are selected either by interacting with receptors in the vesicles or by directly interacting with the coat proteins. The selection of cargo occurs at the budding stage, when the coat proteins begin to distort the donor membrane (e.g. the ER) into a vesicle. Once the cargo has been selected and the coat proteins have been assembled, the vesicle buds off and travels to the acceptor membrane (e.g. the Golgi in the case of COPII vesicles from the ER) either by diffusion or with the help of motor proteins that ‘walk’ the vesicle along the cytoskeleton. The vesicle then fuses with the acceptor membrane, depositing its cargo and constituent lipids ( Figure 11 ).

Vesicle production, transport and fusion.

The main events in vesicular transport of cargo are shown. Cargo is selected and packed into vesicles which are formed by coat proteins (1 and 2). The GTPase Rab is also incorporated on the outside of the vesicle and facilitates the steps illustrated. The vesicle then travels along proteins of the cytoskeleton towards its destination following dissociation of coat proteins (3). The vesicle is tethered to the donor membrane with the help of tethering proteins and the SNARE complexes (4), allowing membrane fusion and release of the cargo (5).

The formation and fusing of vesicles are energetically demanding, as these processes require the stable bilayer to be broken in order to pinch off a new vesicle, and then fused with a different membrane. Both energy and specialized protein machinery are required to overcome this energy barrier.

Once the vesicle has budded, travelled through the cytosol and reached its destination, it must fuse with its receptor membrane. Again this is an energetically unfavourable process, and protein machinery has to be utilized in order to allow the fusion of two bilayers. SNARE proteins are central to the process of vesicle fusion. Vesicles carry v-SNAREs that bind to specific t-SNAREs on the target membranes. Not only does this confer specificity in the targeting of vesicles, but also the SNAREs facilitate membrane fusion on arrival of the vesicle. The interacting v-SNAREs and t-SNAREs form a four-helix bundle at the interface between the two membranes, consisting of three helices from the t-SNARE and one helix from the v-SNARE. This stable interaction is thought to provide the free energy necessary to enable the two membranes to become very close and fuse. As the two bilayers become closer, the lipids in the two outer leaflets can come into contact with one another, thereby increasing the hydrophobic nature of the site and enabling the membranes to join, and overcoming the energy barrier. The transmembrane domains of the SNAREs are also believed to be involved in membrane fusion, as when they are replaced by lipids experimentally, fusion does not occur. Upon fusion, the cargo enters the target compartment, and the lipids and membrane proteins that formed the vesicle diffuse into the target membrane.

Determining the mechanisms of membrane budding is important for understanding how viruses such as the human immunodeficiency virus (HIV) produce new viral particles. Unlike the budding in the secretory pathway described earlier, HIV particles bud away from the cytoplasm, into the extracellular space. This viral budding occurs in the same orientation as the budding that occurs within endosomes. The proteins which enable this budding are referred to as endosomal sorting complexes required for transport (ESCRTs). HIV ‘hijacks’ the ESCRT machinery to enable it to bud from the plasma membrane, out of the cytoplasm and into the extracellular space. Interactions between HIV proteins and ESCRT proteins recruit the host cell ESCRT machinery to the budding vesicle, allowing membrane scission and vesicle release. Other viruses can bud without assistance from the ESCRTs, and it is thought that HIV may also be able to bud in an ESCRT-independent manner. Understanding more about these membrane budding and scission events is crucial to elucidating how viruses proliferate and how we can inhibit processes by means of drug interventions.

Protein trafficking in the secretory pathway

As described earlier, a hydrophobic stretch of 20–30 amino acids with a basic N-terminus and a polar region at the C-terminus emerging from the ribosome causes the protein to be targeted to the ER, where synthesis is completed. This hydrophobic stretch can be cleaved in the case of soluble proteins, or it can remain attached. An uncleaved signal sequence is referred to as a signal anchor sequence, as it both signals ER targeting and then goes on to anchor a protein in the membrane, and becomes a transmembrane domain in the fully folded protein. The SRP-dependent targeting step is common to ER proteins as well as proteins destined for the Golgi or the plasma membrane, or to be secreted from the cell.

ER exit is thought to allow some selection of which proteins remain in the ER and which proteins leave and move in vesicles towards the Golgi. ER exit sites are located in areas of the ER close to the Golgi, and are rich in COPII coat proteins. It is not understood exactly which properties of a protein determine whether it will leave the ER in COPII vesicles, but it is currently thought that the transmembrane domain length is an important factor. Longer transmembrane domains appear to predispose proteins to exit the ER and travel towards the Golgi. This is consistent with the fact that membrane thickness increases through the secretory pathway due to increased cholesterol content, as described earlier.

Upon arrival at the cis -Golgi, proteins can then be retrieved to the ER, remain in the Golgi, or travel onward to the plasma membrane. Retrieval to the ER is not fully understood, but some proteins contain retrieval motifs, such as the KDEL four-amino-acid motif which is recognized by a receptor and enables packaging of the protein into retrograde COPI vesicles. Other proteins appear to cycle between the ER and the Golgi without known retrieval motifs. Proteins move in both anterograde and retrograde directions through the Golgi stack. They can then leave the trans -Golgi and move to the plasma membrane in vesicles.

Proteins at the plasma membrane can move into the cell in vesicles by endocytosis (e.g. when surface receptors are internalized for degradation in lysosomes). Endocytic vesicles are often clathrin coated. Clathrin, like the COPs, distorts the membrane into curved structures, allowing vesicle formation. Clathrin forms a cage-like shape that promotes vesicle formation and scission by virtue of the rigid shape of the protein complexes which form at the membrane. Not all endocytosis is clathrin dependent, and there are other proteins, such as caveolin, which can facilitate the formation of endocytic vesicles.

Mitochondrial and nuclear protein targeting

Newly synthesized proteins destined for the mitochondria or the nucleus are targeted in a different way, independently of the secretory pathway. Some mitochondrial proteins are encoded by the mitochondrial genome, while others are encoded by the nuclear genome. Mitochondria have a double-layered membrane. Therefore targeting signals for mitochondrial proteins need to contain information not only to direct the protein to the organelle, but also to determine in which membrane it will be located (in the case of membrane proteins), or whether it will be located inside the mitochondria (the matrix) or in the intermembrane space between the inner and outer membranes (in the case of soluble proteins). Mitochondrial targeting motifs vary enormously, but generally are located at the N-terminus of the protein and are rich in positively charged and hydrophobic amino acids. Proteins destined for the nucleus are targeted by nuclear localization sequences that direct proteins which have been synthesized in the cytoplasm through nuclear pore complexes. Again these sequences are not very highly conserved, but generally contain clusters of positively charged amino acids.

We have already seen how ion channels and other transport proteins can allow substances to cross the lipid bilayer. Knowledge of how fat-soluble and water-soluble substances cross membranes is important for gaining an understanding of how messages cross membranes and thus how one cell can communicate with another. Cells receive and send messages constantly (e.g. in order to respond to hormone signals, conduct action potentials, and sense external stimuli such as taste and smell).

Messengers: lipid soluble or water soluble?

Substances that send messages are known as messengers, and they vary enormously in their chemical composition, size and hydrophobicity. In order to understand how a cell receives a message, it is important to ascertain first whether the messenger is lipid or water soluble. Hormones are one example of messengers that are released by cells. The human body contains both lipid-soluble and water-soluble hormones. Lipid-soluble hormones are generally transported through the blood, bound to carrier proteins. Steroid hormones such as the androgens and oestrogens are lipid soluble by virtue of their ringed molecular structures, which are derived from cholesterol. This allows these hormones to diffuse freely through the plasma membrane of cells and bind to their receptors, which are located inside cells. In the case of oestrogen, the receptor is located in the cytoplasm and upon ligand binding relocates to the nucleus, where it binds DNA and acts as a transcription factor, altering gene expression. The receptor contains a nuclear localization sequence which is hidden until oestrogen binds, allowing it to be targeted to the nucleus.

Other hormones, such as insulin and adrenaline, are water soluble and therefore cannot pass freely through the membranes of cells. Their receptors are located on the outside of the plasma membrane in order for them to be able to convey a message without entering cells. Insulin binds to the membrane-spanning insulin receptor on the surface of target cells, and initiates a signal cascade that results in an increase in the number of glucose transporters at the cell membrane, and a subsequent increase in glucose uptake.

G proteins and second messengers

Many cell-surface receptors share structural features, including seven membrane-spanning helices. These 7TM receptors bind their ligand (the messenger molecule) on the extracellular side of the membrane, and bind a GTP-binding protein (G protein) on the intracellular side. Due to this interaction with G proteins, these receptors are called G-protein-coupled receptors (GPCRs). When the ligand binds the GPCR, the receptor undergoes conformational changes that are transferred through the membrane-spanning region to the bound G protein. This change in structure allows the G protein to exchange a bound GDP molecule for a GTP molecule, and thereby switch from an inactive state to an active state. G proteins consist of three subunits—α, β and γ. An inactive, GDP-bound G protein consists of all three subunits, with the nucleotide bound in the α subunit. When the GPCR binds the ligand, the G protein is activated and the α subunit, now with GTP bound to it, dissociates from the complex ( Figure 12 ). This activated α subunit now has an exposed face (where the β and γ subunits were bound) and can bind proteins to propagate the signal. An example of this downstream signalling from GPCRs is the activation of adenylate cyclase by the GTP-bound α subunit in the case of the β-adrenergic receptor when it binds its ligand, adrenaline (epinephrine). The effect of this adenylate cyclase activation is an increase in cAMP production from ATP, leading to downstream effects. The dissociated βγ dimer also has downstream effects. The α subunit has GTPase activity so that it can convert the bound GTP back to GDP. The GDP-bound subunit then returns to and binds the β and γ subunits ready for another cycle of signalling.

GPCR and heterotrimeric G-protein signalling.

The ligand bound to the GPCR is shown in red. Binding allows the exchange of GDP for GTP by the associated G protein, and dissociation of the protein into Gα and Gβγ subunits. These then have downstream effects on a range of proteins, thereby propagating the signal from the bound ligand. Yellow arrows indicate either activation (up arrow) or inhibition (down arrow) of the targets. Regulators of G-protein signalling (RGS) proteins aid the GTPase activity of the G protein to turn off the signal. Arrestin can bind the receptor following GPCR phosphorylation by G-protein receptor kinase (GRK), desensitizing the receptor to further signalling. Reproduced from Berridge, M.J. (2012) Cell Signalling Biology; doi:10.1042/csb0001002, with permission.

After the initial ligand interaction with the GPCR and the G-protein dissociation, the message is then carried by second messengers activated by the signal cascade. In the example that has just been given, the G protein associated with the β-adrenergic receptor activates adenylate cyclase, increasing the production of cAMP, which is a widely used second messenger. Most of the effects of cAMP are due to the activation of protein kinase A (PKA). PKA phosphorylates target enzymes to modify their activities. In the case of adrenaline, PKA activates enzymes involved in the production of glucose from glycogen stores, and inhibits enzymes involved in the production of more glycogen.

Around 25% of drugs are targeted at GPCRs, so an understanding of their structures and functions is crucial in the fight against disease. As explained earlier, membrane proteins are notoriously difficult to crystallize due to their hydrophobic nature, and GPCRs have a very small hydrophilic area. Some techniques, such as the production of an antibody–receptor complex to increase hydrophilicity, have been successful in aiding crystallization. Rhodopsin ( Figure 13 ) was crystallized in 2000, followed by the related β 2 -adrenergic receptor in 2007. Since then, several more GPCR structures have been solved, providing valuable information that can help computational biologists to work out the detailed mechanisms of GPCR signalling. Molecular dynamics simulations have been performed on the interactions between GPCRs and their partner G proteins using the crystal structures available to inform the modelling process. These studies will play an important part in helping us to understand how the helices in the GPCRs move and twist in order to convey the extracellular signal to the intracellular G protein.

The crystal structure of rhodopsin.

A ribbon representation of the first crystal structure of rhodopsin is shown in the plane of the membrane ( a ) and from the cytoplasmic side (b). The N- and C-termini are labelled, as are the seven transmembrane helices (I-VII). Adapted from Figure 2 from Palczewski, K., Kumasaka, T., Hori, T., Behnke, C.A., Motoshima, H., Fox, B.A., Le Trong, I., Teller, D.C., Okada, T., Stenkamp, R.E. et al. (2000) Crystal structure of rhodopsin: a G-protein-coupled receptor. Science 289 , 739–745.

Nerve impulses

Nerve impulses are able to occur because biological membranes are impermeable to ions, so a membrane potential can be generated across them, with more of one charge on one side than on the other. These membrane potentials are generated and altered by ion channels. A nerve impulse, or action potential, is generated when a membrane is depolarized upon influx of positively charged ions into the cell. The resting potential in a neuron is around –70 mV, maintained by K + channels and the Na + /K + -ATPase. When an action potential is generated, voltage-dependent Na + channels open once the cell membrane has crossed the threshold potential of around –60 mV. This allows a fast influx of Na + down its electrochemical gradient, increasing the membrane potential (i.e. reducing its negative value). This influx of positive charges enables the inside of the cell to become positively charged compared with the extracellular environment, as the membrane potential exceeds 0 mV. The depolarization itself inhibits the Na + channels, so no more ions enter the cell. To restore the negative resting potential, voltage-dependent K + channels open, allowing K + ions to move out of the cell, thus making the inside of the cell more negative. An after-potential (hyperpolarization) can then occur, whereby the membrane potential decreases below –70 mV before being restored by the action of ion channels and ATPases.

We have seen how membranes, and the membrane proteins within them, function in healthy cells and organisms. We shall now consider what happens in disease, and how we can use our knowledge of membrane proteins to make new drugs to treat disease.

Serious disease results from non-functional ion channels

Cystic fibrosis is an autosomal recessive disease that results from mutations in the CFTR gene. This gene encodes a Cl − channel that has a vital role in regulating the viscosity of mucus on membranes such as those in the lungs. In healthy individuals, transport of Cl − ions out of the cells through CFTR is followed by water, and mucus of the right viscosity is produced. However, lack of Cl − channels results in thick, dehydrated mucus, and consequently cystic fibrosis patients have difficulty in breathing and a predisposition to chest infections. Most cases of cystic fibrosis are caused by the ΔF508 mutation, which is a deletion of a phenylalanine residue at position 508 in the protein. Like nearly all membrane proteins, CFTR is translated on ribosomes at the ER and then moves through the secretory pathway to the plasma membrane, where it carries out its transport role. The single amino acid deletion of F508 causes the protein to misfold, and instead of moving out to the plasma membrane, it is held in the ER by the protein quality control machinery. Therefore very few CFTR molecules reach the plasma membrane in people with the ΔF508 mutation, and this results in serious disease.

Diseases such as this are not easily treated. Blocking the protein quality control machinery is not an option, because it would lead to the release of other misfolded proteins, with potentially disastrous consequences. Although some current drug treatments can ameliorate the symptoms of the disease, it is hoped that gene therapy might become routine as it addresses the cause of the problem. Treatment of patients with an artificial, functional version of the gene enables them to produce a working CFTR protein that can be expressed at the plasma membrane. Although this is not a complete cure, it is a potentially effective way to greatly reduce the symptoms of cystic fibrosis in the lungs. As DNA is a large hydrophilic molecule, it cannot be simply administered like many other drugs. Delivery of gene therapy is a challenge, and this is one reason why it is difficult to treat patients in this way, but methods of delivering new genetic material into cells have been developed. Viruses can be used to deliver the CFTR gene to cells, by harnessing their ability to inject cells with foreign DNA or RNA. Patients may also be able to be given liposomes containing the functional gene, which fuse with cell membranes and deliver the therapeutic gene. Gene therapy is a growing and important area of research, and it is hoped that many diseases, including some cancers, will eventually be able to be treated using DNA.

Membrane proteins provide an entry point for viruses

Viruses that attack the human body can use the body's own membrane proteins to recognize their target cells. HIV attacks cells of the immune system. A protein on the surface of HIV called gp120 binds to CD4 protein molecules on the surface of T-cells that are involved in immunoregulation, and allows fusion of the virus with the host cell. Once the contents of the virus have entered the CD4-positive cell, the HIV genome is integrated with the host genome and uses the host machinery to make new copies of the virus. Over time, the numbers of CD4 T-cells are reduced by the virus, and the patient's immune system eventually becomes so compromised that they are unable to fight invading pathogens. Many therapeutic agents have been created to help to fight HIV, and the interaction between CD4 and gp120 is just one of the points at which drugs can be used to stop the progression of the virus.

Toxins use endocytosis to gain entry to cells and block neurotransmission

Various toxins interfere with the transmission of messages across biological membranes. Tetanus neurotoxin (TeNT) and botulinum neurotoxin (BoNT) are both protein toxins that affect nerve impulse transmission between nerves and muscles. TeNT is produced by a soil bacterium and causes the skeletal muscle spasms that characterize tetanus infection. TeNT-producing bacteria generally enter the body through wounds, and TeNT binds glycolipids enriched at presynaptic membranes of motor neurons ( Figure 14 ). TeNT then undergoes endocytosis and moves up the axon to the dendrites that connect the motor neuron to an inhibitory interneuron. TeNT is released into the synapse between these two cells and is endocytosed into the inhibitory interneuron. Acidification of vesicles containing TeNT causes the protein toxin to break apart into two domains. One of these, the L domain, is translocated into the cytoplasm of the interneuron, where it uses its proteolytic activity to cleave vesicle-associated membrane protein (VAMP). Under normal circumstances, VAMP is part of the protein complex that allows synaptic vesicles to fuse with the presynaptic membrane and release inhibitory neurotransmitters. The action of the L-domain protease of TeNT means that VAMP can no longer function, inhibiting neurotransmitter release across the synapse. As this occurs in inhibitory interneurons, the resulting effect is prolonged skeletal muscle contraction, as no inhibition is conveyed to the motor neuron to allow relaxation.

Mechanism of action of tetanus toxin.

Tetanus toxin (blue circles) enters the presynaptic membranes of motor neurons by endocytosis, and moves up the axon to the dendrites that connect the motor neuron to an inhibitory interneuron. Microtubules (blue and green lines) and actin filaments (red lines) allow retrograde transport of the toxin. TeNT acts on the inhibitory interneuron, where it prevents the release of glycine (red dots), shown by a red cross. Small green dots represent neurotransmitter both inside and being released from vesicles. Adapted from Figure 2 from Rossetto, O., Scorzeto, M., Megighian, A. and Montecucco, C. (2013) Tetanus neurotoxin. Toxicon 66 , 59–63.

Botulinum neurotoxin acts in a similar way to TeNT, but has the opposite effect. Like TeNT, it is released by bacteria. BoNT binds to and is internalized by the presynaptic membrane of motor neurons at the neuromuscular junction. It is released from endocytic vesicles into the cytoplasm of the motor neuron, where it acts on the SNARE complex to inhibit the fusion of synaptic vesicles and release of excitatory neurotransmitters at the neuromuscular junction. This has the effect of blocking muscle contraction and causing paralysis in people infected with botulism. Despite its sometimes lethally toxic nature, BoNT is increasingly used by people who wish to look younger. The toxin is injected into the muscles of the face to cause paralysis, thereby reducing wrinkles and lines in the skin. When used in this way it is referred to as Botox.

Membrane proteins are the target of many drugs

Membrane proteins are important drug targets. As our structural and functional knowledge of membrane proteins expands, it is becoming possible to design more effective medicines. Computational tools are becoming an increasingly important part of the process. One important class of drug targets are pore-forming membrane proteins encoded by viruses. HIV, influenza and polio, among other viruses, encode membrane proteins that form pores in the host cell membranes in order to cause leakage and promote infection. One of these pore-forming proteins was formerly used as a drug target in the treatment of influenza. NMR studies have provided structural information about the pore-forming Vpu protein from HIV-1. Using these data together with structural information about pores with similar sequences, computational models of the structure of the channel in the host membrane can be produced. These models, combined with advanced biophysical techniques, are invaluable for predicting sites for potential drug molecule binding which can then be tested both computationally and experimentally.

More drugs target GPCRs than any other single group of proteins. As explained earlier, conformational changes in the GPCRs permit signals to cross the membrane. However, these large changes in conformation give the proteins flexibility, which makes it difficult for researchers to pinpoint the structures of the GPCRs in any one conformation. Solving the structures of different conformations using X-ray crystallography is a challenging task. As GPCRs are such important drug targets, much research has been focused on solving their structures in order to inform the discovery of new drugs. Computational methods have again proved crucial to understanding the detailed molecular structure of these proteins. Molecular dynamics simulations have been used to aid our understanding of the molecular changes that occur during GPCR activation, and also which lipids are required for the GPCR to function. Although molecular dynamics simulations are a key technique in this area of research, they have some limitations. The biological membranes that surround cells are extremely complex and contain different types of lipids and proteins, both within and associated with the membrane. At present, the time and financial resources required to provide the computational power to simulate such a complex environment are often prohibitive. Simpler models are therefore produced which, although they are able to predict conformational changes in receptors such as GPCRs, may omit other interactions that are important in the activity of membrane proteins. As is often the case, a combination of techniques will be required to gain greater insight into the conformational flexibility of the GPCRs. It is apparent from research conducted to date that these receptors can adopt many different conformations, rather than having just a simple ‘on’ and ‘off’ mechanism. Using different drug molecules to stabilize different conformations in different signalling pathways may be the best approach to finding more effective medicines in the future. There is now much pressure on researchers to replace, refine and reduce the use of animals in drug discovery (an approach referred to as the ‘three Rs’). By using computers in the early stages of the process to model drug–target interactions, researchers can produce much more promising compounds to test in experiments and drug trials.

Biological membranes allow life to exist. From simple unicellular prokaryotes to complex multicellular eukaryotes such as humans, the properties of the membranes that surround cells are remarkably similar. Our understanding of the structure of these lipid bilayers is now expanding rapidly as a result of significant advances in biophysical techniques and the huge computational power now available to researchers. The proteins that inhabit these membranes allow messages to be sent and received so that the cell can communicate with the external environment. Many messages are relayed by hydrophilic molecules that require receptors to transmit information across the bilayer. It is this step that is targeted by the majority of drugs which are on the market today, as it enables us to modify the message before it enters the cell. An understanding of how membrane proteins work, how they reach the correct destinations and how we can alter their functions is key to the fight against human disease.

Abbreviations

botulinum neurotoxin

cystic fibrosis transmembrane conductance regulator

coat protein

endoplasmic reticulum

ER–Golgi intermediate compartment

endosomal sorting complex required for transport

green fluorescent protein

G-protein-coupled receptor

glycosyl-phosphatidylinositol

GTP-binding protein

nuclear magnetic resonance

phosphatidylcholine

Protein Data Bank

phosphatidylethanolamine

protein kinase A

phosphatidylserine

sarco/endoplasmic reticulum Ca 2+ -ATPase

sarcoplasmic reticulum

signal recognition particle

tetanus neurotoxin

vesicle-associated membrane protein

This article is a reviewed, revised and updated version of the following ‘Biochemistry Across the School Curriculum’ (BASC) booklet: Brown B.S., 1996: Biological Membranes; ISBN: 0904498328. For further information and to provide feedback on this or any other Biochemical Society education resource, please contact [email protected] . For further information on other Biochemical Society publications, please visit www.biochemistry.org/publications .

Get Email Alerts

• Online ISSN 1744-1358
• Print ISSN 0071-1365
• Submit Your Work
• Language-editing services
• Recommend to Your Librarian
• Request a free trial
• Accessibility
• Sign up for alerts
• Sign up to our mailing list
• The Biochemist Blog
• Biochemical Society Membership
• Publishing Life Cycle
• Biochemical Society Events
• About Portland Press
• Portland Press Tel
• +44 (0)20 3880 2795
• Portland Press Company no. 02453983
• Biochemical Society Tel
• +44 (0)20 3880 2793
• Email: [email protected]
• Biochemical Society Company no. 00892796
• Registered Charity no. 253894
• VAT no. GB 523 2392 69
• Privacy and cookies
• © Copyright 2024 Portland Press

This Feature Is Available To Subscribers Only

Sign In or Create an Account

Essay on Biomembrane Structure

• Topical Review
• Open access
• Published: 15 March 2019
• Volume 252 , pages 115–130, ( 2019 )

Cite this article

You have full access to this open access article

• Christoph Gerle   ORCID: orcid.org/0000-0002-7265-2804 1 , 2  

10k Accesses

9 Citations

6 Altmetric

Explore all metrics

Of all the macromolecular assemblies of life, the least understood is the biomembrane. This is especially true in regard to its atomic structure. Ideas on biomembranes, developed in the last 200 years, culminated in the fluid mosaic model of the membrane. In this essay, I provide a historical outline of how we arrived at our current understanding of biomembranes and the models we use to describe them. A selection of direct experimental findings on the nano-scale structure of biomembranes is taken up to discuss their physical nature, and special emphasis is put on the surprising insights that arise from atomic scale descriptions.

Similar content being viewed by others

What Is So Unique About Biomembrane Organization and Dynamics?

A Beginner’s Short Guide to Membrane Biophysics

Computational Studies of Biomembrane Systems: Theoretical Considerations, Simulation Models, and Applications

Avoid common mistakes on your manuscript.

Introduction

The three key macromolecules of life are the oligomers of nucleic acids, that is DNA and RNA, oligomers of amino acids, that is proteins, and a multitude of lipids in the aggregate form of cell membranes. DNA and proteins are not only linked via the mRNA but also by the fact that both are linear polymers that assemble into 3D structures which consist of repeating units of only four kinds of nucleic acids or only 20 different kinds of amino acids. Biomembranes are set apart. These are no linear polymers; rather, they are composed of a wide diversity of many single amphiphilic lipid molecules forming a volume enclosing 3D structure whose architecture is not encoded in genetic information. Owing to this, the fact that biomembranes are as fundamental to life as DNA/RNA and proteins might be overlooked. And for a long time, it was doubted whether general statements describing common properties of the biomembrane could be made as such. Despite their very different structures and compositions, the gross structure of all three macromolecules of life is determined by their amphiphilic nature and their chemical environment of liquid water. Together these minimize the decrease in entropy that is caused by the ordering of water through the exposure of their hydrophobic groups to bulk water. In other words, the presence of liquid bulk water surrounding the macromolecules of life gives the entropic term the decisive weight in the energetics that govern their overall structure—an important link for understanding them.

For understanding the properties of the macromolecules of life, atomic scale structures are key. Indeed, before the arrival of experimentally determined structures of DNA and proteins, ideas on how they might appear at the atomic scale were strongly influenced by X-ray structures of inorganic molecules such as crystalline salt [ 1 ]. Since the first experimentally derived atomic scale models of DNA and protein were reported 60 years ago [ 2 – 5 ], many thousands of atomic coordinates have been deposited in the protein data bank archives [ 6 ] (wwPDB; http://www.pdb.org ), which currently contain more than 149,000 entries. However, in stark contrast to this impressive body of knowledge, relatively little is known about the structure of biomembranes: not only are there only several hundred known unique membrane protein structures (Membrane Proteins of Known 3D Structure; http://blanco.biomol.uci.edu/mpstruc/ ) [ 7 ], but also is the rate of their discovery much slower than anticipated (see growth curve in [ 7 ]). Importantly, very few of the deposited structures include a full embedding lipid bilayer and thus far not a single membrane protein structure has been determined under physiological membrane potential. As an example of how difficult it is to know and understand even a seemingly simple geometric property of biomembranes, the question of how thick they are will act as a leitmotif for this essay.

Historical Background

Not long after Matthias Schleiden and Theodor Schwann had proposed in 1838 that the basic unit of all life on our planet is the cell [ 8 , 9 ], Moritz Traube put forward the idea that cells must be surrounded by a semi-permeable barrier. Then, at the turn of the last century, studies on the effectiveness of anesthetics by Meyer and Overton suggested that this barrier might be formed by lipidic substances such as lecithin and cholesterol [ 10 , 11 ]. Work on the structurally simple red blood cells which lack internal membranes proved to be productive, and in 1925 surface area measurements of extracted phospholipids spread on a Langmuir trough gave the correct indication that the cell membrane is formed by a bilayer of lipids [ 12 ]. In the same year, Hugo Fricke measured the electric capacitance of red blood cells for estimating the thickness of their cell membrane [ 13 ]. Assuming a low dielectric constant ε for lipids of ~ 3, his calculations gave a value of 3.3 × 10 −7 cm, i.e. 33 Å, for the thickness of the cell membrane—a value much thinner than anybody expected. Hugo Fricke did not take into account the fact that lipids have a hydrophilic head group of considerable size, and therefore, his experimental set-up did not measure the total thickness of the red blood cell membrane, but only that of its hydrophobic core. As a consequence, he compared the value of 33 Å with the length of lecithin measured in Langmuir trough monolayers, where the lipids take a non-physiological stretched conformation and concluded that red blood cell membranes consist only of a monolayer of lipids. Perhaps, it was this interpretational blunder that put Hugo Fricke’s experimental value of 33 Å on the sideline for many decades to come.

Davson and Danielli, who were the proponents of the very influential—albeit wrong in many aspects, as would eventually be shown—trilamellar membrane model, were aware of Hugo Fricke’s precise value of 33 Å [ 14 ], which clearly did not match their own estimation of around 120 Å for the thickness of biomembranes. Davson and Danielli’s trilamellar model would dominate textbooks for the next 40 years and thus influence basic thinking of life scientists around the world: according to this, the membrane consists of a bilayer of phospholipids sandwiching a disordered central layer of oily molecules such as cholesterol or fatty acids (Fig.  1 a) [ 14 ]. Importantly, this lipid/oil sandwich was conceived to be unstable in itself and thus mainly held together by the strong, localized electrostatic interaction of a layer of water-soluble proteins on both sides. Thinking on biomolecules at the time of Davson and Danielli was profoundly influenced by the success of the Braggs in using X-ray diffraction to build models at the atomic scale for inorganic matter such as crystals of table salt [ 1 ]. These atomistic insights, together with the precise description of (biologically important) water–salt solutions by Debey-Hückel [ 15 ], gave the impression that local, directed strong electrostatic interactions between opposite charges such as salt-bridges, should govern the gross structure of the macromolecules of life. The trilamellar model was further endorsed by early electron micrographs of cell membranes obtained in the 1950s by the Romanian cell biologist Palade and the Danish biologist Sjörstrand [ 16 , 17 ]; their micrographs of OsO 4 fixated and negatively stained cells visualized features matching the predicted trilamellar structure. Importantly, the trilamellar structures found in the images also matched the expected thickness of ~ 120 Å for the cell membrane.

Models of the biomembrane. a The trilamellar membrane model by Davson and Danielli from 1935: a fluid core of lipidic substances is sandwiched by a bilayer of phospholipids whose hydrophilic headgroups (blue) are in electrostatic contact with a cask of water soluble globular proteins; (the membrane proteins) (redrawn after [ 14 ]). b Singer and Nicolson’s fluid mosaic model of the membrane from 1972 with a self-organized lipid bilayer acting as a passive matrix for transmembrane proteins freely floating in two dimensions (redrawn from [ 26 ]). See Table  1 for a list of model properties. c The first direct structural insight into membrane protein structure obtained by electron crystallography of the archaebacterial purple membrane in 1974 by Unwin and Henderson. Seven transmembrane alpha-helices demonstrate the concept of transmembrane proteins to be correct (redrawn from [ 31 ]). (Color figure online)

A further development of the dominant Davson and Danielli trilamellar membrane model was the concept, put forward by Robertson [ 18 ], that all biomembranes should have the same basic architecture. Considering the wide variety of lipids found in the many types of cells harbouring a plethora of different architectures of cell membranes and cell organelle membranes, this concept of the unit membrane, as it was called, is neither self-evident nor was it widely accepted [ 19 ]. Robertson also modified the water-soluble proteins that were thought to enforce the stability of the membrane from globular shape to a sheath of β-sheets forming a stabilizing cask that matched the smooth appearance of the membrane observed in the micrographs of Sjörstrand and Palade. Meanwhile in the quest for the structure of DNA, the trilamellar membrane model found its counterpart in Linus Pauling’s triple helix model of DNA structure. The triple helix model had its hydrophobic purines and pyrimidines pointing out into bulk solution and charged phosphate-sugars building the triple helix’ backbone were held together by the electrostatic interaction afforded through intercalating positive Mg 2+ ions [ 20 ]—a model clearly influenced by ideas of strong and directed molecular interaction. Linus Pauling correctly predicted the structure of α-helices and β-sheets as important architectural elements of protein structure governed by the geometrically precise interactions of individual groups of atoms via hydrogen bonds [ 21 , 22 ]. However, although local and directed chemical bonds determine the structure of organic molecules, non-directional entropic phenomena govern the gross structure of the macromolecules of life. Shortly after Linus Pauling’s proposal, Watson and Crick, who knew of the triple helix model and the X-ray diffraction patterns taken by Rosalind Franklin and Raymond Gosling from calf thymus DNA [ 5 ], built the double helix model of DNA structure that would stand the test of time: hydrophobic bases facing inward on each other and the phosphate-sugar backbone facing bulk water [ 4 ]. A few years later, the first experimentally determined atomic scale 3D model of a macromolecule of life, the whale protein myoglobin at atomic resolution, was presented by the Kendrew/Perutz team in Cambridge [ 2 ]. Though seemingly chaotic at first glance, the α-helical substructure proposed by Pauling had been found. But perhaps even more consequential was the observation that “[...] non-polar residues make up the bulk of the interior of the molecule” and that polar residues dominate the surface [ 23 ]. The following years of pioneering work on the structure of DNA and water soluble proteins laid the groundwork for thermodynamic considerations of what governs their overall structure. This culminated in the general concept of the hydrophobic effect that stabilizes the gross structure of DNA and proteins in liquid water [ 24 , 25 ].

The Fluid Mosaic Model of the Biomembrane

It was Singer and Nicolson’s insight that the same is true for the third class of amphiphilic macromolecules of life. An insight, which freed the “classic” trilamellar model from its stabilizing sheath of water-soluble proteins and enabled the fluid mosaic model (Fig.  1 b) [ 26 ]. In this radical new model of biomembrane structure, it is the entropic punishment of exposing hydrophobic lipid acyl chains to bulk water that ensures the stability of the passive 2D lipid bilayer matrix while allowing mobility of its active ingredient, the transmembrane or lateral bound membrane proteins. Singer and Nicholson posited their new unit membrane to have an expected thickness of ~ 70 to 90 Å; considerably thinner than the ~ 120 Å of the trilamellar membrane. The great conceptual advance and the lucid writing of Singer and Nicolson’s 1972 article made their beautifully drawn cartoon of the new membrane model one of the most influential science illustrations of the twentieth century. The drawing still appears virtually unaltered in all standard biochemistry textbooks and is a powerful meme [ 27 ] within the life science community. In the fluid mosaic model of the membrane, amphiphilic membrane proteins float randomly in the passive matrix of a fluid bilayer of amphiphilic lipids. These, by virtue of the self-organization induced by the hydrophobic effect, have their hydrophobic acyl chain facing inward and their hydrophilic headgroups facing bulk water (see Table  1 for a list of properties of the Fluid Mosaic Membrane Model from 1972). Though it is organized as a 2D matrix, Singer and Nicolson stressed the 3D nature of the biomembrane: both matrix lipids and membrane proteins have direction and an asymmetric distribution. Thus, in an important shift, Singer and Nicolson’s model broke with the symmetric depiction of all previous biomembrane models. This new image of the asymmetric biomembrane fitted well into novel concepts of the biomembrane acting as a permissible capacitor for electric signal transmission in electrophysiology [ 28 ], as well as the paradigm shift announced by Peter Mitchell’s formulation of bioenergetics based on proton gradients across asymmetric biomembranes as the pivotal energy intermediate in oxidative phosphorylation (OXPHOS) [ 29 ]. Since its publication the fluid mosaic model has been modified or extended in numerous publications that deal with one or several aspects where the model’s assumptions collide with experimental evidence (see Table  1 for an incomplete list). For instance, a review by Engelman from 2005 stressed on the patchiness and crowdedness of most cellular membranes which is in clear conflict to the image of free floating membrane proteins depicted in the cartoon of Fig.  1 b [ 30 ]. Notwithstanding, most discussions on the subject, including this essay, still take place in the context established by Singer–Nicolson.

Direct Experimental Insights into the Structure of the Biomembrane

The purple membrane.

Finally, in 1975, the first direct experimental visualization of a membrane protein in the context of its physiological membrane was achieved: bacteriorhodopsin of the natural highly ordered 2D crystal of the purple membrane of the archaebacteria Halobacterium halobium (Fig.  1 c) [ 31 ]. Though limited to about 7 Å resolution, the 3D reconstruction obtained by electron crystallography clearly showed bacteriorhodopsin to span the total of the membrane and thus beautifully demonstrated Singer and Nicolson’s concept of transmembrane, amphiphilic membrane proteins to be correct. At the same time, the purple membrane’s highly ordered, tight arrangement of bacteriorhodopsin membrane proteins—the very reason it was possible to use electron crystallography to analyse its structure—made the first experimental modification to the concept of free floating membrane proteins. In fact, it is only recently that the new technique of high speed AFM made it possible to recognize that bacteriorhodopsin is not only too tightly packed to float freely in the membrane, but also that light-induced movements of the bacteriorhodopsin trimer are possibly coordinated across the whole purple membrane via long range interactions [ 32 ]. In their seminal 1975 study of the purple membrane by electron crystallography, Henderson and Unwin measured the purple membrane’s thickness to be ~ 45 Å, which is considerably thinner than the 70–90 Å that Singer and Nicolson had predicted. It took another 22 years and the development of a cryo-electron microscope equipped with a liquid helium-cooled specimen stage [ 33 ] to improve the 3D map derived from electron crystallography to a level sufficient to yield atomic coordinates for both protein and lipid constituents of the purple membrane (Fig.  2 a) [ 34 , 35 ]. This afforded the first precise measurement of the thickness of a physiological biomembrane: total thickness 42 Å and thickness of the hydrophobic core 30 Å. After a 70-year hiatus, Hugo Fricke’s value of 33 Å for the thickness of the red blood cell’s membrane suddenly seemed to be realistic. However, since the archaebacterial purple membrane from the extremophile Halobacterium halobium has a cell membrane very different from that of mammals, the universality of the findings was not clear. For example, it includes exotic components such as methyl group branched phytanoyl lipid chains (chain length C16), which is specific to archaebacteria, as a major component.

Atomic structures of membrane proteins in the biomembrane. a High-resolution analysis of the purple membrane by electron crystallography enabled the first view of a natural membrane at the atomic scale by direct structural methods [ 34 , 35 ]. Protein in grey, lipid headgroup heteroatoms (phosphor, oxygen and nitrogen) as blue spheres and hydrophobic archaebacterial specific branched phytanoyl acyl chains in yellow. A non-annular lipid can be seen at the upper left edge (PDB: 1AT9). All lipids shown are of the natural membrane modelled according to the crystallographic density map. b Structure of the mammalian water channel aquaporin-0 in a bilayer of the synthetic phospholipid DMPC [ 36 ]. Note that membrane dimensions are very similar to the evolutionary distant purple membrane. Colour coding as in a ; (PDB: 2B6O). All lipids shown are of the reconstituted membrane modelled according to the crystallographic density map. c Cut-through view of the mammalian brain water channel aquaporin-4 determined in the context of a full lipid bilayer by electron crystallography (left panel) [ 37 ] and in the context of detergent micelles by X-ray crystallography (right panel) [ 41 ]. Positions of water molecules are sharply defined (left panel) or smeared out (right side), possibly as a consequence of a dielectric constant-dependent change in the strength of the dipole moment of two short alpha-helices (depicted as ribbon diagrams). Drawing based on [ 42 ]. Colour coding as in a (PDB: 3IYZ and 3GD8). Positioning of the cartoon membrane or detergent micelle relative to the membrane protein is based on the crystallographically determined position of the non-natural reconstituted membrane. (Color figure online)

Water Channels in the Membrane

Following these observations, in 2005, another set of atomic coordinates was reported for a membrane protein as flat as bacteriorhodopsin, together with its embedding membrane composed of the synthetic lipid DMPC (Fig.  2 b) [ 36 ]. Again, this was achieved by electron crystallography making use of the same liquid helium-cooled cryo-electron microscope; but in this instance, the membrane protein was the mammalian eye lens protein aquaporin-0, a member of the water channel family extracted from the eye lens of New Zealand sheep. The lipid was a relatively short (chain length C14) synthetic phosphocholine, yet the dimensions of the membrane were astonishingly similar to those of the archaebacterial purple membrane with a total thickness of ~ 50 Å and a hydrophobic core of ~ 28 Å. A similar membrane thickness was indicated by the visualization of synthetic lipids from both leaflets in direct contact with the membrane protein in 2D crystals of rat aquaporin-4, a water channel found in ganglia cells of the brain [ 37 ]. Perhaps these findings are an indication of an evolutionary pressure to conserve a membrane thickness of ~ 50 Å and a hydrophobic core of ~ 30 Å: a strong case for Robertson’s idea of a unit membrane with similar dimensions across all kingdoms of life.

The biological role of aquaporins is to facilitate the fast (~ 10 9 /s) transfer of water molecules across cell membranes while preventing protons from crossing the membrane. However, water has the ability to form proton wires via the Grotthus mechanism, which is biologically important for membrane proteins involved in proton pumping such as the mammalian cytochrome c oxidase, for example [ 38 ]. For this reason, the avoidance of proton transport in the presence of a chain of transmembrane water molecules is very surprising. A suggestion of how this proton exclusion might work while allowing very fast water transfer across the membrane was based on the first atomic model of a water channel, the water channel aquaporin-1 isolated from human red blood cells [ 39 ]. In this model, the dipole moment of two short alpha helices, tilted and fully immersed in the transmembrane region of the water channel, re-orients water molecules passing through the channel at its centre such that any potential proton wires are disrupted by breaking the continuous inter-water hydrogen bonding between the bulk water phases separated by the membrane. In this context, it is important to note that the strength of the effective alpha helical dipole moment depends on the dielectric constant ε of its chemical environment, with strength increasing for lower ε values (bulk water ε  = ~ 80; lipidic core of a membrane ε  = ~ 2). This results in a relatively strong dipole moment for short helices immersed in the hydrophobic core of the membrane [ 40 ]. The idea that an electric field is involved in breaking the proton wire was given a boost by high-resolution structures of aquaporin-4 determined in the membrane by electron crystallography [ 37 ] and in detergent micelles by X-ray crystallography (Fig.  2 c) [ 41 ]. Despite the protein structures being almost identical, the water position for the structure solved in the low ε environment of a full lipid bilayer were sharply defined, whereas in the high ε environment of the detergent micelle they were smeared out [ 42 ]. Tilted, short alpha helices in the transmembrane region have also been reported for numerous ion channels where they appear to point at the channel traversing ions [ 43 – 45 ]. Interestingly, highly tilted transmembrane alpha-helices are also a hallmark of cation transporting rotary ATPases [ 46 – 51 ]. Electric fields are difficult to visualize at the nano-scale level of biological membranes; however, it seems that cells exploit the hydrophobic core of cell membranes to strengthen the electric dipole moment at the end of membrane immersed alpha helices and thus harness internal electric fields for functions such as coordinating water molecules or ions.

Ion Channels in the Membrane

A well-known function of electric fields in biology is mediated by the voltage sensors attached to ion channels that sense changes in transmembrane potential for the gating of their respective ion channels, enabling the fast transmission of electric signals along the axons of neurons, for instance [ 52 ]. The strength of an electric field depends on the difference in electric potential and the combination of thickness and the value of the dielectric constant ε of the charge-separating insulator. At about 30 Å, the hydrophobic, insulating core of the cell membrane is already much thinner than was originally expected, enabling strong electric fields across the membrane at modest transmembrane potentials. Still, molecular dynamics simulation has suggested that in the close surroundings of voltage sensors (at the region where positively charged arginine residues enable the sensing of an electric field) the membrane is both wetted by the presence of water molecules and thinned beyond the usual ~ 30 Å (Fig.  3 b) [ 54 ]. Given that water molecules confined to few molecule layers and restricted in their orientation have an exceptionally low dielectric constant of ~ 2 [ 53 ], the presence of water in the voltage sensor is unlikely to raise the local ε value. It has been proposed that the combination of wetting and membrane thinning results in the focusing of the electric field generated by the transmembrane potential onto the electric potential sensing charged arginine residues [ 54 ]. The inability to visualize voltage sensors in the context of a lipid bilayer experiencing a transmembrane potential, however, means that experimental verification of this proposal has not been possible yet. Still, in support of this proposal, visualization of individual lipid molecules sandwiched between ion pore and voltage sensor clearly showed that voltage sensors and ion pore do not form one tight protein entity in the membrane (Fig.  3 a) [ 55 ]. Local manipulation of membrane thickness for focusing of the transmembrane electric field onto charged residues might not be limited to voltage sensors of ion conducting channels, but could include ion channel-independent, proton conducting voltage-sensing proteins [ 56 ].

Ion channels in the biomembrane. a Cut-through view of a voltage sensing potassium channel [ 55 ]. Voltage sensor domains are physically separated from the ion channel domain by lipids of the embedding membrane and flexibly connected to the ion pore via extramembranous loops. Protein in grey with voltage sensor domain shaded in red, lipids in yellow, potassium ions in purple (PDB: 2R9R). Positioning of the reconstituted non-natural lipids is based on a crystallographic density map. b An isolated voltage sensor domain with wetted arginine residues and a locally deformed membrane. Protein in grey, ribbon diagram representation with arginines as blue ball and stick models, transmembrane water molecules in red/white. Drawing based on [ 54 ]. Positioning of the cartoon membrane relative to the voltage-sensor domain is based on results from neutron diffraction of reconstituted, non-natural membranes and molecular dynamic simulations. c Cut-through view of tubular crystals of the acetylcholine receptor from postsynaptic membranes of the electric organ from the Atlantic fish Torpedo marmorata. A rare close-up view of a natural cholesterol-rich membrane including large areas of non-annular lipids. For clarity, receptors in the right half of the cut-view are depicted schematically. The position of the membrane is indicated in yellow and blue. Drawing based on [ 59 ]. Positioning of the natural membrane relative to the ion channels is based on a density map obtained by cryo-electron microscopy. (Color figure online)

The only case where an ion channel has been visualized at great detail in the same experimental data set as its physiological membrane is that of the helical tubes of the nicotinic acetyl-choline receptor which can be obtained from the postsynaptic membranes of the electric organ of the Atlantic fish Torpedo ray (Fig.  3 c) [ 57 ]. The density map obtained by cryo-electron microscopy of the tubes, first by electron crystallography [ 58 ] and recently by real space averaging, allows the precise positioning of the natural lipid bilayer with its two hydrophilic headgroup regions and the hydrophobic core relative to the membrane spanning ion channel [ 59 ]. This provides a rare chance for direct measurement of the cell membrane dimension in the presence of a transmembrane protein which unlike bacteriorhodopsin or water channels is not flatly embedded in the membrane, but has large extramembranous hydrophilic domains immersed in bulk water on both sides of the membrane. Though not revealing atomic coordinates of individual lipids, the density map shows the bilayer to have a thickness of ~ 45 Å with a hydrophobic core of ~ 30 Å both at the site of contact with acetylcholine-receptors and the membrane region between them.

Rotary ATPases Manipulate the Membrane

A completely unexpected position for a natural lipid bilayer was found by X-ray crystallography of the sodium ion transporting K-ring of the bacterial V-ATPase from Enterococcus hirae (Fig.  4 d) [ 60 ]. The bilayer visualized is not that of the membrane surrounding the K-ring, but the lipid bilayer occupying the lumen of this ring-shaped transmembrane protein. The inner diameter of the K-ring is wide enough to allow the presence of a number of cardiolipin lipids [ 61 ] large enough to form a full luminal lipid bilayer. Instead of being aligned to the height of the surrounding membrane, the luminal bilayer is shifted to the extracellular side by half the width of a hydrophobic membrane core, i.e. ~ 15 Å. This remarkable shift effectively leaves the lipid headgroups on the inside of the transmembrane ring at the height of the centre of the surrounding membrane, resulting in a thinning of the hydrophobic path between the ring lumen on the intracellular side and the extracellular region of the K-ring. Since V-ATPases, analogous to voltage sensors, are utilizing a transmembrane potential to fulfil their biological roles, shortening of the hydrophobic path between both sides of the membrane might have a role in stabilizing the assembly during rotary catalysis [ 62 ].

Redrawn after [ 63 ]. The cartoon of a mitochondrial cristae was based on observations by freeze-fracture electron microscopy. b In situ electron tomographic analysis of the inner mitochondrial membrane from yeast showed that mitochondrial F-ATP synthase forms dimers at the high positive curvature edges of cristae [ 64 ] (EMDB: 2161). Position and shape of the indicated natural membrane in blue and yellow was obtained by direct structural analysis via subtomogram averaging of cryo-electron tomograms of the natural membrane. c Cryo-EM structures of the detergent solubilized and of the lipid bilayer reconstituted monomeric bovine F-ATP synthase demonstrated that the shape of the mammalian mitochondrial F-ATP synthase alone is sufficient to bend the membrane [ 66 , 67 ] (EMDB: 3167). Position and shape of the cartoon membrane are based on cryo-electron tomograms of in vitro synthetic lipid reconstituted F-ATP synthase. d Cut-through view of the K-ring of the sodium pumping V-ATPase from Enterococcus hirae [ 60 ]. The luminal membrane is off-set relative to the embedding membrane by half a membrane thickness (PDB: 2BL2). Position and thickness of the luminal cartoon membrane are based on the crystallographically visualized natural luminal cardiolipin lipids, whereas the position of the surrounding cartoon membrane is based on the crystallographically visualized bound synthetic detergent molecules and bound sodium ions. (Color figure online)

Rotary ATPases manipulate the biomembrane. a The first model proposing that rows of dimeric mitochondrial F-ATP synthases are shaping the architecture of the inner mitochondrial cristae membrane.

Though the K-ring structure shows that V-ATPases are capable of inducing a vertical shift of rotor ring luminal lipid bilayers in respect to their surrounding bilayer, as in all previous examples, V-ATPases are expected to reside in relatively flat areas of the membrane. In contrast, their molecular cousins in the family of rotary ATPases, the mitochondrial F-ATP synthases, have been shown to form long rows of dimers at the regions of high positive membrane curvature (Fig.  4 a, b) [ 63 , 64 ]. The ridges of the cristae of the inner mitochondrial membrane are the location where the transmembrane potential built up by the electron transport chain is used to generate ATP from ADP and P i via rotary catalysis [ 65 ]. Oligomers of mitochondrial F-ATP synthase dimers are not just preferentially sequestered to regions of high positive curvature; rather, membrane bending is a property of the monomeric F-ATP synthase per se , as indicated by single particle cryo-EM of the detergent micelle solubilized bovine complex (Fig.  4 c) [ 66 ] and conclusively demonstrated by cryo-electron tomography of 2D crystals of the monomeric bovine F-ATP synthase reconstituted into a synthetic lipid bilayer [ 67 ]. The clear tram-track features of the bilayer in the tomograms indicate that, while sharply bent by ~ 40°, bilayer thickness is preserved. A recent high-resolution cryo-EM structure of the transmembrane Fo domain of the mitochondrial F-ATP synthase from yeast suggests that the molecular mechanism by which the membrane is bent might be similar to that of BAR domain proteins [ 68 ]. BAR domain proteins, such as Endophilin A1, are capable to sense and induce positive membrane curvature. The molecular mechanism has been well studied, and it appears that it involves two independent effects: rigid electrostatic interaction with the headgroup region of the membrane, i.e. the kink of the BAR domain protein is forced upon the membrane, or by the insertion of amphiphilic alpha helices into the headgroup region of the membrane. In the latter mechanism displacing lipid headgroups induces the acyl chains of surrounding lipid molecules to enter the space below the inserted amphiphatic alpha helices and thus effectively reducing the local average lipid acyl chain volume in one leaflet of the membrane [ 69 , 70 ]. The lack of an atomic model for the membrane bending subunit of the mitochondrial F-ATP synthase disallows deciding which type of mechanism is pre-dominant and highlights the need for membrane protein structures in the context of their embedding membrane. An unexplained interesting feature of cryo-EM studies by single particle reconstruction of the dimeric form from yeast mitochondria is the unassigned density at the monomer–monomer interface on the matrix side of the complex [ 71 ]. It is unclear whether this density stems from protein or lipid headgroups, or perhaps from detergent that was used to solubilize the membrane protein from its native membrane. In case this density stems from lipid headgroups of a lipid monolayer bridging opposing monomers of a dimer, it will be the first such case in membrane protein biology. And if true, this might be of consequence for the proposal that dimeric mitochondrial F-ATP synthases can form the mitochondrial permeability transition pore [ 72 – 74 ].

The Active Membrane

As described thus far, seen through the lens of high resolution structural studies the membrane takes a passive role. This is in line with the image of a matrix as proposed by Singer and Nicholson, which exerts its functional role chiefly through its composition, form and size. In contrast to this image of a mere matrix, structures of the TRAAK channel have shown that, while remaining chemically passive, the hydrophobic acyl chains of lipids can play a decisive active role in signal transduction (Fig.  5 a) [ 75 , 76 ]. The closing and opening of ion channels is mainly achieved through movement of the protein entity of the channel itself. This movement is a reaction to conformational changes induced by ligand binding, change in transmembrane voltage, or mechanical force exerted on it through the membrane or contacts with intracellular protein. For the TRAAK potassium ion channel, however, X-ray structures of its open and closed state visualized the entering of an individual acyl chain through lateral crevices in the transmembrane domain, blocking the ion pathway in the channel’s closed state. This is a surprising demonstration that lipids not integral to the channel itself can be the active agent in ion channel gating.

The active biomembrane. a X-ray crystal structures of the mechanical stress sensing human TRAAK potassium channel in open (left) and closed (right) conformation demonstrated gating of the ion pore to be achieved by a lipid acyl chain entering through a lateral crevice [ 75 , 76 ]. Protein in grey, potassium ions in purple, and acyl chain in yellow (PDB: 4WFF and 4WFE). b Large-scale movements of the transmembrane domain of the mammalian SERCA calcium pump depicted in grey ribbons are accommodated through rocking motions in the membrane of relatively constant thickness [ 81 ] (PDB: 5XA7 and 5XA8). The position of lipid headgroup phosphor atoms of the reconstituted synthetic bilayer indicated by blue spheres was crystallographically determined. (Color figure online)

X-ray crystallography of 3D crystals of membrane proteins usually does not visualize the physiological environment of the lipid bilayer. This is due to the fact that most 3D crystals are grown from detergent-solubilized membrane proteins in the absence of a lipid bilayer. This is also true for type I 3D crystals consisting of stacked layers of membrane proteins in lipid bilayers, because the lipids present in type I 3D crystals are too disordered to be resolved by analysis of the X-ray diffraction pattern, which originates only from the ordered portion of the 3D crystal. Type I 3D crystals of the Ca 2+ pump SERCA, a P-type ATPase, from the endoplasmic reticulum of rabbit muscle tissue have been used to solve high-resolution structures of many conformations of the enzyme’s pumping cycle [ 77 ]. The assumption that type I 3D crystals of stacked layers of SERCA do indeed contain a lipid bilayer had been confirmed by electron microscopy of thin 3D crystals [ 78 ]. Since SERCA’s Ca 2+ pumping cycle involves large conformational changes in hydrophobic transmembrane alpha helices perpendicular to the assumed membrane plane, i.e. apparently into bulk water, it was expected that the local membrane thickness would adjust itself during the reaction cycle to accommodate the transmembrane movement [ 79 , 80 ]. Recently, application of solvent contrast modulation X-ray crystallography allowed visualization of the lipids surrounding SERCA in the type I 3D crystal [ 81 ], and this approach was used to re-determine several structures of the pumping reaction cycle. By aligning the newly determined structures along the membrane plane, delineated by the phosphor atoms of the lipid headgroups, it became clear that it is not the membrane that adjusts its thickness to accommodate the large transmembrane movements. It is rather the SERCA Ca 2+ pump that adjusts itself in the membrane of a relatively constant hydrophobic thickness of ~ 31–33 Å, by tilting as a whole and thus undergoing rocking motions during the pumping cycle (Fig.  5 b). In this way, the exposure of hydrophobic alpha helices to bulk water during the pumping cycle is avoided. In other words, the membrane thickness dictates the membrane protein’s tilt angle in the membrane more than the membrane protein dictates the thickness of the membrane. Another surprising finding was the active role of lipids in the pumping cycle. Unlike the case of the TRAAK channel, this does not occur via acyl chains, but via interaction of lipid headgroups making electrostatic contacts with charged amino acid residues apparently aiding the movement of the reaction cycle. Thus far, most experimental approaches that visualize membrane proteins are unable to detect individual lipids. Therefore, it might turn out that lipids do play a much more active role than presently suspected.

New approaches such as solvent contrast X-ray crystallography as applied to the type I 3D crystals of the Ca 2+ pump SERCA bring immensely valuable, novel insights into membrane biology; and their results are almost always surprising. Regrettably, though it has become more common with the maturation of the lipid cubic phase approach, growing large type I 3D crystals is far from trivial. Likewise, highly ordered 2D crystals, either naturally occurring like the purple membrane or grown in vitro like aquaporin-0, are very rare. Consequently, the statistics for atomic coordinates of membrane proteins with full lipid bilayers are very poor, with perhaps not more than 7 entries in the protein data bank versus a total number of more than 149,000; of those 7, none was obtained under physiological non-equilibrium conditions, that is, structure determination always occurred in the absence of a membrane potential. Clearly the age of structural membrane biology has hardly begun and new tools are necessary to allow an atomistic view on the biomembranes themselves. An exciting new approach towards structure determination of membrane proteins in their physiological environment of the membrane is realized by single particle cryo-EM of membrane proteins reconstituted into the lipid bilayer of nano-discs. This allows visualization of functionally important lipids together with the membrane protein [ 82 ]. Furthermore, strategies that combine liposomes with reconstituted membrane proteins and single particle cryo-EM might yield structures of integral membrane proteins under truly physiological conditions [ 83 , 84 ]. However, reaching a resolution that is high enough to visualize lipids together with the membrane proteins will be challenging. Perhaps in the long run, cryo-electron tomography of planar lipid bilayers under non-equilibrium conditions will be the best option after the remaining technological challenges of data collection and image analysis of cryo-electron tomography have been solved. Likewise, for probing the thickness of natural biomembranes at the nano-scale in situ cryo-electron tomography is prone to be the best technique with interesting results already at hand [ 85 , 86 ]. In high-resolution structures of integral membrane proteins determined by X-ray crystallography or cryo-EM, the correct assignment of lipids, even when tightly bound, is often error-prone due to the mobility of their headgroups. And less tightly bound lipids interacting with integral membrane proteins are usually lost during detergent mediated membrane protein purification. Thus, though desirable, most lipid–protein interaction is out of reach for direct structural methods. A feasible way out of this conundrum is the combination of molecular dynamic simulations (MD) with insights from X-ray crystallography. This approach is exemplified by the above-described results on SERCA or also by the first in molecular dynamic simulations predicted and then later by X-ray crystallography demonstrated interaction of PIP 2 lipids with the Kir potassium ion channel [ 87 , 88 ]. Very recently, a novel powerful approach in native mass spectrometry of membrane complexes allowed their direct ejection from native membranes [ 89 ]. This completely avoids the artefacts usually associated with detergent-based purification and thus allows to detect lipid species that are only weakly bound to integral membrane proteins. In combination with molecular dynamic simulations, this now opens up a completely new window on probing weak interactions of lipids with protein complexes in the native membrane as was demonstrated here on the interplay of the mitochondrial adenine nucleotide translocase (ANT-1) with fatty acids [ 89 ]. Finally, atomic scale insights into whole assemblies of membrane proteins in membranes of native composition will likely stay accessible only to in silico methods as shown recently for an ensemble of GPCRs in a membrane of physiological composition and asymmetry [ 90 ].

Why ~ 30 Å?

A recurring theme in this essay is the thickness of the biomembrane, in particular of its hydrophobic core. It was shown that, against the expectations of the vast majority of biologists, the biomembrane is much thinner than was anticipated before the arrival of the first “hard structural data” in Henderson and Unwin’s seminal study on the purple membrane. As it turns out, Hugo Fricke’s measurement of 33 Å for the low ε portion of the red blood cell membrane was quite accurate. As noted before, with more than 40,000 lipid species, biomembranes are built up of an extremely diverse set of molecular compounds ( http://www.lipidmaps.org/data/ ). Nevertheless, the examples of direct structural insights into biomembrane situated transmembrane proteins described in this essay suggest that the thickness of their hydrophobic core is astonishingly constant (see Table 2 for a list of thickness values reported in the literature discussed in this essay). And endowed with specific function when deviating from this standard thickness. Naturally, given that the examples discussed here are few and the lipids visualized mostly annular to the embedded membrane proteins, the generalization of the precise thickness of the hydrophobic core of the biomembrane has to be taken with caution. It should also be noted that, of the atomic structures discussed, only that of the lipids in the purple membrane are of a natural membrane. All other are that of reconstituted membranes of selected lipids. In addition, the usually high cholesterol content of plasma membranes is reflected only in the vesicular tubes grown from the natural postsynaptic membranes of the fish Torpedo ray. Interestingly, the tubes apparently contain cholesterol enriched regions in their outer membrane leaflet, possibly aiding acetylcholine receptors’ gating [ 59 ]. The likely important interaction of acetylcholine receptors and cholesterol in membranes of mixed lipid content was recently explored by coarse-grained MD detecting a clear de-mixing of lipid species and receptors [ 91 ]. Early X-ray and neutron diffraction studies on 40% cholesterol/60% lecithin model membranes in the 1970ies gave a leaflet to leaflet distance for the cholesterol headgroup hydroxyls of ~ 39 Å—indicating a thickening of the hydrophobic core by several ångstrøm [ 92 ]. The calculated carbonyl to carbonyl distance of ~ 30 Å supports the notion of cholesterol induced membrane thickening. Indeed cholesterol-rich domains are thought to be crucial for the formation of thicker and functionally important raft-domains. Again this underlines that deviation of membrane thickness is connected to specific biological function. A later study by Wiener & White from 1991 that combined X-ray and neutron diffraction of hydrated pure DOPC bilayers by Wiener and White allowed to build a detailed model of the bilayer in the fluid state [ 93 – 95 ]. The high B-factors of the different chemical components that constitute the fluid bilayer such as phosphates, carbonyl groups or water molecules, demonstrated a rather broad Gaussian distribution of their position transversal to the membrane plane, i.e. strong transbilayer thermal motion. Nevertheless, the average carbonyl to carbonyl distance is, as expected, ~ 30 Å. In the universal structural elements of DNA/RNA and proteins, the thickness of an alpha helix or the helical pitch of the DNA double helix is constant dictated by the architecture of the molecular backbone of these biological polymers. In biomembranes, such universal structural elements common to all life forms are missing. Notwithstanding, perhaps the thickness of ~ 30 Å for the hydrophobic core of membranes is a biological constant that might teach us something about the biophysical essence of life. I am not aware of any compelling reason why the hydrophobic thickness of biomembranes should be ~ 30 Å, and not for instance ~ 15 Å or ~ 60 Å. Nor do I know a compelling reason why it appears to be so similar among the many different life forms on our planet while at the same time being composed of such a wide variety of lipid molecules. A possible reason might be evolutionary origin. Perhaps the thickness was fixed by the range of organic molecules available in the chemical environment of submarine hydrothermal vent microstructures for the build-up of the primordial membrane in the Last Universal Common Ancestor [ 96 , 97 ]. Rapid lateral gene transfer in early life forms of an RNA world [ 98 , 99 , 100 ] with bioenergetics based on leaky membranes [ 101 , 102 ], and a stream of basic hydrogen saturated water, might then have spread a normed transmembrane thickness of the earliest membrane proteins. Or perhaps ~ 30 Å is the minimum thickness that will block the passage of ions, in particular protons, and provide a sufficient stability by the hydrophobic effect, while still allowing a level of compliance compatible with membrane fusion events and sharp membrane bending. It might also be that a hydrophobic thickness of ~ 30 Å is ideal for the generation of electric fields by transmembrane potentials and alpha helical dipole moment. The experimental atomic scale probing of electric fields in biomembranes is a largely unchartered territory that only recently starts to yield to exploration in silico [ 103 – 105 ]. Therefore, geometrical constraints of membrane and membrane protein architecture might be determined more by the physics that govern transmembrane potential fall-off and the resulting electric fields than is clear at present.

Lateral and Transversal Variance

Though the leitmotif of this essay is the by structural methods relative accessible property of membrane thickness, the importance of lateral and transversal heterogeneity has to be mentioned. As described above, it appears that membrane thickness is more uniform than originally anticipated. In contrast, the lateral distribution of lipid species and transversal properties such as lateral pressure seem to be subject to far more variance than suggested by the fluid mosaic model of the biomembrane. A prominent example for lateral deviation from uniformity is lipid rafts [ 106 ]. These are cholesterol-rich domains in the lipid bilayer exhibiting lower fluidity and segregation from the more fluid and less ordered areas of the membrane. Lipid rafts are thought to act as specialized functional areas separating and clustering membrane proteins for differing physiological and pathophysiological functions such as signalling or virus budding. Their existence in the living cell and physiological relevance has been a matter of debate mostly due to the fact that good methods for their detection in the living cell were missing. This is a consequence of lipid rafts’ limited size which is in the nanometer range and their dynamic nature which is in the millisecond range [ 107 ]. However, the recent advent of super resolution light microscopy provides the field with a tool that can be used to dissect form and functions of lipid rafts in the membranes under physiological conditions [ 108 , 109 ]. Transversal non-uniformity of the biomembrane is immediately apparent in density profiles of wetted, fluid lipid bilayers which reveal two main areas of approximately equal thickness and radically different hydrophobicity and chemical environment: interfacial headgroup regions and the hydrocarbon core region. An environment far more complex than that of bulk water and crucial for determining the shape of membrane proteins [ 110 ]. Variations in the transversal properties of biomembranes as a function of membrane depth such as dipole moment or lateral pressure are even less accessible by direct experimental means than that of lateral variations in lipid composition. As a consequence, despite its biological importance, the property of lateral pressure remains the domain of in silico studies. Its significant variation in dependence on the presence of the non-lamellar lipid DOPE has been demonstrated by computer simulations [ 111 ]. These showed an almost constant membrane thickness while lateral pressure in the headgroup region decreased by up to 50% and increased in the lipid tail region by up to ~ 400%. This provides an indication of how lipid composition might modify membrane properties for the regulation of integral membrane protein function. Likewise the effect of general anesthetics has been proposed to be an effect of transversal changes in the lateral pressure profile resulting in a change in the conformational landscape of, e.g. sodium ion channels [ 112 ]. This proposal is now corroborated by molecular dynamic simulations which demonstrated that general anesthetics like chloroform, diethylether or enflurane insert into the hydrocarbon core and can cause an increase of lipid area at roughly constant membrane thickness resulting in a decrease in lateral pressure in the headgroup region at the position of the carbonyl groups [ 113 ].

Closing Statement

Cellular membranes are outside of the central dogma of molecular biology. Yet still they are as fundamental to life as are DNA, RNA and proteins. Naturally, the current limitations in our understanding of membrane biology are highly regrettable. That our knowledge on biomembranes lags so far behind that of the other molecules of life is mostly a consequence of the experimental difficulties of dealing with lipid bilayers and membrane proteins. Therefore, it is desirable and anticipated that advances especially in biophysical and computational methods will bring about a more realistic picture of the biomembranes allowing it to come into the place occupied by the fluid mosaic model for already more than 40 years.

Bragg WL (1914) The analysis of crystals by the X-ray spectrometer. Proc R Soc Lond A 89(613):468–489

Article   Google Scholar  

Kendrew JC, Dickerson RE, Strandberg BE, Hart RG, Davies DR, Phillips DC, Shore VC (1960) Structure of myoglobin. Nature 185(422):427–1960

Google Scholar  

Perutz MF, Rossmann MG, Cullis AF, Muirhead H, Will G, North ACT (1960) Structure of haemoglobin: a three-dimensional Fourier synthesis at 5.5-Å. resolution, obtained by X-ray analysis. Nature 185(4711):416

Article   CAS   PubMed   Google Scholar  

Watson JD, Crick FH (1953) Molecular structure of nucleic acids. Nature 171(4356):737–738

Franklin RE, Gosling RG (1953) Molecular configuration in sodium thymonucleate. Nature 171(4356):740

Berman H, Henrick K, Nakamura H (2003) Announcing the worldwide protein data bank. Nat Struct Mol Biol 10(12):980

Article   CAS   Google Scholar  

Membrane Proteins of Known 3D Structure; http://blanco.biomol.uci.edu/mpstruc/ . Accessed 13 Mar 2019

Schleiden MJ (1838) Beiträge zur Phytogenesis. Arch Anat Physiol Wiss Med (J Müller) 137–176

Schwann T (1839) Mikroscopische Untersuchungen über die Übereinstimmung in der Struktur und dem Wachstum der Thiere und Pflanzen. W Engelmann: Leipzig No 176

Meyer H (1899) Welche Eigenschaft der Anaesthetica bedingt ihre narkotische Wirkung? Naunyn-Schmiedeberg’s Arch Exp Pathol Pharmakol 42:109–118

Overton E (1901) Studien über die Narkose. Fischer, Jena

Gorter E, Grendel FJEM (1925) On bimolecular layers of lipoids on the chromocytes of the blood. J Exp Med 41(4):439–443

Article   CAS   PubMed   PubMed Central   Google Scholar  

Fricke H (1925) The electric capacity of suspensions with special reference to blood. J Gen Physiol 9(2):137–152

Danielli JF, Davson H (1935) A contribution to the theory of permeability of thin films. J Cell Physiol 5(4):495–508

Debye P, Hückel E (1923) De la theorie des electrolytes. I. abaissement du point de congelation et phenomenes associes. Physikalische Zeitschrift 24(9):185–206

CAS   Google Scholar  

Palade GE (1952) The fine structure of mitochondria. Anat Rec 114(3):427–451

Sjostrand FS (1953) Electron microscopy of mitochondria and cytoplasmic double membranes. Nature 171(4340):30

Robertson J, David (1959) The ultrastructure of cell membranes and their derivatives. Biochemical society symposium, vol 16

Korn ED (1966) Structure of biological membranes. Science 153(3743):1491–1498

Pauling L, Corey RB (1953) A proposed structure for the nucleic acids. Proc Natl Acad Sci 39(2):84–97

Pauling L, Corey RB, Branson HR (1951) The structure of proteins: two hydrogen-bonded helical configurations of the polypeptide chain. Proc Natl Acad Sci 37(4):205–211

Pauling L, Corey RB (1951) The pleated sheet, a new layer configuration of polypeptide chains. Proc Natl Acad Sci 37(5), 251–256

Kendrew JC (1963) Myoglobin and the structure of proteins. Science 139(3561):1259–1266

Tanford C (1962) Contribution of hydrophobic interactions to the stability of the globular conformation of proteins. J Am Chem Soc 84(22):4240–4247

Tanford C (1978) The hydrophobic effect and the organization of living matter. Science 200(4345):1012–1018

Singer SJ, Nicolson GL (1972) The fluid mosaic model of the structure of cell membranes. Science 175(4023):720–731

Dawkins R (1976) The selfish gene. Oxford university press, New York

Hodgkin AL, Katz B (1949) The effect of sodium ions on the electrical activity of the giant axon of the squid. J Physiol 108(1):37–77

Mitchell P (1961) Coupling of phosphorylation to electron and hydrogen transfer by a chemi-osmotic type of mechanism. Nature 191(4784):144–148

Engelman DM (2005) Membranes are more mosaic than fluid. Nature 438(7068):578–580

Henderson R, Unwin PNT (1975) Three-dimensional model of purple membrane obtained by electron microscopy. Nature 257(5521):28

Shibata M, Uchihashi T, Yamashita H, Kandori H, Ando T (2011) Structural changes in bacteriorhodopsin in response to alternate illumination observed by high-speed atomic force microscopy. Angew Chem 123(19):4502–4505

Fujiyoshi Y, Mizusaki T, Morikawa K, Yamagishi H, Aoki Y, Kihara H, Harada Y (1991) Development of a superfluid helium stage for high-resolution electron microscopy. Ultramicroscopy 38(3–4):241–251

Mitsuoka K, Hirai T, Murata K, Miyazawa A, Kidera A, Kimura Y, Fujiyoshi Y (1999) The structure of bacteriorhodopsin at 3.0 Å resolution based on electron crystallography: implication of the charge distribution1. J Mol Biol 286(3):861–882

Grigorieff N, Ceska TA, Downing KH, Baldwin JM, Henderson R (1996) Electron-crystallographic refinement of the structure of bacteriorhodopsin. J Mol Biol 259(3):393–421

Gonen T, Cheng Y, Sliz P, Hiroaki Y, Fujiyoshi Y, Harrison SC, Walz T (2005) Lipid–protein interactions in double-layered two-dimensional AQP0 crystals. Nature 438(7068):633

Tani K, Mitsuma T, Hiroaki Y, Kamegawa A, Nishikawa K, Tanimura Y, Fujiyoshi Y (2009) Mechanism of aquaporin-4’s fast and highly selective water conduction and proton exclusion. J Mol Biol 389(4):694–706

Shimada A, Kubo M, Baba S, Yamashita K, Hirata K, Ueno G et al (2017) A nanosecond time-resolved XFEL analysis of structural changes associated with CO release from cytochrome c oxidase. Sci Adv 3(7):e1603042

Murata K, Mitsuoka K, Hirai T, Walz T, Agre P, Heymann JB et al (2000) Structural determinants of water permeation through aquaporin-1. Nature 407(6804):599

Sengupta D, Behera RN, Smith JC, Ullmann GM (2005) The α helix dipole: screened out? Structure 13(6):849–855

Ho JD, Yeh R, Sandstrom A, Chorny I, Harries WE, Robbins RA et al (2009) Crystal structure of human aquaporin 4 at 1.8 Å and its mechanism of conductance. Proc Natl Acad Sci 106(18):7437–7442

Tani K, Fujiyoshi Y (2014) Water channel structures analysed by electron crystallography. Biochimica et Biophysica Acta (BBA) General Subjects 1840(5):1605–1613

Doyle DA, Cabral JM, Pfuetzner RA, Kuo A, Gulbis JM, Cohen SL et al (1998) The structure of the potassium channel: molecular basis of K + conduction and selectivity. Science 280(5360):69–77

Dutzler R, Campbell EB, Cadene M, Chait BT, MacKinnon R (2002) X-ray structure of a ClC chloride channel at 3.0 Å reveals the molecular basis of anion selectivity. Nature 415(6869):287

Payandeh J, Scheuer T, Zheng N, Catterall WA (2011) The crystal structure of a voltage-gated sodium channel. Nature 475(7356):353

Allegretti M, Klusch N, Mills DJ, Vonck J, Kühlbrandt W, Davies KM (2015) Horizontal membrane-intrinsic α-helices in the stator a-subunit of an F-type ATP synthase. Nature 521(7551):237

Zhou A, Rohou A, Schep DG, Bason JV, Montgomery MG, Walker JE et al (2015) Structure and conformational states of the bovine mitochondrial ATP synthase by cryo-EM. Elife 4:e10180

Article   PubMed   PubMed Central   Google Scholar  

Sobti M, Smits C, Wong AS, Ishmukhametov R, Stock D, Sandin S, Stewart AG (2016) Cryo-EM structures of the autoinhibited E. coli ATP synthase in three rotational states. Elife 5:e21598

Hahn A, Vonck J, Mills DJ, Meier T, Kühlbrandt W (2018) Structure, mechanism, and regulation of the chloroplast ATP synthase. Science 360(6389):eaat4318

Nakanishi A, Kishikawa JI, Tamakoshi M, Mitsuoka K, Yokoyama K (2018) Cryo EM structure of intact rotary H+-ATPase/synthase from Thermus thermophilus. Nat Commun 9(1):89

Roh SH, Stam NJ, Hryc CF, Couoh-Cardel S, Pintilie G, Chiu W, Wilkens S (2018) The 3.5-Å cryoEM structure of nanodisc-reconstituted yeast vacuolar ATPase Vo proton channel. Mol Cell 69(6):993–1004

Hille B (2001) Ion channels of excitable membranes, vol 507. Sinauer, Sunderland

Fumagalli L, Esfandiar A, Fabregas R, Hu S, Ares P, Janardanan A et al (2018) Anomalously low dielectric constant of confined water. Science 360(6395):1339–1342

Krepkiy D, Mihailescu M, Freites JA, Schow EV, Worcester DL, Gawrisch K et al (2009) Structure and hydration of membranes embedded with voltage-sensing domains. Nature 462(7272):473

Long SB, Tao X, Campbell EB, MacKinnon R (2007) Atomic structure of a voltage-dependent K+ channel in a lipid membrane-like environment. Nature 450(7168):376

Takeshita K, Sakata S, Yamashita E, Fujiwara Y, Kawanabe A, Kurokawa T, Okochi Y, Matsuda M, Narita H, Okamura Y, Nakagawa A (2014) X-ray crystal structure of voltage-gated proton channel. Nat Struct Mol Biol 21(4):352

Brisson A, Unwin PN (1984) Tubular crystals of acetylcholine receptor. J Cell Biol 99(4):1202–1211

Miyazawa A, Fujiyoshi Y, Unwin N (2003) Structure and gating mechanism of the acetylcholine receptor pore. Nature 423(6943):949

Unwin N (2017) Segregation of lipids near acetylcholine-receptor channels imaged by cryo-EM. IUCrJ 4(4):393–399

Murata T, Yamato I, Kakinuma Y, Leslie AG, Walker JE (2005) Structure of the rotor of the V-Type Na+-ATPase from Enterococcus hirae. Science 308(5722):654–659

Zhou M, Morgner N, Barrera NP, Politis A, Isaacson SC, Matak-Vinković D et al (2011) Mass spectrometry of intact V-type ATPases reveals bound lipids and the effects of nucleotide binding. Science 334(6054):380–385

Gerle C (2011) Stabilization of Fo/Vo/Ao by a radial electric field. Biophysics 7:99–104

Allen RD, Schroeder CC, Fok AK (1989) An investigation of mitochondrial inner membranes by rapid-freeze deep-etch techniques. J Cell Biol 108(6):2233–2240

Davies KM, Anselmi C, Wittig I, Faraldo-Gómez JD, Kühlbrandt W (2012) Structure of the yeast F1Fo-ATP synthase dimer and its role in shaping the mitochondrial cristae. Proc Natl Acad Sci 109(34), 13602–13607

Boyer PD (1997) The ATP synthase—a splendid molecular machine. Ann Rev Biochem 66(1):717–749

Baker LA, Watt IN, Runswick MJ, Walker JE, Rubinstein JL (2012) Arrangement of subunits in intact mammalian mitochondrial ATP synthase determined by cryo-EM. Proc Natl Acad Sci 109(29), 11675–11680

Jiko C, Davies KM, Shinzawa-Itoh K, Tani K, Maeda S, Mills DJ et al (2015) Bovine F1Fo ATP synthase monomers bend the lipid bilayer in 2D membrane crystals. Elife 4:e06119

Guo H, Bueler SA, Rubinstein JL (2017) Atomic model for the dimeric FO region of mitochondrial ATP synthase. Science 358(6365):936–940

Peter BJ, Kent HM, Mills IG, Vallis Y, Butler PJG, Evans PR, McMahon HT (2004) BAR domains as sensors of membrane curvature: the amphiphysin BAR structure. Science 303(5657):495–499

Ambroso MR, Hegde BG, Langen R (2014) Endophilin A1 induces different membrane shapes using a conformational switch that is regulated by phosphorylation. Proc Natl Acad Sci 111:6982–6987

Hahn A, Parey K, Bublitz M, Mills DJ, Zickermann V, Vonck J et al (2016) Structure of a complete ATP synthase dimer reveals the molecular basis of inner mitochondrial membrane morphology. Mol Cell 63(3):445–456

Giorgio V, Von Stockum S, Antoniel M, Fabbro A, Fogolari F, Forte M et al (2013) Dimers of mitochondrial ATP synthase form the permeability transition pore. Proc Natl Acad Sci 110(15), 5887–5892

Bernardi P, Rasola A, Forte M, Lippe G (2015) The mitochondrial permeability transition pore: channel formation by F-ATP synthase, integration in signal transduction, and role in pathophysiology. Physiol Rev 95(4):1111–1155

Gerle C (2016) On the structural possibility of pore-forming mitochondrial FoF1 ATP synthase. Biochimica et Biophysica Acta (BBA)-Bioenergetics 1857(8):1191–1196

Brohawn SG, Campbell EB, MacKinnon R (2014) Physical mechanism for gating and mechanosensitivity of the human TRAAK K + channel. Nature 516(7529):126

Brohawn SG, Su Z, MacKinnon R (2014) Mechanosensitivity is mediated directly by the lipid membrane in TRAAK and TREK1 K+ channels. Proc Natl Acad Sci 111(9):3614–3619

Toyoshima C (2008) Structural aspects of ion pumping by Ca2+-ATPase of sarcoplasmic reticulum. Arch Biochem Biophys 476(1):3–11

Ogawa H, Stokes DL, Sasabe H, Toyoshima C (1998) Structure of the Ca2 + pump of sarcoplasmic reticulum: a view along the lipid bilayer at 9-Å resolution. Biophys J 75(1):41–52

Toyoshima C, Mizutani T (2004) Crystal structure of the calcium pump with a bound ATP analogue. Nature 430(6999):529

Sørensen TLM, Møller JV, Nissen P (2004) Phosphoryl transfer and calcium ion occlusion in the calcium pump. Science 304(5677):1672–1675

Norimatsu Y, Hasegawa K, Shimizu N, Toyoshima C (2017) Protein–phospholipid interplay revealed with crystals of a calcium pump. Nature 545(7653):193

Gao Y, Cao E, Julius D, Cheng Y (2016) TRPV1 structures in nanodiscs reveal mechanisms of ligand and lipid action. Nature 534(7607):347

Sejwal K, Chami M, Baumgartner P, Kowal J, Müller SA, Stahlberg H (2017) Proteoliposomes–a system to study membrane proteins under buffer gradients by cryo-EM. Nanotechnol Rev 6(1):57–74

Wang L, Sigworth FJ (2009) Structure of the BK potassium channel in a lipid membrane from electron cryomicroscopy. Nature 461(7261):292

Engel BD, Schaffer M, Cuellar K, Villa L, Plitzko E, J. M., & Baumeister W (2015) Native architecture of the Chlamydomonas chloroplast revealed by in situ cryo-electron tomography. eLife 4:e04889

Ader NR, Hoffmann PC, Ganeva I, Borgeaud AC, Wang C, Youle RJ, Kukulski W (2019) Molecular and topological reorganizations in mitochondrial architecture interplay during Bax-mediated steps of apoptosis. eLife 8:e40712

Stansfeld PJ, Hopkinson R, Ashcroft FM, Sansom MS (2009 ) PIP2-binding site in Kir channels: definition by multiscale biomolecular simulations. Biochemistry 48(46):10926–10933

Hansen SB, Tao X, MacKinnon R (2011) Structural basis of PIP 2 activation of the classical inward rectifier K+ channel Kir2. 2. Nature 477(7365):495

Chorev DS, Baker LA, Wu D, Beilsten-Edmands V, Rouse SL, Zeev-Ben-Mordehai T, Jiko C, Samsudin F, Gerle C, Khalid S, Stewart AG (2018) Protein assemblies ejected directly from native membranes yield complexes for mass spectrometry. Science 362(6416):829–834

Koldsø H, Sansom MS (2015) Organization and dynamics of receptor proteins in a plasma membrane. J Am Chem Soc 137(46):14694–14704

Sharp L, Salari R, Brannigan G (2019) Boundary lipids of the nicotinic acetylcholine receptor: spontaneous partitioning via coarse-grained molecular dynamics simulation. Bbamem. https://doi.org/10.1016/j.bbamem.2019.01.005

Franks NP (1976) Structural analysis of hydrated egg lecithin and cholesterol bilayers. I. X-ray diffraction. J Mol Biol 100(3):345–358

Wiener MC, White SH (1991 Jan) Fluid bilayer structure determination by the combined use of X-ray and neutron diffraction. I. Fluid bilayer models and the limits of resolution. Biophys J 59(1):162–173

Wiener MC, White SH (1991) Fluid bilayer structure determination by the combined use of X-ray and neutron diffraction. II.” Composition-space” refinement method. Biophys J 59(1):174–185

Wiener MC, White SH (1992) Structure of a fluid dioleoylphosphatidylcholine bilayer determined by joint refinement of X-ray and neutron diffraction data. III. Complete structure. Biophys J 61(2):434–447

Russell MJ, Hall AJ, Martin W (2010) Serpentinization as a source of energy at the origin of life. Geobiology 8(5):355–371

Lane N, Allen JF, Martin W (2010) How did LUCA make a living? Chemiosmosis in the origin of life. BioEssays 32(4):271–280

Crick FH (1968) The origin of the genetic code. Journal of molecular biology 38(3):367–379

Orgel LE (1968) Evolution of the genetic apparatus. J Mol Biol 38(3):381–393

Gilbert W (1986) Origin of life: the RNA world. Nature 319(6055):618

Mulkidjanian AY, Makarova KS, Galperin MY, Koonin EV (2007) Inventing the dynamo machine: the evolution of the F-type and V-type ATPases. Nat Rev Microbiol 5(11):892

Mulkidjanian AY, Galperin MY, Koonin EV (2009) Co-evolution of primordial membranes and membrane proteins. Trends Biochem Sci 34(4):206–215

Capponi S, Heyden M, Bondar AN, Tobias DJ, White SH (2015) Anomalous behavior of water inside the SecY translocon. Proc Natl Acad Sci 112(29):9016-9021

Freites JA, Schow EV, White SH, Tobias DJ (2012) Microscopic origin of gating current fluctuations in a potassium channel voltage sensor. Biophys J 102(11):L44-L46

Article   CAS   PubMed Central   Google Scholar  

Vermaas JV, Tajkhorshid E (2016) Differential membrane binding mechanics of synaptotagmin isoforms observed in atomic detail. Biochemistry 56(1):281–293

Sezgin E, Levental I, Mayor S, Eggeling C (2017) The mystery of membrane organization: composition, regulation and roles of lipid rafts. Nat Rev Mol Cell Biol 18(6):361

Levental I, Veatch SL (2016) The continuing mystery of lipid rafts. J Mol Biol 428(24):4749–4764

Eggeling C, Ringemann C, Medda R, Schwarzmann G, Sandhoff K, Polyakova S, Belov VN, Hein B, von Middendorff C, Schönle A, Hell SW (2009) Direct observation of the nanoscale dynamics of membrane lipids in a living cell. Nature 457(7233):1159

Owen DM, Williamson DJ, Magenau A, Gaus K (2012) Sub-resolution lipid domains exist in the plasma membrane and regulate protein diffusion and distribution. Nat Commun 3:1256

White SH, Ladokhin AS, Jayasinghe S, Hristova K (2001) How membranes shape protein structure. J Biol Chem 276(35):32395–32398

Ding W, Palaiokostas M, Wang W, Orsi M (2015) Effects of lipid composition on bilayer membranes quantified by all-atom molecular dynamics. J Phys Chem B 119(49):15263–15274

Cantor RS (1997) The lateral pressure profile in membranes: a physical mechanism of general anesthesia. Biochemistry 36(9):2339–2344

Fábián B, Sega M, Voloshin VP, Medvedev NN, Jedlovszky P (2017) Lateral pressure profile and free volume properties in phospholipid membranes containing anesthetics. J Phys Chem B 121(13):2814–2824

Download references

Acknowledgements

This essay is based on a lecture given at the graduate school of biomedical sciences at the University of Padova upon invitation by Paolo Bernardi. I would like to thank Nigel Unwin for critical reading of the manuscript and also for providing the map that was used to produce Fig.  3 c. I am also grateful for graphical support by Bernhard C. Ludewig and help with the English language by Andrew C. Elliott.

This study was funded by a CREST Grant JPMJCR13M4 (to Genij Kurisu and C.G.) from JST, Japan, a Grants-in-Aid for Scientific Research (Kiban B: 17H03647) from MEXT, Japan, and a BINDS grant from AMED, Japan.

Author information

Authors and affiliations.

Institute for Protein Research, Osaka University, Suita, Japan

Christoph Gerle

Core Research for Evolutional Science and Technology, Japan Science and Technology Agency, Kawaguchi, Japan

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Christoph Gerle .

Ethics declarations

Conflict of interest.

The authors declare that they have no conflict of interest.

Research Involving Human and Animal Participants

This article does not contain any studies with human participants or animals performed by any of the authors.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License ( http://creativecommons.org/licenses/by/4.0/ ), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Reprints and permissions

About this article

Gerle, C. Essay on Biomembrane Structure. J Membrane Biol 252 , 115–130 (2019). https://doi.org/10.1007/s00232-019-00061-w

Download citation

Received : 06 January 2019

Accepted : 28 February 2019

Published : 15 March 2019

Issue Date : 15 June 2019

DOI : https://doi.org/10.1007/s00232-019-00061-w

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Membrane structure
• Membrane protein
• Lipid bilayer
• Singer–Nicolson
• History of science
• Structural biology
• Find a journal
• Publish with us
• Track your research

• school Campus Bookshelves
• menu_book Bookshelves
• perm_media Learning Objects
• login Login
• how_to_reg Request Instructor Account
• hub Instructor Commons
• Download Page (PDF)
• Download Full Book (PDF)
• Periodic Table
• Physics Constants
• Scientific Calculator
• Reference & Cite
• Tools expand_more
• Readability

selected template will load here

This action is not available.

7: Module 5- Cell Membranes

• Last updated
• Save as PDF
• Page ID 43580
• 7.1: Why It Matters- Cell Membranes
• 7.2: Introduction to Structure of the Membrane
• 7.3: Structure of the Cell Membrane
• 7.4: Phospholipids
• 7.5: Introduction to Kinds of Transport
• 7.6: Passive Transport
• 7.7: Active Transport
• 7.8: Membranes and Transport
• 7.9: Introduction to Endocytosis and Exocytosis
• 7.10: Endocytosis
• 7.11: Exocytosis
• 7.12: Putting It Together- Cell Membranes
• 7.13: Assignment- Membranes Alive!

If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

To log in and use all the features of Khan Academy, please enable JavaScript in your browser.

Unit 11: More about membranes

Cell membranes.

• Cell membrane introduction (Opens a modal)
• Cell membrane overview and fluid mosaic model (Opens a modal)
• Fluid mosaic model: cell membranes article (Opens a modal)
• Cell membrane proteins (Opens a modal)
• The cell membrane review (Opens a modal)
• Water potential example (Opens a modal)
• Plasma membranes 4 questions Practice
• Membrane permeability 4 questions Practice
• Passive transport review (Opens a modal)
• Active transport review (Opens a modal)
• Mechanisms of transport: tonicity and osmoregulation 4 questions Practice
• Facilitated diffusion 4 questions Practice
• Membrane transport 4 questions Practice

• International
• Schools directory
• Resources Jobs Schools directory News Search

Top Mark A-Level Biology essay - The membranes of different types of cells & their functions

Subject: Biology

Age range: 16+

Resource type: Assessment and revision

Last updated

26 September 2018

• Share through email
• Share through twitter
• Share through linkedin
• Share through facebook
• Share through pinterest

Top Mark A-Level Biology essay addressing the title: The membranes of different types of cells are involved in many different functions

• Holistic introduction
• Properties of membranes
• Components present in membranes - Proteins and channels
• Passage of substances across membranes - Passive and facilitated diffusion, co-transport, active transport
• Ultrafiltration
• Plasma receptors - Blood glucose control, immune response
• Membranes in nervous communication
• Synaptic plasticity

Contains information from beyond the syllabus. Examples are structured in a clear and logical format.

Useful for revision for this specific topic and to see a sample structure of a high grade essay.

Tes paid licence How can I reuse this?

Get this resource as part of a bundle and save up to 71%

A bundle is a package of resources grouped together to teach a particular topic, or a series of lessons, in one place.

Top mark A-Level Biology Essays

A collection of high mark, sample essays for A-Level Biology, providing helpful revision content and examples of logical structuring. New Specification! Includes essays on the titles: - The Importance of movement in living organisms - Explain the importance of shapes fitting together in cells and organisms - Carbon dioxide may affect organisms directly or indirectly. - How is energy transferred within and between organisms - Explain the importance of shapes fitting together in cells and organisms - Different types of relationships and interactions between organisms - The membranes of different types of cells are involved in many different functions - The importance of DNA in science and technology Save over 70%!

Your rating is required to reflect your happiness.

It's good to leave some feedback.

Something went wrong, please try again later.

This resource hasn't been reviewed yet

To ensure quality for our reviews, only customers who have purchased this resource can review it

Report this resource to let us know if it violates our terms and conditions. Our customer service team will review your report and will be in touch.

Not quite what you were looking for? Search by keyword to find the right resource:

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Portland Press Opt2Pay

Biological membranes

Biological membranes allow life as we know it to exist. They form cells and enable separation between the inside and outside of an organism, controlling by means of their selective permeability which substances enter and leave. By allowing gradients of ions to be created across them, membranes also enable living organisms to generate energy. In addition, they control the flow of messages between cells by sending, receiving and processing information in the form of chemical and electrical signals. This essay summarizes the structure and function of membranes and the proteins within them, and describes their role in trafficking and transport, and their involvement in health and disease. Techniques for studying membranes are also discussed.

Structure and organization of membranes

Membranes are composed of lipids, proteins and sugars.

Biological membranes consist of a double sheet (known as a bilayer) of lipid molecules. This structure is generally referred to as the phospholipid bilayer. In addition to the various types of lipids that occur in biological membranes, membrane proteins and sugars are also key components of the structure. Membrane proteins play a vital role in biological membranes, as they help to maintain the structural integrity, organization and flow of material through membranes. Sugars are found on one side of the bilayer only, and are attached by covalent bonds to some lipids and proteins.

Three types of lipid are found in biological membranes, namely phospholipids, glycolipids and sterols. Phospholipids consist of two fatty acid chains linked to glycerol and a phosphate group. Phospholipids containing glycerol are referred to as glycerophospholipids. An example of a glycerophospholipid that is commonly found in biological membranes is phosphatidylcholine (PC) ( Figure 1 a), which has a choline molecule attached to the phosphate group. Serine and ethanolamine can replace the choline in this position, and these lipids are called phosphatidylserine (PS) and phosphatidylethanolamine (PE), respectively. Phospholipids can also be sphingophospholipids (based on sphingosine), such as sphingomyelin. Glycolipids can contain either glycerol or sphingosine, and always have a sugar such as glucose in place of the phosphate head found in phospholipids ( Figure 1 b). Sterols are absent from most bacterial membranes, but are an important component of animal (typically cholesterol) and plant (mainly stigmasterol) membranes. Cholesterol has a quite different structure to that of the phospholipids and glycolipids. It consists of a hydroxyl group (which is the hydrophilic ‘head’ region), a four-ring steroid structure and a short hydrocarbon side chain ( Figure 1 c).

( a ) Phosphatidylcholine, a glycerophospholipid. ( b ) Glycolipid. ( c ) A sterol.

The sugars attached to lipids and proteins can act as markers due to the structural diversity of sugar chains. For example, antigens composed of sugar chains on the surface of red blood cells determine an individual's blood group. These antigens are recognized by antibodies to cause an immune response, which is why matching blood groups must be used in blood transfusions. Other carbohydrate markers are present in disease (e.g. specific carbohydrates on the surface of cancer cells), and can be used by doctors and researchers to diagnose and treat various conditions.

Amphipathic lipids form bilayers

All membrane lipids are amphipathic—that is, they contain both a hydrophilic (water-loving) region and a hydrophobic (water-hating) region. Thus the most favourable environment for the hydrophilic head is an aqueous one, whereas the hydrophobic tail is more stable in a lipid environment. The amphipathic nature of membrane lipids means that they naturally form bilayers in which the hydrophilic heads point outward towards the aqueous environment and the hydrophobic tails point inward towards each other ( Figure 2 a). When placed in water, membrane lipids will spontaneously form liposomes, which are spheres formed of a bilayer with water inside and outside, resembling a tiny cell ( Figure 2 b). This is the most favourable configuration for these lipids, as it means that all of the hydrophilic heads are in contact with water and all of the hydrophobic tails are in a lipid environment.

Spontaneous formation of bilayers by membrane lipids. The hydrophilic heads (pink circles) will always face the aqueous environment in bilayers ( a ) and liposomes ( b ). The hydrophobic tails will face inward away from the water.

Early experiments by E. Gorter and F. Grendel in 1925 were the first to demonstrate that biological membranes are bilayers. These researchers extracted the lipids from red blood cells and found that they occupied a space that was twice the surface area of the cell. Red blood cells contain no internal membranes, so they deduced that the plasma membrane must be composed of two layers of lipids.

Biological membranes and the fluid mosaic model

The fluid mosaic model proposed by Jonathan Singer and Garth Nicolson in 1972 describes the dynamic and fluid nature of biological membranes. Lipids and proteins can diffuse laterally through the membrane. Phospholipids can diffuse relatively quickly in the leaflet of the bilayer in which they are located. A phospholipid can travel around the perimeter of a red blood cell in around 12 s, or move the length of a bacterial cell within 1 s. Phospholipids can also spin around on their head-to-tail axis, and their lipid tails are very flexible. These different types of movements create a dynamic, fluid membrane which surrounds cells and organelles. Membrane proteins can also move laterally in the bilayer, but their rates of movement vary and are generally slower than those of lipids. In some cases, membrane proteins are held in particular areas of the membrane in order to polarize the cell and enable different ends of the cell to have different functions. One example of this is the attachment of a glycosyl-phosphatidylinositol (GPI) anchor to proteins to target them to the apical membrane of epithelial cells and exclude them from the basolateral membrane.

Fluorescence photobleaching is one experimental method that is used by scientists to demonstrate visually the motility of proteins and lipids in a bilayer ( Figure 3 ). A lipid or membrane protein located on the surface of a cell is tagged with a fluorescent marker such as green fluorescent protein (GFP). A beam of laser light is then focused on to a small area of the cell surface using a fluorescence microscope in order to bleach the fluorescent tags in this area so that they no longer emit a fluorescence signal. This small area of membrane is observed over time and gradually the fluorescence increases again, indicating that other tagged proteins or lipids are diffusing into this region from elsewhere in the membrane. This demonstrates that the lipid bilayer surrounding cells is fluid in nature and allows lateral diffusion of both lipids and membrane proteins.

Cells expressing a GFP-labelled protein in the endoplasmic reticulum were subjected to photobleaching. ( a ) A cell before bleaching. ( b ) The same cell immediately after bleaching of the square section shown. ( c ) The same cell 5 min after photobleaching. Adapted from Figure 1b from Lippincott-Schwartz, J., Snapp, E. and Kenworthy, A. (2001) Studying protein dynamics in living cells. Nat. Rev. Mol. Cell. Biol. 2 , 444–456.

Despite all this movement of lipids and proteins in the bilayer, vertical movement, or ‘flip-flop’, of lipids and proteins from one leaflet to another occurs at an extremely low rate. This is due to the energetic barrier encountered when forcing the hydrophilic head (in the case of lipids) or hydrophilic regions (in the case of proteins) through the hydrophobic environment of the inside of the membrane. This near absence of vertical movement allows the inner and outer leaflets of the bilayer to maintain different lipid compositions, and enables membrane proteins to be inserted in the correct orientation for them to function. However, some enzymes facilitate the process of lipid flip-flop from one leaflet to another. These flippases, or phospholipid translocators, use ATP to move lipids across the bilayer to the other leaflet. In eukaryotic cells, flippases are located in various organelles, including the endoplasmic reticulum (ER), where they flip-flop newly synthesized lipids.

How membranes are made

Biological membranes are formed by adding to a pre-existing membrane. In prokaryotes this occurs on the inner leaflet of the plasma membrane, facing the cytoplasm. In eukaryotes, membrane synthesis takes place at the ER on the cytoplasmic leaflet of the ER membrane (termed the ‘inside’ of the cell). Lipids then leave the ER and travel through the secretory pathway for distribution to various subcellular compartments or the plasma membrane.

In eukaryotic cells, enzymes that span the ER catalyse the formation of membrane lipids. In the cytoplasmic leaflet of the ER membrane, two fatty acids are bound, one by one, to glycerol phosphate from the cytoplasm. This newly formed diacylglycerol phosphate is anchored in the ER membrane by its fatty acid chains. The phosphate is then replaced by the head group (e.g. phosphate and choline). Flippases in the ER membrane can then move some of these newly formed lipids to the luminal side of the ER membrane. Similarly, flippases in prokaryotes can transfer new lipids from the inner leaflet of the plasma membrane to the outer leaflet. These flippases are responsible for adjusting the lipid composition of each layer of the membrane. In eukaryotes, lipids must then be distributed to the various intracellular membranes. The traffic of vesicles between organelles in combination with signals that direct particular lipids to specific locations is required to create the correct lipid composition in all of the cellular membranes ( Figure 4 ). Vesicles bud from the ER and travel via the ER–Golgi intermediate compartment (ERGIC) to join with the Golgi, where sorting of lipids takes place. The Golgi then sends lipids in vesicles to various destinations, including the plasma membrane and lysosomes. Lipids and proteins are internalized from the plasma membrane into endosomes. Organelles, such as mitochondria, acquire lipids from the ER by a different mechanism. Water-soluble proteins called phospholipid-exchange proteins remove phospholipids from the ER membrane and deposit them in the membranes of the appropriate organelles.

The main compartments of eukaryotic cells are shown. Arrows indicate movement of lipid vesicles between them, with colours at the tail end indicating origin and those at the head end indicating destination.

Distribution of lipids

The inner and outer leaflets of bilayers differ in their lipid composition. In mammalian cells, the outer leaflet of the plasma membrane contains predominantly PC and sphingomyelin, whereas PS and PE are found on the inner leaflet. During programmed cell death (apoptosis), PS is no longer restricted to the inner leaflet of the plasma membrane. It is exposed on the outer leaflet by the action of an enzyme called scramblase which is a type of flippase enzyme. PS is negatively charged, unlike PC, which has no net charge. The movement of PS into the outer leaflet therefore changes the charge of the plasma membrane as viewed from the outside of the cell. This change in surface charge labels the apoptotic cell for phagocytosis by phagocytic cells such as macrophages.

Lipid composition also varies between the organelles within eukaryotic cells. Cholesterol is synthesized in the ER, but the ER membrane has a relatively low cholesterol content, as much of the cholesterol is transported to other cellular membranes. The prevalence of cholesterol in membranes increases through the secretory pathway, with more in the Golgi than in the ER (the trans -Golgi network is richer in cholesterol than the cis -Golgi), and most in the plasma membrane. This increase in cholesterol through the secretory pathway results in slightly thicker membranes in the late Golgi and plasma membrane compared with the ER, and is thought to be a contributing factor to protein sorting through the pathway, as membrane proteins in the plasma membrane generally have longer hydrophobic transmembrane domains than membrane proteins that reside in the ER.

Membrane proteins

Membrane proteins are the nanomachines that enable membranes to send and receive messages and to transport molecules into and out of cells and compartments. Without membrane proteins the phospholipid membrane would present an impenetrable barrier and cells would be unable to communicate with their neighbours, transport nutrients into the cell or waste products out of it, or respond to external stimuli. Both unicellular and multicellular organisms need membrane proteins in order to live. The membrane proteins that are present in a particular membrane determine the substances to which it will be permeable and what signal molecules it can recognize.

Synthesis of membrane proteins

In eukaryotic cells, the synthesis of membrane proteins destined for the plasma membrane, ER or any other membrane-bound compartment begins on cytosolic ribosomes. After a short segment of protein has been synthesized, the ribosome, mRNA and nascent protein chain associate with the ER, where the rest of the protein is made and simultaneously inserted into the membrane. This phenomenon was first explained by Günter Blobel, David Sabatini and Bernhard Dobberstein in the 1970s. These scientists proposed that there is a ‘binding factor’ which recognizes the emerging protein chain and can dock the ribosome at the ER membrane. We now know that there is an N-terminal signal sequence within membrane proteins. These signal sequences are not identical but share a common motif, namely a hydrophobic stretch of 20–30 amino acids, a basic region at the N-terminus and a polar domain at the C-terminus of the signal. These N-terminal signal sequences are recognized by the signal recognition particle (SRP), which has binding sites for the signal sequence, ribosome and the SRP receptor which is embedded in the ER membrane. Upon binding the SRP, the ribosome pauses protein synthesis. The SRP binds to the SRP receptor, adjacent to a translocon pore in the ER membrane. The translocon is a protein pore through which membrane protein chains can be threaded into the membrane. It has a laterally opening gate to allow newly synthesized proteins into the ER membrane. Once the ribosome is at the translocon, the SRP dissociates and protein synthesis resumes. This process is referred to as co-translational targeting, and the main events are summarized in Figure 5 .

The key steps of ER targeting are summarized. Each component is labelled and the ER membrane is represented by double blue lines. The signal sequence (shown in black) becomes the first transmembrane domain of the protein in this example.

Co-translational targeting is the dominant mechanism for protein delivery to the ER in higher eukaryotes, whereas yeast and prokaryotes favour post-translational targeting, whereby proteins are delivered to the ER after completion of synthesis. Post-translational targeting also occurs in higher eukaryotes, often when a membrane protein is so small that the signal sequence does not emerge until the whole protein has been synthesized. Post-translational targeting can be carried out both by SRP-dependent and by SRP-independent mechanisms.

Structure and function of membrane proteins

Membrane-spanning proteins are diverse in structure and function. They can be constructed of α-helices or from β-barrels. The β-barrel membrane proteins often function as pores, with hydrophobic amino acids facing out into the bilayer. In addition, there are other non-spanning proteins which associate with the bilayer, often using a hydrophobic anchor. Here we shall focus on the α-helical membrane proteins. These proteins have at least one α-helical hydrophobic stretch of amino acids, around 20 residues in length, which corresponds to around 30 Å (the thickness of an average phospholipid bilayer). If an α-helical membrane protein spans the membrane more than once, it will have more than one of these hydrophobic sections. For example, the Ca 2+ -ATPase of the ER and sarcoplasmic reticulum (SR) spans the membrane 10 times, so it has 10 hydrophobic stretches of around 20 amino acids each.

Membrane proteins control what enters and leaves the cell

A vital class of membrane proteins are those involved in active or passive transport of materials across the cell membrane or other subcellular membranes surrounding organelles. For a cell or an organism to survive, it is crucial that the right substances enter cells (e.g. nutrients) and the right substances are transported out of them (e.g. toxins).

Passive and active transport

Molecules can cross biological membranes in several different ways depending on their concentration on either side of the membrane, their size and their charge. Some molecules, including water, can simply diffuse through the membrane without assistance. However, large molecules or charged molecules cannot cross membranes by simple diffusion. Charged molecules such as ions can move through channels passively, down electrochemical gradients. This movement is described as ‘downhill’, as the ions or molecules travel from an area of high concentration to an area of low concentration. This requires channel proteins but no energy input. Passive transport can also be mediated by carrier proteins that carry specific molecules such as amino acids down concentration gradients, again without any requirement for energy. Active transport moves species against concentration gradients and requires energy, which is obtained from ATP, from light, or from the downhill movement of a second type of molecule or ion within the same transporter ( Figure 6 ).

The different types of membrane proteins involved in passive and active transport are shown.

Passive transport

Passive transport is the movement of molecules across biological membranes down concentration gradients. This type of transport does not require energy. Channels form water-filled pores and thus create a hydrophilic path that enables ions to travel through the hydrophobic membrane. These channels allow downhill movement of ions, down an electrochemical gradient. Both the size and charge of the channel pore determine its selectivity. Different channels have pores of different diameters to allow the selection of ions on the basis of size. The amino acids that line the pore will be hydrophilic, and their charge will determine whether positive or negative ions travel through it. For example, Ca 2+ is positively charged, so the amino acids lining the pores of Ca 2+ channels are generally basic (i.e. they carry a negative charge).

Channels are not always open. They can be gated by ligands which bind to some part of the protein, either by a change in membrane potential (voltage gated) or by mechanical stress (mechanosensitive). The nicotinic acetylcholine receptor is an example of a ligand-gated ion channel which opens upon binding the neurotransmitter acetylcholine ( Figure 7 ). The nicotinic acetylcholine receptor is a pentameric membrane protein composed of five subunits arranged in a ring, with a pore through the centre. In the closed state, the pore is blocked by large hydrophobic amino acid side chains which rotate out of the way upon acetylcholine binding to make way for smaller hydrophilic side chains, allowing the passage of ions through the pore. Opening of the nicotinic acetylcholine receptor allows rapid movement of Na + ions into the cell and slower movement of K + ions out of the cell, in both cases down the electrochemical gradient of the ion. The difference in gradients between Na + and K + across the membrane means that more Na + enters the cell than K + leaves it. This creates a net movement of positive charges into the cell, resulting in a change in membrane potential. Acetylcholine released by motor neurons at the neuromuscular junction travels across the synapse and binds to nicotinic acetylcholine receptors in the plasma membrane of the muscle cells, causing membrane depolarization. This depolarization of the muscle cells triggers Ca 2+ release and muscle contraction.

The pentameric structure of the receptor is shown, with the pore region (P) indicated. Transmembrane helices (M1–M4) are labelled in each subunit. The bilayer is shown in orange. Reproduced from Berridge, M.J. (2012) Cell Signalling Biology; doi:10.1042/csb0001003, with permission.

Carrier proteins are the other class of membrane proteins, apart from channels, which can facilitate passive transport of substances down concentration gradients. Carrier proteins transport molecules much more slowly than channels, as a number of conformational changes in the carrier are required for the transport of the solute across the membrane. A molecule such as a sugar binds to the carrier protein on one side of the membrane where it is present at a high concentration. Upon binding, the carrier changes conformation so that the sugar molecule then faces towards the opposite side of the membrane. The concentration of sugar on this side is lower, so dissociation occurs and the sugar is released. This is downhill movement and, although slower than movement through channels, it requires no energy.

The cystic fibrosis transmembrane conductance regulator (CFTR) is an ATP-dependent chloride ion (Cl − ) channel that has an important role in regulating the viscosity of mucus on the outside of epithelial cells. ATP is used to gate the channel, but the movement of Cl − occurs down its electrochemical gradient, so does not require energy. A heritable change in the CFTR gene which results in a single amino acid deletion in the protein causes cystic fibrosis. This is a serious illness in which thick mucus accumulates in the lungs, causing a significantly lower than average life expectancy in patients who have the disease. Unimpaired ion transport is vital for our survival and health, and conditions such as cystic fibrosis highlight the need for research into these types of proteins.

Active transport

The transport of molecules across a membrane against a concentration gradient requires energy, and is referred to as active transport. This energy can be obtained from ATP hydrolysis (primary active transport), from light (as, for example, in the case of the bacterial proton pump bacteriorhodopsin), or from an electrochemical gradient of an ion such as Na + or H + (secondary active transport).

Calcium ions signal many events, including muscle contraction, neurotransmitter release and cellular motility. However, high cytoplasmic concentrations of Ca 2+ are toxic to the cell. Therefore Ca 2+ must be tightly regulated and removed from the cytoplasm either into internal stores (the ER, and the SR in muscle cells) or into the extracellular space. This Ca 2+ removal is carried out by a family of Ca 2+ -ATPases, including the sarco/endoplasmic reticulum Ca 2+ -ATPase (SERCA), which hydrolyse ATP to move Ca 2+ against its electrochemical gradient into the ER and SR ( Figure 8 ). There are Ca 2+ -ATPases in the ER, Golgi and plasma membrane, and despite their sequence similarity, these proteins are differentially targeted to the appropriate membrane. These Ca 2+ pumps are primary active transporters. SERCA moves two Ca 2+ ions into the ER or SR for every ATP molecule that is hydrolysed. The pump undergoes a cycle of binding ATP and phosphorylation, and undergoes large conformational changes every time it transports a pair of Ca 2+ ions. SERCA is a P-type ATPase (so called because it is phosphorylated during ion transport). There are many P-type ATPases, and they are conserved in evolution across many species. The Na + /K + -ATPase is one of these P-type ATPases, and it works in a similar way to SERCA to pump Na + out of the cell and K + into the cell using energy derived from the hydrolysis of ATP. We have now obtained three-dimensional structures of SERCA in a number of conformational states, which allow scientists to visualize the transport process.

The crystal structure of SERCA in the ADP- and Ca 2+ -bound state is shown. D351 (in red) is the residue phosphorylated during the movement of Ca 2+ ions into the ER or SR. The three cytoplasmic domains, phosphorylation (P), nucleotide binding (N) and actuator (A) are labelled. ADP is shown in yellow and Ca 2+ ions in green. Protein Data Bank (PDB) code 2ZBD, rendered using PDB Protein Workshop.

Secondary active transport requires an ion electrochemical gradient to drive the uphill transport of another solute. The downhill movement of one species drives the uphill movement of the other. This can be symport (in which both types of molecule or ion travel across the membrane in the same direction) or antiport (in which the two species travel in opposite directions), as shown in Figure 9 .

The two types of co-transport are shown, with examples.

In order to transport glucose into cells, the Na + –glucose symporter uses the electrochemical gradient of Na + across the plasma membrane. The concentration of Na + is much higher outside the cell, and the inside of the cell is negatively charged relative to the outside, so by allowing Na + to travel down its electrochemical gradient, these transporters can move glucose uphill, into the cell and against its concentration gradient. This is referred to as symport, as both Na + and glucose travel in the same direction—in this case into the cell. In order for this symport to be sustainable, the Na + gradient must be maintained. This is done by the Na + /K + -ATPase, which uses ATP to pump the Na + back into the extracellular space, thus maintaining a low intracellular Na + concentration.

Both Na + and Ca 2+ are present at much higher concentrations outside the cell than inside it. Like the Na + –glucose symporter, the Na + –Ca 2+ exchanger uses the electrochemical gradient of Na + across the plasma membrane to move a second species (Ca 2+ ) against its electrochemical gradient. However, in this case the transporter is an antiporter, as it uses the concentration gradient of one substance moving in (Na + ) to move another (Ca 2+ ) out of the cell. This antiporter has an exchange rate of three Na + ions in to two Ca 2+ ions out. It moves Ca 2+ out of the cell faster than the plasma membrane equivalents of SERCA, but has a lower affinity for Ca 2+ than these P-type ATPases. Again this transporter relies on the Na + /K + -ATPase to maintain the low intracellular Na + concentration.

Solving the structure of membrane proteins

In order to understand more fully the mechanisms of action of membrane proteins such as the transporters described here, we can determine their three-dimensional protein structures. As a result of huge advances in structural biology in the last 50 years, we now have access to many thousands of protein structures in online databases. This enables researchers to visualize the structure of their protein of interest, and thus gain insight into its mechanism.

X-ray crystallography

The structure of whale myoglobin was solved in 1958 using X-ray crystallography, earning John C. Kendrew and Max Perutz the Nobel Prize in Chemistry. This was the first protein structure to be solved using this technique, and since then thousands of proteins have been solved using this method. X-ray crystallography works by firing a beam of X-rays at a crystalline structure and measuring the diffraction of the X-rays after they have passed through the structure of interest. This generates an electron density map, showing where different atoms in the structure are located. For regular crystalline solids such as salts this is relatively straightforward, but for large irregular molecules such as proteins it can present many technical challenges. Before a protein is subjected to X-ray beams, it must first be purified and crystallized. In nature, proteins exist in the busy milieu of a cell, surrounded by thousands of other types of proteins, as well as lipids and other molecules. A common method of obtaining enough of the protein of interest involves expressing the relevant gene in a system such as bacteria. The gene is tagged with a small protein tag which can be used to isolate the protein of interest. Bacterial systems allow large amounts of protein to be produced cheaply and quickly. However, if the protein of interest is from a species that is only distantly related to that in which it is normally expressed (e.g. a human protein produced in Escherichia coli ( E. coli )), the lack of correct glycosylation enzymes and the differences in protein folding and assembly may prevent the production of a biologically active protein. In addition, the expression of membrane proteins that make pores or channels can kill the host organism.

A pure protein sample is then crystallized by allowing water to evaporate away, in exactly the same way as a solution of salt will form crystals naturally when left to dry. Optimum conditions for this must be determined, and crystallization conditions are not always straightforward, as they differ from one protein to another. For soluble proteins such as myoglobin this is easier than for insoluble membrane proteins. Membrane proteins have lipid-soluble domains that will not dissolve in an aqueous medium. This significantly decreases the ease with which membrane protein structures can be solved using X-ray diffraction. However, there are ways in which scientists can overcome this difficulty. Generally, membrane proteins are removed from the membrane in which they were made and placed in an environment of lipids and detergents for crystallization. Sometimes the lipids associated with the protein are apparent in the crystal structure.

The number of solved crystal structures of proteins is constantly growing as technology improves and expertise is shared among scientists to help to optimize conditions for crystal production. The Protein Data Bank (PDB) is an online archive of protein structures which can be freely accessed by scientists worldwide. At the time of writing, 88% of the structures in the PDB have been solved by X-ray crystallography, and there are currently just under 70 000 X-ray crystal structures in the database. The number of membrane protein structures in the PDB is increasing rapidly with the refinement of crystallization techniques ( Figure 10 ).

The increase in the number of solved crystal structures of membrane proteins is shown from 1985, when the first such structure was solved. Adapted from White, S.H. (2009) Biophysical dissection of membrane proteins. Nature 459 , 344–346.

Other structural techniques

Nuclear magnetic resonance (NMR) spectroscopy is another valuable technique for elucidating membrane protein structure. Molecules are placed in a magnetic field and the resonance properties of different atomic nuclei are measured, which gives an indication of where in a particular molecule different atoms are located. Generally, NMR is limited to smaller proteins, below around 35 kDa in size. It also offers the potential to visualize proteins in a more physiologically relevant environment (e.g. in lipid bilayers or micelles). Another advantage of NMR is that it does not require the protein to be locked in a crystal lattice—a structure which can distort the natural shape of the protein.

Electron microscopy can also be used to study membrane protein structure. By freezing membrane proteins in their natural lipid environments, it is possible to investigate their structure using high-resolution electron microscopy. This provides a snapshot of the naturally occurring conformation of individual proteins in the bilayer.

Interactions between lipids and proteins in biological membranes

The lipids that surround membrane proteins in biological membranes play an important role in the activity of these proteins. As was mentioned earlier, some membrane protein crystal structures include lipids bound to the outside surface of the transmembrane domains of the proteins. It is thought that these lipids bind tightly to the protein, and have a long-lived interaction with the transmembrane region. In other cases, lipids are thought to interact briefly with membrane proteins, rapidly moving away and being replaced by other membrane lipids. The activity of membrane proteins is considered to be dependent to some extent on the lipids that surround them in the membrane. Certain types of K + channel are thought to bind to negatively charged membrane lipids, as the activity of these channels increases at higher anionic lipid concentrations. These types of interaction can be studied by placing a purified form of the protein of interest in an artificial bilayer and measuring its activity. By altering the types of lipid present in the artificial bilayer, deductions can be made about the lipids that the protein requires in order to be active. Fluorescence spectroscopy and electron spin resonance are two techniques that are used to measure how strongly membrane proteins interact with specific lipids around them.

Molecular dynamics simulations use computer algorithms to work through theoretical problems. These simulated experiments are useful for investigating interactions between membrane proteins and lipids, as in real membranes these interactions are often so fleeting that they are very difficult to measure. Molecular dynamics simulations have predicted that in the case of the nicotinic acetylcholine receptor, the negatively charged lipid, phosphatidic acid, is required for activity. These simulations have also shown that cholesterol stabilizes the receptor and that the phosphatidic acid forms a shell around the protein which is more long-lasting than the interactions with other membrane lipids. Although molecular dynamics simulations are extremely useful, they are limited by the assumptions and approximations on which they are based. As in many areas of biology, a combination of experimental and computational research is required if real progress is to be made in understanding the complexity of biological membranes.

Internal membranes in eukaryotic cells form organelles

Inside the plasma membrane that surrounds eukaryotic cells lie many other membranes which define the intracellular compartments, or organelles. Each of these organelles has distinct functions and contains specific complements of proteins adapted for these roles. With the exception of a few proteins that are coded for by the mitochondrial genome, synthesis of all of the proteins that are required in these organelles begins on ribosomes in the cytoplasm, and therefore the proteins must be directed to the correct destination. We have seen earlier how this is achieved with membrane proteins, and most organelles have some kind of signal sequence that can be recognized by various receptors and which ensures that the protein arrives at the correct organelle.

Organelles have distinct lipid compositions

Besides the specific protein complement of each organelle, the lipid make-up of the bilayers surrounding organelles varies. Lipids are synthesized in the ER, and flippases move lipid molecules between leaflets of the bilayer. For organelles in the secretory pathway and the plasma membrane, lipid transport into these compartments is mediated by vesicular membrane traffic through the pathway. The cholesterol concentration in membranes increases from the ER through the Golgi to the plasma membrane. Cholesterol makes membranes thicker and more rigid, so the low levels of cholesterol in the ER membrane render it thin and facilitate the insertion of newly synthesized membrane and secretory proteins. PC becomes relatively less abundant through this pathway, with more found in the ER than at the plasma membrane. PS and PE are found throughout the secretory pathway in the cytosolic leaflet of the membranes. This differential lipid composition through the secretory pathway is achieved by targeting specific lipids into transport vesicles. Proteins included in these vesicles act as labels and direct the lipids to the right compartment. Forward-moving (anterograde) vesicles destined for the plasma membrane are rich in cholesterol. Lipids also move backwards through the secretory pathway, from the plasma membrane towards the ER. This is known as retrograde traffic. Retrograde vesicles from the Golgi are enriched in lipids such as PC, which are concentrated in the ER.

The lipid composition of the mitochondria is very different from that of the secretory pathway compartments. Mitochondrial membranes are much richer in PE and cardiolipin than is the ER. Cardiolipin is synthesized in the mitochondria and is predominantly confined to this organelle. As membrane proteins have evolved along with their organelles and surrounding lipids, it follows that different lipid compositions are required in different organelles for the optimum activity of the proteins within their membranes. The structure of the ADP/ATP carrier in mitochondria has been solved and was found to include cardiolipin and PC molecules bound to the protein. The activity of this carrier protein is dependent on the presence of cardiolipin, which is relatively abundant in mitochondrial membranes.

Proteins must be targeted to the correct organelle for cells to function

The targeting of newly synthesized membrane and secretory proteins to the ER has already been briefly discussed. However, there are many different destinations within the cell to which a protein can be sent, and sometimes proteins are located in more than one of these. The signals and protein machinery that are required to target proteins to the correct compartment are many and various, and much of the detail of the exact mechanisms involved has yet to be clarified.

Vesicular transport

Traffic through the secretory pathway is by vesicular transport in both anterograde and retrograde directions. Proteins and lipids can be included and excluded from vesicles by various means in order to selectively determine which molecules move forward or backward through the pathway. Vesicles are coated with proteins that determine their destination. Generally these coat proteins (COPs) are directional—COPII coats anterograde vesicles, and COPI coats retrograde vesicles. Proteins that travel in vesicles (referred to as cargo) are selected either by interacting with receptors in the vesicles or by directly interacting with the coat proteins. The selection of cargo occurs at the budding stage, when the coat proteins begin to distort the donor membrane (e.g. the ER) into a vesicle. Once the cargo has been selected and the coat proteins have been assembled, the vesicle buds off and travels to the acceptor membrane (e.g. the Golgi in the case of COPII vesicles from the ER) either by diffusion or with the help of motor proteins that ‘walk’ the vesicle along the cytoskeleton. The vesicle then fuses with the acceptor membrane, depositing its cargo and constituent lipids ( Figure 11 ).

The main events in vesicular transport of cargo are shown. Cargo is selected and packed into vesicles which are formed by coat proteins (1 and 2). The GTPase Rab is also incorporated on the outside of the vesicle and facilitates the steps illustrated. The vesicle then travels along proteins of the cytoskeleton towards its destination following dissociation of coat proteins (3). The vesicle is tethered to the donor membrane with the help of tethering proteins and the SNARE complexes (4), allowing membrane fusion and release of the cargo (5).

The formation and fusing of vesicles are energetically demanding, as these processes require the stable bilayer to be broken in order to pinch off a new vesicle, and then fused with a different membrane. Both energy and specialized protein machinery are required to overcome this energy barrier.

Once the vesicle has budded, travelled through the cytosol and reached its destination, it must fuse with its receptor membrane. Again this is an energetically unfavourable process, and protein machinery has to be utilized in order to allow the fusion of two bilayers. SNARE proteins are central to the process of vesicle fusion. Vesicles carry v-SNAREs that bind to specific t-SNAREs on the target membranes. Not only does this confer specificity in the targeting of vesicles, but also the SNAREs facilitate membrane fusion on arrival of the vesicle. The interacting v-SNAREs and t-SNAREs form a four-helix bundle at the interface between the two membranes, consisting of three helices from the t-SNARE and one helix from the v-SNARE. This stable interaction is thought to provide the free energy necessary to enable the two membranes to become very close and fuse. As the two bilayers become closer, the lipids in the two outer leaflets can come into contact with one another, thereby increasing the hydrophobic nature of the site and enabling the membranes to join, and overcoming the energy barrier. The transmembrane domains of the SNAREs are also believed to be involved in membrane fusion, as when they are replaced by lipids experimentally, fusion does not occur. Upon fusion, the cargo enters the target compartment, and the lipids and membrane proteins that formed the vesicle diffuse into the target membrane.

Determining the mechanisms of membrane budding is important for understanding how viruses such as the human immunodeficiency virus (HIV) produce new viral particles. Unlike the budding in the secretory pathway described earlier, HIV particles bud away from the cytoplasm, into the extracellular space. This viral budding occurs in the same orientation as the budding that occurs within endosomes. The proteins which enable this budding are referred to as endosomal sorting complexes required for transport (ESCRTs). HIV ‘hijacks’ the ESCRT machinery to enable it to bud from the plasma membrane, out of the cytoplasm and into the extracellular space. Interactions between HIV proteins and ESCRT proteins recruit the host cell ESCRT machinery to the budding vesicle, allowing membrane scission and vesicle release. Other viruses can bud without assistance from the ESCRTs, and it is thought that HIV may also be able to bud in an ESCRT-independent manner. Understanding more about these membrane budding and scission events is crucial to elucidating how viruses proliferate and how we can inhibit processes by means of drug interventions.

Protein trafficking in the secretory pathway

As described earlier, a hydrophobic stretch of 20–30 amino acids with a basic N-terminus and a polar region at the C-terminus emerging from the ribosome causes the protein to be targeted to the ER, where synthesis is completed. This hydrophobic stretch can be cleaved in the case of soluble proteins, or it can remain attached. An uncleaved signal sequence is referred to as a signal anchor sequence, as it both signals ER targeting and then goes on to anchor a protein in the membrane, and becomes a transmembrane domain in the fully folded protein. The SRP-dependent targeting step is common to ER proteins as well as proteins destined for the Golgi or the plasma membrane, or to be secreted from the cell.

ER exit is thought to allow some selection of which proteins remain in the ER and which proteins leave and move in vesicles towards the Golgi. ER exit sites are located in areas of the ER close to the Golgi, and are rich in COPII coat proteins. It is not understood exactly which properties of a protein determine whether it will leave the ER in COPII vesicles, but it is currently thought that the transmembrane domain length is an important factor. Longer transmembrane domains appear to predispose proteins to exit the ER and travel towards the Golgi. This is consistent with the fact that membrane thickness increases through the secretory pathway due to increased cholesterol content, as described earlier.

Upon arrival at the cis -Golgi, proteins can then be retrieved to the ER, remain in the Golgi, or travel onward to the plasma membrane. Retrieval to the ER is not fully understood, but some proteins contain retrieval motifs, such as the KDEL four-amino-acid motif which is recognized by a receptor and enables packaging of the protein into retrograde COPI vesicles. Other proteins appear to cycle between the ER and the Golgi without known retrieval motifs. Proteins move in both anterograde and retrograde directions through the Golgi stack. They can then leave the trans -Golgi and move to the plasma membrane in vesicles.

Proteins at the plasma membrane can move into the cell in vesicles by endocytosis (e.g. when surface receptors are internalized for degradation in lysosomes). Endocytic vesicles are often clathrin coated. Clathrin, like the COPs, distorts the membrane into curved structures, allowing vesicle formation. Clathrin forms a cage-like shape that promotes vesicle formation and scission by virtue of the rigid shape of the protein complexes which form at the membrane. Not all endocytosis is clathrin dependent, and there are other proteins, such as caveolin, which can facilitate the formation of endocytic vesicles.

Mitochondrial and nuclear protein targeting

Newly synthesized proteins destined for the mitochondria or the nucleus are targeted in a different way, independently of the secretory pathway. Some mitochondrial proteins are encoded by the mitochondrial genome, while others are encoded by the nuclear genome. Mitochondria have a double-layered membrane. Therefore targeting signals for mitochondrial proteins need to contain information not only to direct the protein to the organelle, but also to determine in which membrane it will be located (in the case of membrane proteins), or whether it will be located inside the mitochondria (the matrix) or in the intermembrane space between the inner and outer membranes (in the case of soluble proteins). Mitochondrial targeting motifs vary enormously, but generally are located at the N-terminus of the protein and are rich in positively charged and hydrophobic amino acids. Proteins destined for the nucleus are targeted by nuclear localization sequences that direct proteins which have been synthesized in the cytoplasm through nuclear pore complexes. Again these sequences are not very highly conserved, but generally contain clusters of positively charged amino acids.

Sending messages across membranes

We have already seen how ion channels and other transport proteins can allow substances to cross the lipid bilayer. Knowledge of how fat-soluble and water-soluble substances cross membranes is important for gaining an understanding of how messages cross membranes and thus how one cell can communicate with another. Cells receive and send messages constantly (e.g. in order to respond to hormone signals, conduct action potentials, and sense external stimuli such as taste and smell).

Messengers: lipid soluble or water soluble?

Substances that send messages are known as messengers, and they vary enormously in their chemical composition, size and hydrophobicity. In order to understand how a cell receives a message, it is important to ascertain first whether the messenger is lipid or water soluble. Hormones are one example of messengers that are released by cells. The human body contains both lipid-soluble and water-soluble hormones. Lipid-soluble hormones are generally transported through the blood, bound to carrier proteins. Steroid hormones such as the androgens and oestrogens are lipid soluble by virtue of their ringed molecular structures, which are derived from cholesterol. This allows these hormones to diffuse freely through the plasma membrane of cells and bind to their receptors, which are located inside cells. In the case of oestrogen, the receptor is located in the cytoplasm and upon ligand binding relocates to the nucleus, where it binds DNA and acts as a transcription factor, altering gene expression. The receptor contains a nuclear localization sequence which is hidden until oestrogen binds, allowing it to be targeted to the nucleus.

Other hormones, such as insulin and adrenaline, are water soluble and therefore cannot pass freely through the membranes of cells. Their receptors are located on the outside of the plasma membrane in order for them to be able to convey a message without entering cells. Insulin binds to the membrane-spanning insulin receptor on the surface of target cells, and initiates a signal cascade that results in an increase in the number of glucose transporters at the cell membrane, and a subsequent increase in glucose uptake.

G proteins and second messengers

Many cell-surface receptors share structural features, including seven membrane-spanning helices. These 7TM receptors bind their ligand (the messenger molecule) on the extracellular side of the membrane, and bind a GTP-binding protein (G protein) on the intracellular side. Due to this interaction with G proteins, these receptors are called G-protein-coupled receptors (GPCRs). When the ligand binds the GPCR, the receptor undergoes conformational changes that are transferred through the membrane-spanning region to the bound G protein. This change in structure allows the G protein to exchange a bound GDP molecule for a GTP molecule, and thereby switch from an inactive state to an active state. G proteins consist of three subunits—α, β and γ. An inactive, GDP-bound G protein consists of all three subunits, with the nucleotide bound in the α subunit. When the GPCR binds the ligand, the G protein is activated and the α subunit, now with GTP bound to it, dissociates from the complex ( Figure 12 ). This activated α subunit now has an exposed face (where the β and γ subunits were bound) and can bind proteins to propagate the signal. An example of this downstream signalling from GPCRs is the activation of adenylate cyclase by the GTP-bound α subunit in the case of the β-adrenergic receptor when it binds its ligand, adrenaline (epinephrine). The effect of this adenylate cyclase activation is an increase in cAMP production from ATP, leading to downstream effects. The dissociated βγ dimer also has downstream effects. The α subunit has GTPase activity so that it can convert the bound GTP back to GDP. The GDP-bound subunit then returns to and binds the β and γ subunits ready for another cycle of signalling.

The ligand bound to the GPCR is shown in red. Binding allows the exchange of GDP for GTP by the associated G protein, and dissociation of the protein into Gα and Gβγ subunits. These then have downstream effects on a range of proteins, thereby propagating the signal from the bound ligand. Yellow arrows indicate either activation (up arrow) or inhibition (down arrow) of the targets. Regulators of G-protein signalling (RGS) proteins aid the GTPase activity of the G protein to turn off the signal. Arrestin can bind the receptor following GPCR phosphorylation by G-protein receptor kinase (GRK), desensitizing the receptor to further signalling. Reproduced from Berridge, M.J. (2012) Cell Signalling Biology; doi:10.1042/csb0001002, with permission.

After the initial ligand interaction with the GPCR and the G-protein dissociation, the message is then carried by second messengers activated by the signal cascade. In the example that has just been given, the G protein associated with the β-adrenergic receptor activates adenylate cyclase, increasing the production of cAMP, which is a widely used second messenger. Most of the effects of cAMP are due to the activation of protein kinase A (PKA). PKA phosphorylates target enzymes to modify their activities. In the case of adrenaline, PKA activates enzymes involved in the production of glucose from glycogen stores, and inhibits enzymes involved in the production of more glycogen.

Around 25% of drugs are targeted at GPCRs, so an understanding of their structures and functions is crucial in the fight against disease. As explained earlier, membrane proteins are notoriously difficult to crystallize due to their hydrophobic nature, and GPCRs have a very small hydrophilic area. Some techniques, such as the production of an antibody–receptor complex to increase hydrophilicity, have been successful in aiding crystallization. Rhodopsin ( Figure 13 ) was crystallized in 2000, followed by the related β 2 -adrenergic receptor in 2007. Since then, several more GPCR structures have been solved, providing valuable information that can help computational biologists to work out the detailed mechanisms of GPCR signalling. Molecular dynamics simulations have been performed on the interactions between GPCRs and their partner G proteins using the crystal structures available to inform the modelling process. These studies will play an important part in helping us to understand how the helices in the GPCRs move and twist in order to convey the extracellular signal to the intracellular G protein.

A ribbon representation of the first crystal structure of rhodopsin is shown in the plane of the membrane ( a ) and from the cytoplasmic side (b). The N- and C-termini are labelled, as are the seven transmembrane helices (I-VII). Adapted from Figure 2 from Palczewski, K., Kumasaka, T., Hori, T., Behnke, C.A., Motoshima, H., Fox, B.A., Le Trong, I., Teller, D.C., Okada, T., Stenkamp, R.E. et al. (2000) Crystal structure of rhodopsin: a G-protein-coupled receptor. Science 289 , 739–745.

Nerve impulses

Nerve impulses are able to occur because biological membranes are impermeable to ions, so a membrane potential can be generated across them, with more of one charge on one side than on the other. These membrane potentials are generated and altered by ion channels. A nerve impulse, or action potential, is generated when a membrane is depolarized upon influx of positively charged ions into the cell. The resting potential in a neuron is around –70 mV, maintained by K + channels and the Na + /K + -ATPase. When an action potential is generated, voltage-dependent Na + channels open once the cell membrane has crossed the threshold potential of around –60 mV. This allows a fast influx of Na + down its electrochemical gradient, increasing the membrane potential (i.e. reducing its negative value). This influx of positive charges enables the inside of the cell to become positively charged compared with the extracellular environment, as the membrane potential exceeds 0 mV. The depolarization itself inhibits the Na + channels, so no more ions enter the cell. To restore the negative resting potential, voltage-dependent K + channels open, allowing K + ions to move out of the cell, thus making the inside of the cell more negative. An after-potential (hyperpolarization) can then occur, whereby the membrane potential decreases below –70 mV before being restored by the action of ion channels and ATPases.

Membranes in health and disease

We have seen how membranes, and the membrane proteins within them, function in healthy cells and organisms. We shall now consider what happens in disease, and how we can use our knowledge of membrane proteins to make new drugs to treat disease.

Serious disease results from non-functional ion channels

Cystic fibrosis is an autosomal recessive disease that results from mutations in the CFTR gene. This gene encodes a Cl − channel that has a vital role in regulating the viscosity of mucus on membranes such as those in the lungs. In healthy individuals, transport of Cl − ions out of the cells through CFTR is followed by water, and mucus of the right viscosity is produced. However, lack of Cl − channels results in thick, dehydrated mucus, and consequently cystic fibrosis patients have difficulty in breathing and a predisposition to chest infections. Most cases of cystic fibrosis are caused by the ΔF508 mutation, which is a deletion of a phenylalanine residue at position 508 in the protein. Like nearly all membrane proteins, CFTR is translated on ribosomes at the ER and then moves through the secretory pathway to the plasma membrane, where it carries out its transport role. The single amino acid deletion of F508 causes the protein to misfold, and instead of moving out to the plasma membrane, it is held in the ER by the protein quality control machinery. Therefore very few CFTR molecules reach the plasma membrane in people with the ΔF508 mutation, and this results in serious disease.

Diseases such as this are not easily treated. Blocking the protein quality control machinery is not an option, because it would lead to the release of other misfolded proteins, with potentially disastrous consequences. Although some current drug treatments can ameliorate the symptoms of the disease, it is hoped that gene therapy might become routine as it addresses the cause of the problem. Treatment of patients with an artificial, functional version of the gene enables them to produce a working CFTR protein that can be expressed at the plasma membrane. Although this is not a complete cure, it is a potentially effective way to greatly reduce the symptoms of cystic fibrosis in the lungs. As DNA is a large hydrophilic molecule, it cannot be simply administered like many other drugs. Delivery of gene therapy is a challenge, and this is one reason why it is difficult to treat patients in this way, but methods of delivering new genetic material into cells have been developed. Viruses can be used to deliver the CFTR gene to cells, by harnessing their ability to inject cells with foreign DNA or RNA. Patients may also be able to be given liposomes containing the functional gene, which fuse with cell membranes and deliver the therapeutic gene. Gene therapy is a growing and important area of research, and it is hoped that many diseases, including some cancers, will eventually be able to be treated using DNA.

Membrane proteins provide an entry point for viruses

Viruses that attack the human body can use the body's own membrane proteins to recognize their target cells. HIV attacks cells of the immune system. A protein on the surface of HIV called gp120 binds to CD4 protein molecules on the surface of T-cells that are involved in immunoregulation, and allows fusion of the virus with the host cell. Once the contents of the virus have entered the CD4-positive cell, the HIV genome is integrated with the host genome and uses the host machinery to make new copies of the virus. Over time, the numbers of CD4 T-cells are reduced by the virus, and the patient's immune system eventually becomes so compromised that they are unable to fight invading pathogens. Many therapeutic agents have been created to help to fight HIV, and the interaction between CD4 and gp120 is just one of the points at which drugs can be used to stop the progression of the virus.

Toxins use endocytosis to gain entry to cells and block neurotransmission

Various toxins interfere with the transmission of messages across biological membranes. Tetanus neurotoxin (TeNT) and botulinum neurotoxin (BoNT) are both protein toxins that affect nerve impulse transmission between nerves and muscles. TeNT is produced by a soil bacterium and causes the skeletal muscle spasms that characterize tetanus infection. TeNT-producing bacteria generally enter the body through wounds, and TeNT binds glycolipids enriched at presynaptic membranes of motor neurons ( Figure 14 ). TeNT then undergoes endocytosis and moves up the axon to the dendrites that connect the motor neuron to an inhibitory interneuron. TeNT is released into the synapse between these two cells and is endocytosed into the inhibitory interneuron. Acidification of vesicles containing TeNT causes the protein toxin to break apart into two domains. One of these, the L domain, is translocated into the cytoplasm of the interneuron, where it uses its proteolytic activity to cleave vesicle-associated membrane protein (VAMP). Under normal circumstances, VAMP is part of the protein complex that allows synaptic vesicles to fuse with the presynaptic membrane and release inhibitory neurotransmitters. The action of the L-domain protease of TeNT means that VAMP can no longer function, inhibiting neurotransmitter release across the synapse. As this occurs in inhibitory interneurons, the resulting effect is prolonged skeletal muscle contraction, as no inhibition is conveyed to the motor neuron to allow relaxation.

Tetanus toxin (blue circles) enters the presynaptic membranes of motor neurons by endocytosis, and moves up the axon to the dendrites that connect the motor neuron to an inhibitory interneuron. Microtubules (blue and green lines) and actin filaments (red lines) allow retrograde transport of the toxin. TeNT acts on the inhibitory interneuron, where it prevents the release of glycine (red dots), shown by a red cross. Small green dots represent neurotransmitter both inside and being released from vesicles. Adapted from Figure 2 from Rossetto, O., Scorzeto, M., Megighian, A. and Montecucco, C. (2013) Tetanus neurotoxin. Toxicon 66 , 59–63.

Botulinum neurotoxin acts in a similar way to TeNT, but has the opposite effect. Like TeNT, it is released by bacteria. BoNT binds to and is internalized by the presynaptic membrane of motor neurons at the neuromuscular junction. It is released from endocytic vesicles into the cytoplasm of the motor neuron, where it acts on the SNARE complex to inhibit the fusion of synaptic vesicles and release of excitatory neurotransmitters at the neuromuscular junction. This has the effect of blocking muscle contraction and causing paralysis in people infected with botulism. Despite its sometimes lethally toxic nature, BoNT is increasingly used by people who wish to look younger. The toxin is injected into the muscles of the face to cause paralysis, thereby reducing wrinkles and lines in the skin. When used in this way it is referred to as Botox.

Membrane proteins are the target of many drugs

Membrane proteins are important drug targets. As our structural and functional knowledge of membrane proteins expands, it is becoming possible to design more effective medicines. Computational tools are becoming an increasingly important part of the process. One important class of drug targets are pore-forming membrane proteins encoded by viruses. HIV, influenza and polio, among other viruses, encode membrane proteins that form pores in the host cell membranes in order to cause leakage and promote infection. One of these pore-forming proteins was formerly used as a drug target in the treatment of influenza. NMR studies have provided structural information about the pore-forming Vpu protein from HIV-1. Using these data together with structural information about pores with similar sequences, computational models of the structure of the channel in the host membrane can be produced. These models, combined with advanced biophysical techniques, are invaluable for predicting sites for potential drug molecule binding which can then be tested both computationally and experimentally.

More drugs target GPCRs than any other single group of proteins. As explained earlier, conformational changes in the GPCRs permit signals to cross the membrane. However, these large changes in conformation give the proteins flexibility, which makes it difficult for researchers to pinpoint the structures of the GPCRs in any one conformation. Solving the structures of different conformations using X-ray crystallography is a challenging task. As GPCRs are such important drug targets, much research has been focused on solving their structures in order to inform the discovery of new drugs. Computational methods have again proved crucial to understanding the detailed molecular structure of these proteins. Molecular dynamics simulations have been used to aid our understanding of the molecular changes that occur during GPCR activation, and also which lipids are required for the GPCR to function. Although molecular dynamics simulations are a key technique in this area of research, they have some limitations. The biological membranes that surround cells are extremely complex and contain different types of lipids and proteins, both within and associated with the membrane. At present, the time and financial resources required to provide the computational power to simulate such a complex environment are often prohibitive. Simpler models are therefore produced which, although they are able to predict conformational changes in receptors such as GPCRs, may omit other interactions that are important in the activity of membrane proteins. As is often the case, a combination of techniques will be required to gain greater insight into the conformational flexibility of the GPCRs. It is apparent from research conducted to date that these receptors can adopt many different conformations, rather than having just a simple ‘on’ and ‘off’ mechanism. Using different drug molecules to stabilize different conformations in different signalling pathways may be the best approach to finding more effective medicines in the future. There is now much pressure on researchers to replace, refine and reduce the use of animals in drug discovery (an approach referred to as the ‘three Rs’). By using computers in the early stages of the process to model drug–target interactions, researchers can produce much more promising compounds to test in experiments and drug trials.

Closing remarks

Biological membranes allow life to exist. From simple unicellular prokaryotes to complex multicellular eukaryotes such as humans, the properties of the membranes that surround cells are remarkably similar. Our understanding of the structure of these lipid bilayers is now expanding rapidly as a result of significant advances in biophysical techniques and the huge computational power now available to researchers. The proteins that inhabit these membranes allow messages to be sent and received so that the cell can communicate with the external environment. Many messages are relayed by hydrophilic molecules that require receptors to transmit information across the bilayer. It is this step that is targeted by the majority of drugs which are on the market today, as it enables us to modify the message before it enters the cell. An understanding of how membrane proteins work, how they reach the correct destinations and how we can alter their functions is key to the fight against human disease.

Abbreviations

This article is a reviewed, revised and updated version of the following ‘Biochemistry Across the School Curriculum’ (BASC) booklet: Brown B.S., 1996: Biological Membranes; ISBN: 0904498328. For further information and to provide feedback on this or any other Biochemical Society education resource, please contact gro.yrtsimehcoib@noitacude . For further information on other Biochemical Society publications, please visit www.biochemistry.org/publications .

Recommended reading and key publications

• Berman H.M., Coimbatore Narayanan B., Di Costanzo L., Dutta S., Ghosh S., Hudson B.P., Lawson C.L., Peisach E., Prlić A, Rose P.W., et al. Trendspotting in the Protein Data Bank. FEBS Lett. 2013; 587 :1036–1045. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Blobel G., Dobberstein B. Transfer of proteins across membranes. 1. Presence of proteolytically processed and unprocessed nascent immunoglobulin light chains on membrane-bound ribosomes of murine myeloma. J. Cell Biol. 1975; 67 :835–851. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Blobel G., Sabatini D.D. Ribosome–membrane interaction in eukaryotic cells. Biomembranes. 1971; 2 :193–195. [ Google Scholar ]
• Bonifacino J.S., Glick B.S. The mechanisms of vesicle budding and fusion. Cell. 2004; 116 :153–166. [ PubMed ] [ Google Scholar ]
• Butler J., Watson H.R., Lee A.G., Schuppe H.J., East J.M. Retrieval from the ER-Golgi intermediate compartment is key to the targeting of C-terminally anchored ER-resident proteins. J. Cell Biochem. 2011; 112 :3543–3548. [ PubMed ] [ Google Scholar ]
• Caffrey M. Membrane protein crystallization. J. Struct. Biol. 2003; 142 :108–132. [ PubMed ] [ Google Scholar ]
• Devaux P.F., López-Montero I., Bryde S. Proteins involved in lipid translocation in eukaryotic cells. Chem. Phys. Lipids. 2006; 141 :119–132. [ PubMed ] [ Google Scholar ]
• Fischer W.B., Hsu H.-J. Viral channel forming proteins — modeling the target. Biochim. Biophys. Acta. 2011; 1808 :561–571. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Gorter E., Grendel F. On bimolecular layers of lipoids on the chromocytes of the blood. J. Exp. Med. 1925; 41 :439–443. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Grossfield A. Recent progress in the study of G protein-coupled receptors with molecular dynamics computer simulations. Biochim. Biophys. Acta. 2011; 1808 :1868–1878. [ PubMed ] [ Google Scholar ]
• Hurley J.H., Boura E., Carlson L.-A., Różycki B. Membrane budding. Cell. 2010; 143 :875–887. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Hurst R., Rollema H., Bertrand D. Nicotinic acetylcholine receptors: from basic science to therapeutics. Pharmacol. Ther. 2013; 137 :22–54. [ PubMed ] [ Google Scholar ]
• Inesi G., Prasad A.M., Pilankatta R. The Ca 2+ ATPase of cardiac sarcoplasmic reticulum: physiological role and relevance to diseases. Biochem. Biophys. Res. Commun. 2008; 369 :182–187. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Kendrew J.C., Bodo G., Dintzis H.M., Parrish R.G., Wyckoff H., Phillips D.C. A three-dimensional model of the myoglobin molecule obtained by X-ray analysis. Nature. 1958; 181 :662–666. [ PubMed ] [ Google Scholar ]
• Lange Y. Disposition of intracellular cholesterol in human fibroblasts. J. Lipid Res. 1991; 32 :329–339. [ PubMed ] [ Google Scholar ]
• Lee A.G. How lipids affect the activities of integral membrane proteins. Biochim. Biophys. Acta. 2004; 1666 :62–87. [ PubMed ] [ Google Scholar ]
• Lippincott-Schwartz J., Snapp E., Kenworthy A. Studying protein dynamics in living cells. Nat. Rev. Mol. Cell. Biol. 2001; 2 :444–456. [ PubMed ] [ Google Scholar ]
• Lundmark R., Carlsson S.R. Driving membrane curvature in clathrin-dependent and clathrin-independent endocytosis. Semin. Cell Dev. Biol. 2010; 21 :363–370. [ PubMed ] [ Google Scholar ]
• Lynes E.M., Simmen T. Urban planning of the endoplasmic reticulum (ER): how diverse mechanisms segregate the many functions of the ER. Biochim. Biophys. Acta. 2011; 1813 :1893–1905. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Palczewski K., Kumasaka T., Hori T., Behnke C.A., Motoshima H., Fox B.A., Le Trong I., Teller D.C., Okada T., Stenkamp R.E., et al. Crystal structure of rhodopsin: a G protein-coupled receptor. Science. 2000; 289 :739–745. [ PubMed ] [ Google Scholar ]
• Risselada H.J., Grubmüller H. How SNARE molecules mediate membrane fusion: recent insights from molecular simulations. Curr. Opin. Struct. Biol. 2012; 22 :187–196. [ PubMed ] [ Google Scholar ]
• Rossetto O., Scorzeto M., Megighian A., Montecucco C. Tetanus neurotoxin. Toxicon. 2013; 66 :59–63. [ PubMed ] [ Google Scholar ]
• Singer S.J., Nicolson G.L. The fluid mosaic model of the structure of cell membranes. Science. 1972; 175 :720–731. [ PubMed ] [ Google Scholar ]
• Smith A.W. Lipid–protein interactions in biological membranes: a dynamic perspective. Biochim. Biophys. Acta. 2012; 1818 :172–177. [ PubMed ] [ Google Scholar ]
• Tate C.G. A crystal clear solution for determining G-protein-coupled receptor structures. Trends Biochem. Sci. 2012; 37 :343–352. [ PubMed ] [ Google Scholar ]
• Tighe A.P., Schiavo G. Botulinum neurotoxins: mechanism of action. Toxicon. 2013; 67 :87–93. [ PubMed ] [ Google Scholar ]
• Vaidehi N. Dynamics and flexibility of G-protein-coupled receptor conformations and their relevance to drug design. Drug Discov. Today. 2010; 15 :951–957. [ PubMed ] [ Google Scholar ]
• van Meer G., de Kroon A.I. Lipid map of the mammalian cell. J. Cell Sci. 2011; 124 :5–8. [ PubMed ] [ Google Scholar ]
• Watson H.R., Butler J., Schuppe H.J., Lee A.G., East J.M. The localization of the ER retrieval sequence for the calcium pump SERCA1. Mol. Membr. Biol. 2011; 28 :216–226. [ PubMed ] [ Google Scholar ]
• White S.H. Biophysical dissection of membrane proteins. Nature. 2009; 459 :344–346. [ PubMed ] [ Google Scholar ]
• Wimley W.C. The versatile β-barrel membrane protein. Curr. Opin. Struct. Biol. 2003; 13 :404–411. [ PubMed ] [ Google Scholar ]