27 Biomembranes-II

Prof. M. N. Gupta

1. Objectives

•  To learn about our early notions regarding role of proteins in membranes
•  To learn about three different kinds of proteins: peripheral, integral and lipid-linked
•  To learn about structure of integral proteins glycophorin, bacteriorodopsin and photosynthetic reaction centre.
•  To learn about how lipids are linked to the proteins and general location of such lipid linked proteins.

2. Concept Map

3. Description

We view proteins as water soluble globular proteins with the hydrophobic amino acids getting buried inside. These are the proteins which we mostly study. For example, well known hydrolases such as chymotrypsin and trypsin.

So, given that restricted version, it is difficult to understand their co-existence with lipids in membranes. Their co-existence is the result of mutual adjustment! Lipids become slightly polar and proteins organize themselves in such a way as to present their hydrophobic surface to lipophilic milieu.

It was a matter of slow realization that many biological processes occur at membranes.

Transport processes were the early focus.

Later, it was found that even many metabolic process like photosynthesis and oxidative phosphorylation involve membranes.

Most of our tools to characterize biological molecules especially proteins were suited to deal with samples in aqueous solutions. Characterizing solid samples has been a challenge.

Thus electron microscopy was valuable in looking at proteins in membranes.

It was the concept of homeostasis initially given by Claude Bernard in relation to blood and lymph in higher animals which constituted “milieu interieur” (internal milieu). It refers to the “constancy of the internal environment”

It is in the context of concentrations of inorganic cations and anions both inside and outside cells, their homeostasis, that membranes attracted initial attention from physiologists and biochemists.

In that era, the concept of membranes as semipermeable membranes which allow some specie to move across them was introduced. That these membranes consisted of proteins and lipids was known. Practically nothing was known about proteins of the membranes.

Again it is the pursuit of understanding how relative concentrations of Na+/K+ are maintained inside the cell, that involvement of a protein acting as the pump was identified. A 1968 text book mentions that “the enzyme cannot be separated from the lipid matrix of the cell with retention of these properties. Accordingly, the protein is thought to exist as a lipoprotein in lipid matrix.”

The main worry was the ability of proteins to interact with lipids. We did not know enough about the protein structure of these membrane proteins. As indicated elsewhere, even their isolation and purification in active forms was not easy.

Another membrane protein which attracted attention was E. coli permease which was part of Monod’s Lac operon enzyme. Kennedy isolated a protein from membrane fraction of E. coli cells induced by thiogalactoside. The protein showed appropriate binding affinity for the galactosides.

Even with the meagre tools, the biochemists nevertheless had classified transport into active and passive ones. Also conformational changes and phosphorylation/dephosphorylation were being implicated in the role of protein in transport across membranes.

Eventually, as tools to look at membranes developed and enzymologists learnt how to isolate and characterize proteins associated with membranes, a broad classification emerged.

Integral proteins

Integral or intrinsic proteins are firmly bound to the other components of the membranes (i.e. lipids) by hydrophobic interactions. Such proteins, in fact, cannot be isolated without disrupting the membrane structure, while, no covalent bonds are involved, the hydrophobic milieu ensures that this binding is quite strong.

Organic solvents, detergents like SDS or triton x-100 and even chaotropic compounds can be used to disturb the membrane structure and hence help in releasing such proteins, chaotropic compounds.

Chaotropic compounds/ions are substances which disturb the water structure and facilitate interaction between nonpolar substances and water, thereby enhancing solubility of the former.

Guanidinium chloride and urea in the range of 5-10 M are good chaotropes in nature.

Once released, these integral proteins tend to aggregate extensively and form precipitates. To prevent that these are generally kept in solution by using detergents or water miscible organic solvents. The use of butanol in protein extraction in classical enzymology was an empirical approach, the basis of which is clear now.

These integral proteins are seldom completely buried inside the membrane. These normally have three components. One is the extracellular one and the second is in the cytoplasm. It is the middle component which interacts with the membrane. Hence, the word transmembrane proteins is sometime used for such proteins.

Peripheral or extrinsic membrane proteins

Many proteins are bound to membranes weakly. This originates in the fact that their interactions are linked to exterior part of the membranes. Washing with buffers of appropriate pH and ionic strength releases these proteins.

Peripheral proteins do not interact predominantly by hydrophobic interactions. Hence, their surfaces are not extensively hydrophobic. Thus, once extracted these are soluble in aqueous buffers. These are comparatively easy to work with.

In many cases, peripheral proteins interact with outer parts of the integral proteins through H-bonding or electrostatic interactions. Cyt c is the example which is a familiar protein. Cyt c is a part of the respiratory chain. As a peripheral proteins, it is associated with the outer surface of the inner mitochondrial membrane.

Lipid linked proteins

Some proteins are anchored to the membrane by being a lipoprotein. The lipid is covalently conjugated to the protein at post translational covalent modification step.

Lipids, like any other conjugated structure, affects the structure and function of the protein. It mediates the interaction of protein with the milieu. It also mediates the interaction with other proteins.

There are three kinds of lipid linked proteins. Prenylayted proteins, fatty acylated proteins and glycosylphosphatidylinositol-linked proteins. In some cases, more than one lipid molecule is linked to the protein molecule.

Figure 1: Two different ways in which transmembrane proteins may be embedded in a lipid bilayer (grey).

There are two ways an integral or transmembrane protein is present in association with the membranes. In the first mode, the polypeptide chain crosses through the membrane interior only once.

In such cases, we talk of hydrophilic N-terminal, C-terminal and the hydrophobic transmembrane part. The transmembrane part is almost always a transmembrane helix.

Such amphiphilic proteins have the functional globular domain at either end of the protein. Such domains can often be obtained undisturbed in their structure by proteolytic cleavage.

Haemagglutin and neuraminidase components of the influenza virus and HLA proteins are well known examples of such functional entities obtained by the proteolytic cleavage.

In order to understand this, it may be useful to recollect fundamental notions of domains in protein structure.

As fine structures of many proteins came to be known from X-ray diffraction studies, it was realised that few types of secondary structures often occur in combination in the same way in many proteins. Such arrangements are called supersecondary structures or motifs.

In some cases, even a single motif is responsible for a specific function, in other cases, it is a part of larger functional assembly.

Domain can be considered a unit of tertiary structure. It is a part of the polypeptide chain that can independently fold into a stable tertiary level structure.

In many case domains are capable of an independent function. A protein may contain one domain or often several domains.

A good analogy will be to view protein molecules as obeying federal form of the governance. The protein has a federal structure: domains akin to the state or regional governments, geographically (structurally) separated and functionally independent in many ways, are integral part of the protein.

Another kind of integral protein has a different kind of association with the membrane. In such cases the polypeptide chain crosses the membrane several times. The protein in such cases has three kinds of parts which are hydrophilic.

Both the N-terminal and C-terminals are hydrophilic just like in the first kinds of membrane proteins. The novel hydrophilic components are loops which stick out as the transmembrane protein emerges out of and re enters the membrane.

The loop may be a simple one or a bunch of loops. All permutation and combination are known.

In such cases, proteolytic cleavage almost never produces the functional domains.

Understandably, many fragments of different sizes can be obtained.

Human erythrocyte glycophorin A is an example of well studied integral protein. Three clear domains can be identified in this protein.

The N-terminal domain has 72 amino acid residues and is heavily glycosylated with 16 oligosaccharide chains. Out of these 15 are O-linked and one is a N-linked oligosaccharide.

The transmembrane part of this protein has 19 amino acid residues, almost all of this membrane spanning segment consist of hydrophobic amino acids. There is not even a single amino acid with a charged side chain. There is a single threonine but it is flanked by two of the most hydrophobic amino acids: Ile and Leu.

The cytoplasmic C-terminal has many amino acids which are acidic and basic. Again, whenever there is an occasional hydrophobic amino acid like at 98 and 99 (Leu and Ile), the amino acids preceding these are two Arg and Lys each!

In fact two variants of glycophorins are known. Glycophorins AM has Ser and Gly whereas Glycophorin AN has Leu and Glu at positions 1 & 5 which are, of course part of external N-terminal domain.

As we deal with few more integral proteins we will see that transmembrane part of such proteins is helical at least in part. So, such proteins have either one helix or several in their transmembrane parts.

Presence of α helix in such cases can be predicted by computing the free energy change in transferring the amino acid segment from the interior of the membrane to water. The G higher than +85 KJ/mol suggest the possibility of α helix formation.

Among the amino acid side chains, Ala (A), Glu (E), Leu (L) and Met (M) promote α-helix formation. On the other hand pro (P), Gly (G), Tyr (Y) and Ser (S) are poor in forming α-helix.

However, this information alone lead to poor prediction about α-helix formation. The most favoured position of α-helix is on the outside of a protein structure. In such a situation, one side of the helix faces exterior while other side faces the hydrophobic interior of the protein.

Hence, it is not surprising that transmembrane segments prefer α-helix formation. By its very design, one side of the helix anyway has a thermodynamically stable existence. α-helices can be also completely buried in such cases or partially buried.

The above table has amino acid sequences from three helices present in the different proteins. The first one, 260-270 segment from citrate synthase forms a completely buried helix. The second is 355-365 ADH segment represents a partially buried helix. The troponin C sequence (87-97 residues) forms a helix which is completely exposed.

This behaviour of α helix to co-exist with hydrophilic and hydrophobic surroundings originates in its own design. The peptide bond has a dipole moment with fractional charges on the atoms of the peptide units as shown.

The dipoles of these peptides units are aligned along the axis of the α-helix. The α-helix has 3-6 residues per turn, so with a periodicity of every 3 to 4 residues, side chains tend to change from hydrophobic to hydrophilic.

Keeping in mind that each turn is 3.6 amino acid residues long, each residue can be plotted as part of concentric circles (forming the spiral) at 360/3.6= 100º. This kind of plot displays position of the residue on a plane perpendicular to the α-helix axis.

In cases of ADH, it is clear that one side of the α-helix has hydrophilic residues and the other side has hydrophobic residues. This is in line with the fact that then α-helix is partially exposed.

With citrate synthase, predominantly hydrophobic character is obvious. In case of troponin C, the helix has all polar amino acids and the helix is completely exposed to hydrophilic milieu.

Before we go any further, it is useful to take this discussion further and try to understand why in some cases, several helices occur as a part of the transmembrane segment of the membrane proteins.

In fact, such α-domain structures are quite common. As indicated earlier, these consist of a bundle of α-helices with a loop connecting these α-helices.

α-helices, in fact favour structures in which these co-exist. A very common occurrence is 4-helix bundle. When such bundle exist, the adjacent α-helix are usually antiparellal with respect to each other.

This arrangement leads to a hydrophobic core in the middle of the bundle along its length. The side chains facing inwards exclude water.

these proteins just like other protein which are soluble in aqueous buffers. However, mild detergent like octyl glucoside if added, solublilize these proteins in their native conformation.

Such protein-detergent complexes consist of hydrophilic parts of the detergent forming the surface as it is the hydrophobic part which combines with the hydrophobic surface of the membrane proteins.

Such protein-detergent complexes have proven useful for both purification and crystallization of membrane proteins. For crystallization, a suitable amphipathic molecule (identified by hit-and trial) is added to the protein-detergent complexes.

Apparently, this molecule facilitates the necessary packing interactions in the crystal. Some of the early membrane proteins which could be obtained in big enough crystal sizes for X-ray diffraction are an outer membrane protein from E. coli called porin; an enzyme prostaglandin synthase and photosynthetic reaction centres. The photosynthetic reaction centres are from two bacterial species called Rhodopseudomonas viridis and Rhodobacter sphaeroides.

Thus, while glycophorin was the first transmembrane protein to be sequenced, the detailed picture about higher order structure of a membrane protein came from high resolution electron microscopy of 2-dimensional crystals. Such crystalline sheet are easier to obtain. In many cases, they are present as such in vivo.

Rhodopsin from a purple bacteria Halobacterium halobium is a membrane protein and is called bacteriorhodopsin. The bacterium has been extensively studied in the context of photosynthesis. This protein occurs as an ordered sheet.

Richard Henderson and Nigel Unwin at MRC, Cambrige looked at it by tilted low dose electron microscopy. The high resolution technique enabled them to construct 3-D image from such 2-dimensional samples.

This excellent work had two important implications. Firstly, it discovered 7- transmembrane helices. While the earlier work in 1975 was at 7 Aº resolution, later follow up in 1989 at 3 Aº resolution confirmed and gave a clearer picture of the 7-helical structures.

These studies in fact were the first ones to provide the experimental proof that our predictive models to predict helices based upon amino-acid sequence were quite good.

The technique later on was used to obtain similar information about acetylcholine receptor, ion pumps, gap junctions, and cytochrome oxidase. The last protein is an enzyme which crystallises in two dimensions.

We had earlier mentioned the propensity of the α-helices to form helix bundle structures. The seven-helix bacteriorhodopsin is actually a light driven proton pump. The protein has 247 amino acid residues with a retinal moiety attached to its Lys216.

The cell membrane has ~ 0.5 µm wide patches and contains 75% rhodopsin and 25% lipid. The helices of the protein are ~25 amino acid rods spanning the bilayer of lipids in an almost perpendicular fashion to their plane.

These helices are also ordered in an antiparallel fashion and connected with loops of varying size. The charged side chains of the protein are exposed to the aqueous environment. The loops, ofcourse, are also largely made of hydrophilic amino acids.

It was in a 1982 conference in Sicily that Hartmont Michel presented the X-ray diffraction pattern of a crystal of large complex of polypeptide chains which constitute photosynthetic reaction centre from this bacterium Rhodopsudomonas viridis. Later, 2.5 Aº resolution data was also obtained by him in collaboration with Hans Deisenhofer and Robert Huber. Robert Huber was awarded Nobel prize.

The protein has nonidentical subunits of ~300 residues and contains 4 chlorophyll molecules, 4 other pigments and one nonheme Fe. The photosynthetic centre has 1187 amino acid residues. The membrane spanning part has 11 α-helices forming a 45 Aº long cylinder with hydrophobic exterior.

An interesting finding has been that even in such proteins, the buried amino acids are hydrophobic. However, the exterior (membrane exposed) amino acids are also and if anything, even more hydrophobic. The surfaces of the proteins are designed keeping in mind the polarity (or nonpolarity) of their milieu.

Lipid linked proteins

Generally palmitic acid and myristic acid are linked to membrane proteins. Myristolyation with the C14 saturated fatty acids occurs to the N-terminal Gly. Such proteins occur in cytosol, endoplamic reticulum, plasma membrane and even nucleus.

Palmitoylation (C16 saturated fatty acid) is via thioester linkage to specific Cys. Such proteins are only found on the cytoplasmic face of plasma membranes, where they are involved in signalling.

Proteins are farnesylated or geranylgeranylated at Cys via a thioester linkeage. GPI-linkages are more common in parasitic protozoa but do occur in all eukaryotes. These are located on the exterior of the plasma membranes.

Single or multiple helix bundles are not the only ways transmembrane domains of integral protein are organized. Porins, the gatekeeper in membranes which allow entry of necessary nutrients in the cell, contain transmembrane β-barrels. We will look at that when we talk of the membrane transport.

Electron microscopy contributed very much to our understanding of membrane protein as these proteins were found to be difficult to crystallize.

The X-ray diffraction picture of the photosynthetic centre, when presented in that conference was the first X-ray diffraction study of membrane protein. Given that and its complex structure and immense biological importance actually made scientists listen to the presentation in pindrop silence. Science, does have its exciting moments.

Summary

We learnt about:

•  Early ideas about role of membranes and its proteins

•  Three different kinds of membrane proteins

•  Structures of Glycophorin, bacteriorhodopsin and photosynthetic reaction centre

•  Lipid linked proteins