26 Biomembranes I

Prof. M. N. Gupta

1. Objectives

•  To understand how centrifugation helped us in cellular fractionation and localization of biological processes in membranes.

•  To learn how microscopy, especially high resolution electron microscopy helped us in understanding structural features of the membranes.

•  To learn how lipid bilayer ultimately evolved into fluid mosaic model and what are the essential features of this model

•  To learn about finer aspects of composition of the bilayers, membrane asymmetry and membrane biogenesis

2. Concept Map

3. Description

For quite some time, it was thought that the cellular activity was restricted to the cytoplasm. Membranes were just thought to be some sort of sac in which biologically active molecules, superamolecular assemblies, metabolic activities were contained.

Gradually, this concept changed. Today, the biomembranes play as much role as the cytoplasmic entities. It is not an inert but active structure. We start our discussion on biomembranes and their functions with this modules.

Even till 1968-69, no text book had a separate chapter dealing with biomembranes. The general view of the biomembranes can be exemplified by how these were viewed in a standard text book of that era:

“The structure of the cytoplasmic membrane is of great importance in asmuch as it is the barrier between the external and internal environments of the cell”

The knowledge about plant and microbial membranes was somewhat greater than membranes of the mammalian cell. It was known that bacterial membranes are involved in oxidative metabolism.

It was known that these are essentially lipoproteins. The difference between active and passive transport were not clear. The involvement of membranes in generating Na+-K+ gradient across was known. So, vaguely the notion was there that these are not just “sacs or containers”

Two tools contributed critically to our early understanding of cellular organization in general and biomembranes as well. The first was microscopy. Electron microscopy which gave much higher resolution than light microscopy started revealing finer details of the various cells: animal cells and their organelles and microbial cells.

The second was high speed centrifugation. We tend to underestimate the contribution of this (by now a common instrument in all biochemistry labs) tool. It is not widely known that the debate about proteins/enzymes being macromolecules was settled when their centrifugation was done by Svedberg! The centrifugation, thereby also provided first good estimates of the molecular weights of proteins/enzymes. It also told us the rough (& relative sizes) of cellular organelles.

Centrifugation also was a useful tool in fractioning cellular organelles. Duve, Kuff, Anderson and Allfrey made notable contributions. The fractions obtained were characterized in terms of markers or indicator enzymes.

Thus, it was common to talk of acid phosphatase particles for lysosomes. One difficulty which did arise and created confusion is that many techniques which were used to create these “cell homogenates” destroyed the integrity of some organelles. Thus, “microsomes” were obtained from endoplasmic reticulum and electron transport particles” (ETP) were obtained from mitochondria.

Much misleading information about morphology of membranous structures accumulated as these structures were prone to getting pinched off and form vesicular components during attempts to break the cellular structure and isolate subcellular entities.

It was found that many enzymes (especially those not present on the membrane surfaces) had “latent” or “cryptic” activity. This referred to the fact that these enzymes showed no activity when present in the membranes presumably since substrate molecules could not access the binding sites. In some cases, their kinetic behaviour was different from their soluble counterparts. This relates to presence of isoforms of both soluble and membrane bound proteins.

It is here that cytochemicals approaches were useful in ruling out many artifacts. A particular worrisome fact was the finding that highly charged molecules like cyt c or rRNA could detach from their natural structures and bind to something else.

It is necessary to appreciate that for few decades, while impressive evidences accumulated about structural aspects of soluble proteins/enzymes, we practically knew nothing about membrane bound proteins. For example, all early proteins which were isolated, purified and sequenced were soluble proteins: insulin. Ribonuclease A, lysozymes, Serine protease.

Early efforts at isolation of membrane proteins were not successful. While surfactants were used to detach these from membranes, removal of surfactants often led to their spontaneous aggregation. Attempts at establishing their primary sequence by breaking the chain and fractionating peptides were not straight forward. Many of these peptides were highly hydrophobic and aggregated!

Gradually, biochemists learnt to deal with membrane proteins. Protocols for their isolation and purification were developed. Looking at fine structures of solids is possible by far fewer techniques than looking at water soluble macromolecules. FT-IR can help but that came much later. Solid state NMR is also not as easy as NMR of solutions. Similar challenges exist in the areas of immobilized enzymes and lyophilized powders used in low water enzymology.

As early as 1925, E. Gorter and F. Grendel gave the idea of a lipid bilayer. Their idea was based upon their measurements of the total surface area of a RBC and the area occupied by monolayers from the lipids isolated from RBC.

The next contribution came in the form of models proposed by J. Danielli and H. Davson in 1925 and by J.D. Robertson few decades later in 1959. These models retained the bilayer concept but placed proteins on the outer surfaces of the bilayers.

The current model called fluid mosaic model was given in 1972 by J.J. Singer and G. L. Nicolson. The name fluid mosaic model was actually used in the original title of their paper in Science Journal in that year.

By that time, a lot more information was available about structures of lipids in general and phospholipids in particular. Their fluid mosaic model also retained the bilayer concept.

According to the fluid mosaic model, phospholipid molecules are arranged in a repeating pattern to form bilayers. Each bilayer had a thickness of about 50Aº.

The lipid bilayer apart from acting as a permeability barrier also contained proteins. Some proteins were more specifically associated with some lipid molecules. The later were essential for the biological function of these proteins.

While, in general, proteins could move laterally within the membranes, their free rotation was not allowed.

Some proteins were localized at one of the two surfaces (cytoplasmic or extracellular) of the membrane. Others encompassed the membrane bilayers. As we now know, former are called peripheral proteins while the later are called integral proteins.

All in all, fluid mosaic model viewed biomembranes as more dynamic structures rather than a static one. In fact, “two dimensional solution” like phrases have been used to describe membranes according to this model.

Figure 2: (Top) Relative electron densities of rabbit optic nerve and sciatic nerve myelin membranes as a function of distance from the centre. (Bottom) Structural interpretation of the density profile.

The fluid mosaic model was supported by experimental data of various kinds. Low angle X-ray diffraction showed two peaks at the periphery with trough in the middle in the electron density plot. The polar heads of phospholipids (and Glycolipids) show higher electron density than nonpolar tails in the bilayers.

These conclusions were in line with the X-ray diffraction data obtained with phospholipid bilayers in solution which were studied as model system. Some other physicochemical data were consistent with the bilayer structure of the membranes.

It is known that hydrophilic molecules cannot pass through the biomembranes unless some integral protein act as facilitators. We will discuss this at length in subsequent modules.

At the same time, membrane proteins required the hydrophobic milieu of the membranes to be functional and often lost their biological activity when placed in aqueous solution. More often than not, these simply aggregated!

Membranes are asymmetric in structure. While this asymmetry is ensured by integral proteins not being allowed to rotate, that constraint is due to the structure of the integral proteins. Such proteins are invariably glycoproteins with oligosaccharides present on the extracellular side. These hydrophilic proteins during rotation will pass through hydrophobic interior. That is not favoured.

The asymmetry of the membranes is a critical design feature. Pumps and channels which are of tremendous biological importance depend upon this. For example, the well known Na+-K+ pump (which we will discuss in considerable detail in a later module) has the orientation that it pumps out Na+ and lets in K+.

The membranes are built by growth on pre-existing membranes. Integral proteins are inserted in an asymmetric fashion. Similarly, glycolipids also are not allowed free rotation. Different lipids like phospholipids and sphingolipids are not positioned arbitrarily.

While fluid mosaic model, by its name emphasized the fluidity of the membranes, this fluidity is not unlimited. On the other hand, depending upon its location, overall function, each membrane has fluidity tailored within a range.

Various factors define and control the membrane fluidity, the fatty acyl chains in the lipid bilayers of phospholipids can have either trans conformation or gauche conformation. Trans conformation and gauche conformations are converted into each other by rotation by 120 º either clockwise or anticlockwise.

The trans conformation leads to highly ordered state where as gauche conformation introduces elements of disorder. The membrane fluidity is controlled by this factor.

Raising temperature makes membrane non fluid. DSC of model phosphotidyl choline bilayers show the influences of the trans/Gauche conformation.

The bilayers have characteristic Tm. At this Tm, as DSC curve show there is a transition. This corresponds to all trans rigid forms changing to partly gauche forms. Length of the acyl chain of the fatty acids and their degree of unsaturation dictates the Tm.

The saturated fatty acids, if forming the acyl chains results in higher Tm. Such chains apart from favourable hydrophobic interactions, pack nicely. Presence of a double bond, as in unsaturated fatty acids, it may be recollected are in cis-conformation.

Such bends disrupts the packing, introduce a disorder and lowers Tm of the resulting membrane structure. This means more fluidity in the structure of the membranes. It is estimated that each additional –CH2- contributes -0.5 Kcal/mole to the free energy of interaction between the saturated hydrocarbon chains.

Thus, as seen in figure, three molecules of C18 saturated fatty acids stearic acid pack neatly. On the other hand, the corresponding C18 fatty acid oleic acid with just one double bond disrupts the packing in a significant manner.

In prokaryotes, these parameters viz. length of the fatty acid chain and % and degree of unsaturation in the chain are used to regulate the membrane fluidity.

If the growth temperature of E. coli is lowered from 42ºC to 27 ºC, the ratio of saturated fatty acids to unsaturated fatty acids drops from 1.6 to 1.0. Less % of saturated fatty acids ensures that at lower temperature of growth, E. coli membranes are able to retain the desired fluidity. Thus, as mentioned earlier, membrane fluidity is a critical parameter.

In eukaryotes, steroid cholesterol is a key regulator of membrane fluidity. While discussing steroids in an earlier module, we had mentioned that cholesterol should not be perceived as an undesirable molecule. It is essential to provide the desired fluidity to eukaryotic membranes. This necessary amount of cholesterol, if not available from the diet, is synthesized by the body.

If enough cholesterol is available, the cholesterol precursor HMG CoA gets converted to alternative products. Thus, the control at that branch point in the metabolic pathway ensures that enough cholesterol to modulate membrane fluidity is available at all times and at the same time body not end up synthesizing more cholesterol than it needs.

Cholesterol molecules become part of the bilayers. The –OH group of this steroid H-bonds to –C=O group of the phospholipid head. The hydrophobic part interacts with the hydrophobic tails of the phospholipids in the bilayers.

At high concentrations of cholesterol, the bilayers donot stack very well. So much so that there is no phase transition between ordered state and disordered state.

At the same time, presence of bulky steroid nucleus sterically hinders the motions of other wise flexible acyl chains of fatty acids in phospholipids. Thus, cholesterol modulates the membranous flexibility by both stopping molecular level motions as well as disrupting the ordered nature of the bilayers.

In addition to cholesterol, the membrane of the prokaryotic cells also do not contain sphingomyelin and glycolipids. The lipopolysaccharides are of course, present. Undecaprenol, also called bactoprenol, is an isoprenoid lipid in bacteria where is participates in lipopolysacharides synthesis.

Table 1: Chemical Composition of some cell membranes

general membranes contain lipids, proteins and carbohydrates. The carbohydrates are part of lipids as either as lipopoysacharides or glycolipids. With proteins, oligosaccharide chains glycosylating at post translational covalent modification stage form glycoproteins.

The functions as a containment and permeability control are carried by lipids, proteins discharge specific functions which we will discuss. As can be seen from the table, the ratio of the protein to lipid is different for the membranes of the different intracellular organelles.

Golgi membrane has least amount of protein: lipid ratio, endoplasmic reticulum has slightly higher.

Golgi also have least amount of cholesterol as compared to other lipids.

The two leaflets of the bilayers have unequal distribution of some lipids. As early as in 1972 Bretscher found that phosphatidyl choline and sphingomyelin were in the outer leaflet of the bilayer. On the other hand much of the phosphatidyl ethanolamine and phosphatidyl serine were present in the inner leaflet.

These preferential locations of the lipids were identified by chemical modification and enzymatic reactions such as use of phosholipases and sphingomyelinase.

It is worth noting that the overall amount of the lipid in both layers was similar, only its nature differed. Lipid asymmetry has been further probed by many techniques including TNBS (2, 4, 6 tri nitro benzene sulfonic acid) reactions.

Both phophatidyl ethanolamine and phosphatidyl serine contain primary amino group. TNBS is a well known reagent for quantifying free amino groups in proteins. Hence, accessibility of these two phospholipids to TNBS has provided useful information about asymmetry partitioning of lipids between two bilayers.

TNBS reaction in intact cells showed that these two phospholipids are in the outer bilayer of the cell membrane. In fact, both modified and modified forms of these phospholipids can be separated and quantified.

NMR, EPR and X-ray diffraction analysis have confirmed these conclusions arrived from chemical and enzymatic studies that asymmetric distribution of lipids in the two leaflets of the bilayer is the general property of the biomembranes.

One reason for this is that integral proteins associated with membranes are generally glycosylated and face outer surface. Reversal of this orientation will bring oligosaccharides in the membrane interior. This is hydrophobic and hence will be thermodynamically prohibited.

Equally important, many integral proteins are associated with specific lipids. Thus, orientation of these lipids is dictated by orientation of the integral proteins.

Asymmetric membrane structures are formed right at the biogenesis stage. Even a bacterial cell membrane contains > 100 different proteins and many kinds of phospholipids. Each kind of phospholipid may have several different fatty acid composition. Hence, biogenesis of membranes is a complex exercise.

Assembly of membranes

All growing cells ultimately divide and hence also needs to worry about synthesis of membrane components. The new cells will require their membranes. So, its synthesis becomes also the responsibility of the parent membrane.

At all stages, the membrane has to retain its functions, the control of permeability. This indicated that new lipids and proteins have to be synthesized and inserted into pre-existing membranes without disturbing its structural and functional integrity.

It so turned out that in both prokaryotes and eukaryotes, the biosynthetic enzymes for membrane lipids synthesis are integral proteins present right in the membranes.

The membrane synthesis requires expansion, self assembly and ensuring that right orientation of lipids and proteins is achieved. These factors also require a fine tuned logistics.

In most of the bacteria, lipids are only phospholipids. In such cases, both fatty acid acylation of glycerol-3-Phosphate and subsequent steps of phospholipids biosynthesis takes place with in the cell membrane.

In eukaryotic cells, the expansion of existing membrane takes place with the help of endoplasmic reticulum. The new membrane is formed as part of endoplasmic reticulum, pinched off it as vesicles which migrate and integrate with cell membrane.

The key design element in membrane is self assembly. Self assembly is based upon fundamental nature of noncovalent forces. Thermodynamic considerations ensures that polar structures favour hydrophilic milieu and hydrophobic interactions leads to assembly of nonpolar structures.

It is this self assembly features which allow vesicles to exist and get integrated with cell membrane. The property of amphiphilic lipids to be able to exist in multiple forms (vesicles, bilayers, fluid mosaics, liposomes) is exploited by eukaryotic cell membranes in outsourcing the synthesis of new membrane to endoplasmic reticulum.

In animal cells, lipid biosynthetic enzymes have been found to be firmly part of the endoplasmic membrane. This supported the above pathway.

In both cases, whether the site of synthesis is cell membrane (in bacteria) or endoplasmic reticulum (in eukaryotes), the enzymes synthesizing lipids are so placed that their active site faces cytoplasm.

This is a facile design as precursors for lipid biosynthesis originates in cytoplasm. Their removal from cytoplasm and being picked up by these active sites embedded in membranes is what chemists employ in phase transfer catalysis and now exploited by biotechnologists in design of multiphase reactors.

This strategy however raises another challenge.

How do these phospholipids reach outer leaflets of the membrane. If you look at the bilayer carefully, it will be realised that it also will need flip-flop of the phospholipid.

Rothman and kennedy used the gram positive bacteria Bacillus magatarium as a system of 32P label in short pulses to track synthesis of new phoshatidyl ethanolamine molecules. Use of TNBS, again proved useful to find out where such newly synthesized molecules of this phospholipid with free amino group were present.

Their conclusions were that these new phospholipids occur initially in the inner leaflets only but move over quite fast to the outer leaflet. It was estimated that this ill require 105 times faster flip flop rates than generally are thought to be permissible in biomembranes.

This flip-flop is an energy driven process, requires ATP and some enzymes to catalyse it. Provision for all this is available only for flip-flop of newly synthesized lipids, this is ensured by requirements of enzymatic catalysis.

Membrane proteins are synthesized in two different ways. Some are synthesized by polyribosomes in the cytoplasm. These are later inserted into membranes which is called post translational insertion.

In other cases, polyribosomes are bound to the membranes. In such cases, protein are inserted as these are synthesized. This is called co-translational insertion. It appears that the choice between two modes depends upon the protein sequence.

The two membrane component: lipids and proteins have turnover rates. Phospholipids have a shorter half-life, the half-lives of proteins vary in a wider range.

These turnover rates allow the membrane to renew themselves and adjust to any change in their external environments. Such external parameters include temperature and availability of the nutrition (and its composition).

Summary

•  Microsocopy and centrifugation helped in gaining early insights into membrane components
•  Lipid bilayer model
•  Fluid Mosaic model