Moss - What genes cant do - 2003 (522929), страница 23
Текст из файла (страница 23)
It also plays an intergral role in the cellular interactions thatinduce changes in cellular states, and in turn it is subject to being chemically transformed by cells undergoing changes of state. Even terminallydifferentiated cells which are stabilized within some tissue matrix displaycell-surface receptors that are responsive to blood-borne chemicalmessengers such as insulin, growth factors, and the like.
Cell-surfacereceptors are themselves glycoconjugates, i.e., protein molecules witholigosaccharide attachments. The biochemistry of the oligosaccharides86Chapter 3may significantly impact the binding properties of the active site of thereceptor.The construction of the glycocalyx occurs in a stepwise fashion.
Theelaboration of oligosaccharides attachments to proteins and lipids takesplace as the proteins and lipids pass through the membrane system justdescribed. Each compartment, each “pancake in the stack,” has its owndistinctive biosynthetic capacity. The glycoprotein chain is modifiedaccording to sequential exposure to the glyscosyl-transferase enzymes ofthe respective compartment and the availability of saccharide primers.The possibilities of oligosaccharide chain elongation are highlycomplex. Sugar units (the building blocks) may be removed as well asadded along the way, chain branching may or may not occur, or different sugar-adding enzymes (glycosyltransferases) may or may not bepresent in a certain compartment at a certain time, and alternative glycosyltransferases may compete for the same growing chain.
With all ofthese variables, subtle diffences in the cellular context may influence theglycosylation pathway. Changes in cell shape, for example, that wouldbe influenced by cell-cell or cell-matrix attachments could affect glycoslyation patterns. But glycosylation patterns, in turn, which alter theglycocalyx of the cell can affect cell-cell and cell-matrix adhesion andthus the inductive state of the cell, the shape of the cell, and so on. Onecan envisage complex feedback loops that may become established withor without the involvement of changes in the transcriptional activationstate of DNA.
Further, one can plausibly postulate the possibility of“organizational mutations” which become stabilized and propagatedduring the subsequent developmental history of the organism and evenconceivably transmitted to progeny.Thus far I’ve described the differentiation along the radial axis of theprincipal system of cellular membranes. Now we turn to a considerationof the organization and movement of proteins in the plane of themembrane (orthogonal to the axis just discussed); in so doing will wedraw on some theoretical work of Max Delbrück. Whether Delbrückconceived of this work as continuous with, and a logical extensionof, the issues broached earlier by him and Schrödinger I can’t say, but Ihope that it will become evident why the suggestion of such continuityA Critique of Pure (Genetic) Information87is warranted.
The membrane dynamics to be discussed focus on experimental work performed on the plasma membrane, but there should beno reason why the findings couldn’t pertain as well to the many internalpancake membranes that are not as readily accessible to experimentalinquiry.In 1972 Singer and Nicolson first described the concept of cellularmembranes as bilayers of amphipathic phospholipids with embeddedproteins as the “fluid-mosaic” model. It had already been shown by Fryeand Edidin (1970) that if a mouse and human cell are fused (by use ofa Sendai virus “fusogen”), both the mouse and human antigens becomequickly mixed together and distributed generally around the surfaceof the “heterokaryon” cell produced.
Cell-surface antigens are generallytaken to be glycoproteins and glycolipids embedded in the membrane,and these early results were suggestive of free diffusion in the plane ofthe membrane. In 1975 Max Delbrück and a colleague provided a theoretical model for the diffusion of proteins, approximated as cylindricalobjects diffusing freely in a membrane. For a cylinder with radius adiffusing in a viscous sheet with thickness h and viscosity m, which isbordered by a less-viscous fluid of viscosity m¢, the coefficient for lateraldiffusion DL is given as DL = KBT/4pmh(ln mh/m¢a - g) with KB = Boltzman’s constant, T = temperature, and g = Euler’s constant (~0.5772)(Saffman & Delbrück, 1975).
Using “ballpark” values for the respectiveparameters, a = 25Å, h = 35Å, m = 1 - 10 poise, m¢ = 10-2 poise, KBT =4 ¥ 10-14 ergs, results in a predicted DL of 6.0 ¥ 10-8 to 6.0 ¥ 10-9 cm2/sec(Cherry 1979).I have previously indicated the strong evidence for the nonrandom distribution of proteins at the different levels of the membranous stacksthat comprise the principal membrane system of the cell. The Saffmanand Delbrück model provided a handle for determining whether themovement of proteins within the plane of a membrane is purely diffusional (and thus random) or constrained in ways that could prove to beof biological significance.
One can see how Delbrück’s exploration ofthis question follows the same lines of interest expressed by Schrödingerconcerning the physical basis of the preservation of biological order.Random patterns of distribution are information-poor and require88Chapter 3little by way of explanation. Highly ordered and specific patterns oforganization, as we’ve seen with respect to the radial or concentricpattern of compartmentalization in the cell, do pose explanatory challenges with respect to the preservation of information. We will nowconsider evidence as to whether the movement of proteins in the planeof the plasma membrane—and thus largely orthogonal to the directionof the differentiation of the membranous strata—also poses such achallenge.The principal technique that has been used for examining the lateralmovement of membrane-embedded constituents has been fluorescencerecovery after photobleaching (FRAP).
In this method, cell-surface components are labeled with a fluorescent dye. A laser is focused on a small(1 to 10 mm2) area on the surface of the labeled cell. Fluorescence in thisarea is monitored by a photomultiplier. By momentarily increasing theintensity of the laser by 103 to 105-fold the fluorophores in the regioncan be photochemically bleached. The laser is then attenuated andfluorescence once again measured. The time it takes for full recovery offluorescence (due to the diffusional replacement of the bleached fluorophores) is used to calculate the DL of the membrane-embedded cellsurface components. The level of fluorescence immediately after bleaching is designated f0. The complete or highest fluorescent level is designated fµ.".
From the times it takes to recover maximum fluorescentrecovery is derived the t1/2, or half-recovery time. The diffusion coefficient, DL, is then given by the equation DL = (w2/4t1/2) g, where n is theradius of the beam, t1/2 the half-recovery time, and g a parameter whichaccounts for the degree of bleaching and beam profile. Under typical conditions g = 1.3 (Cherry 1979).Simple-model membranes can be constructed in the laboratory inwhich nothing is present above or below the plane of the membrane.Various proteins can be incorporated into these model membranes andthe lipid composition can be altered in order to evaluate the influence oflipid composition on diffusional coefficients.
The model membranes provided a good opportunity to test the predictions of the Saffman-Delbrückequation. In a number of FRAP studies using different lipid compositions and different proteins, DL’s were found to be in the 10-7 to 10-8range, which is in fairly good accordance with the Saffman-DelbrückA Critique of Pure (Genetic) Information89predictions (Smith et al. 1979, Vaz et al.
1981). The model membraneswere also used to assess the effects on diffusion of the presence ofhydrophilic components in the diffusing protein. Structurally, the membrane can be imagined as being like a thick sandwich with fairly thinslices of bread. The bread would constitute the hydrophilic parts of themembrane with the much thicker hydrophobic lipids sandwichedbetween. The hydrophilic aspects of a membrane protein (which mayinclude oligosaccharide chains) will be “dangling” above or below theplane of the membrane, being significantly longer than the thin width ofthe hydrophilic part of the membrane. The model membranes allow forthe effects on diffusion of the hydrophilic protrusions to be uncoupledfrom any effects due to the interaction between such protrusionsabove or below the plane of the membrane with other intracellular orextracellular components of a real cell.
Comparison of the diffusioncoefficients of the memberane protein gramicidin, which possesses nohydrophilic portion, with that of glycophorin, which has a largehydrophilic portion, revealed little difference, suggesting that hydrophilicprotrusions from the membrane as such are not important for determining diffusion rates. This will be seen to become important as thepossibility of diffusional constraints based on biologically specific interactions between hydrophilic groups and cellular or extracellular constituents outside of the membrane is considered.FRAP studies carried out on a number of different living mammaliancells (in culture) revealed two major findings: (1) cell-surface proteinsappeared to be present in a mixture of mobile and immobile fractionsand (2) the mobile fraction is generally at least two orders of magnitude(100-fold) slower (~10-10 cm2/sec) than the range of diffusion mobilitiesfound in the model membranes.
Subsequent studies provide strong evidence for the cause of both of the above to be based on constraininginteractions between the membrane-embedded proteins and cellular constituents within and adjacent to the cytoplasmic face of the membrane.The possibility that a reduction of diffusion rate might be due toincreased viscosity within membranes in which the protein-to-lipid ratiois very high was discounted by the finding that the photoreceptorpigment protein rhodopsin, which is packed at a maximally dense ratioof approximately one to one with membrane lipids in the outer segment90Chapter 3of retinal rod cells, actually diffused at the comparatively rapid rate ofDL = ~3.5 ¥ 10-9 cm2/sec (Poo & Cone 1974).The advent of electron microscopy in cell biology revealed, amongmany other things, the presence of various kinds of filamentousstructures within the cell, collectively referred to as “cytoskeleton.” Thepresence of something like a cytoskeleton in cells—especially if it ishighly specific in form with respect to cell type, developmental stage, andso forth—is clearly relevant to our concerns about structural information, its transmission, and its relationship to the genome.
I will introducethe cytoskeleton now within the context of its putative role in constraining the free diffusion of membrane-embedded proteins.Support for the cytoskeletal interaction with integral membrane proteins was provided by studies of diffusion in membrane “blebs.” Blebsare protrusions of the plasma membrane that are thought to be detachedfrom the underlying cytoskeletal connections. Blebbing is induced byvarious factors, such as cross-linking of membrane proteins, anoxia,physical injury, and prolonged protease treatment of a cell. Wu et al.(1981) induced blebbing in mouse lymphocytes and compared the difference of diffusion rates of both protein and lipids between the blebbedand normal cell-surface membrane. The diffusion rate of protein in theblebs was 1000-fold faster than in unblebbed normal membrane, whereblebbing only enhanced lipid diffusion by a factor of 4. This differencelends support to the idea that membrane proteins, but not lipids, becomeassociated with cytoskeletal structures which limit their mobility.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
