Indian Journal of Ophthalmology

: 1972  |  Volume : 20  |  Issue : 2  |  Page : 55--62

The subunit structure of chick lens crystallins and its relationship to their antigenic properties

D.E.S Truman, RM Clayton, A.T.H Burns, JC Campbell 
 M. R. C. Epigenetics Research Group, Institute of Animal Genetics, West Mains Road, Edinburgh EH9, 3JN, United Kingdom

Correspondence Address:
D.E.S Truman
M. R. C. Epigenetics Research Group, Institute of Animal Genetics, West Mains Road, Edinburgh EH9, 3JN
United Kingdom

How to cite this article:
Truman D, Clayton R M, Burns A, Campbell J C. The subunit structure of chick lens crystallins and its relationship to their antigenic properties.Indian J Ophthalmol 1972;20:55-62

How to cite this URL:
Truman D, Clayton R M, Burns A, Campbell J C. The subunit structure of chick lens crystallins and its relationship to their antigenic properties. Indian J Ophthalmol [serial online] 1972 [cited 2020 Nov 25 ];20:55-62
Available from:

Full Text

Immunglogical methods have long played an important role in studies of the crystallins, and because of the absence of enzyma­tic function from these structural proteins it is probable that anti­genic reactions will continue tc predominate among the methods available for research on crystal­lins. As techniques available have improved, so the complexity of the immunological data has in­creased. We may hope eventually to understand this complexity in terms of the detailed molecular structure of the crystallins, but this goal still remains somewhat distant. Nevertheless it is now becoming possible to relate immu­nological results obtained with crystallins of some species to the quaternary structure of the pro­tein molecules and so to move to­wards a molecular understanding of crystallin structure and func­tion.

Our studies of the crystallins have concentrated mainly on those of the chick as this species is most convenient for work on embryonic development, but we have benefited from the large amount of work that has been done on other species, particularly the chemical work on bovine cry­stallins. It is our belief that results obtained with chick lenses may ultimately be of value in under­standing developmental processes in other species and so eventually contribute to clinical ophthalmo­logy.

The particular value of the lens as a system for studying cell diffe­rentiation has been emphasised by Clayton (1970), and much of our data discussed in the context of background knowledge on crystal­lin structure, function and deve­lopment. This article seeks to bring together our own work on the immunological behaviour of chick crystallins and the way in which this may be understood in terms of protein structure. Unfor­tunately the nomenclature of the chick crystallins has been un­settled for some time and in our earlier papers we have followed different schemes of nomenclature from that now used, which is based on that used by Zwaan and Ikeda (1968).

It was shown by Bloemendal. Bent, Jongkind and Wisse (1962) that bovine α- crystallin molecules, with a molecular weight of 810000. could be broken down into much smaller subunits by treat­ment with reagents such as 7M­urea, which breaks hydrogen bonds but does not break the co­valent bonds that hold together protein molecules. It thus appear­ed that the large molecules of α-­crystallin were composed of many smaller subunits held into aggre­gates by hydrogen bonds. Subse­quent work has shown that each of these subunits is composed of a single polypeptide chain, and that more than one type of subunit is present in the total α-crystallin molecule. There has been some controversy about the molecular weight of the subunits, and a value of less than 20000 has been sug­gested by Hoenders. De Groot. Gerding and Bloemendal (1969) Some work has been done on the determination of the amino acid sequences of the α-crystallin sub­units and there are apparently very great similarities between the different types of subunit (Corran and Walley, 1969).

Largely on the basis of studies of antigenic specificity we pro­posed in 1967 that all of the three major classes of crystallin in the chick lens were probably com­posed of subunits and that within each class more than one type of polypeptide chain was found (Clay­ton and Truman, 1967). Subsequent studies have confirmed these views and given us more informa­tion about the relationship between the antigenic properties and pro­tein structure of these crvstallins

Chick α-crystalin. Hoenders (1965) found by analytical ultra­centrifugation that the proteins of this group had a molecular weight of 470000, while Zwaan (1968), using a method dependent on elec­trophoresis in polyacrylamide gel, found three separate molecular species in this class with molecular weights of 400.000. 350,000 and 60.000 Three major components in the α- crystallin group were also shown by Truman (1968), though the resolution of these compo­nents becomes less clear as the lens ages (Truman, Brown and Campbell, 1972). Corresponding to these three size classes of α-cry­stal.lin, we also find that immuno­electrophoresis may reveal three concentric precipitin arcs (Tru­man, Clayton and Campbell, 1967). The three components may also be revealed in quantitative immunoelectrophoresis and in the Osserman tests when embryonic lens extracts are compared with adult crystallins (Truman et al., 1972). During embryonic develop­ment there appears to be a change either in the proportions or in the properties of these three classes of α-crystallin.

There has in the past been some controversy about changes in the physical properties of α-crystallin of various species during embryo­nic development (reviewed by Zwaan 1963), and a number of workers have shown that the sub­unit composition of bovine α-cry­stallin changes during development (Schoenmakers and Bloemendal, 1968; Plamer and Papaconstantinou, 1968, 1969). Direct evidence of the subunit structure of chick α- crystalline and its change in com­position during development has been put forward by Rana and Maisel (1969), and by Clayton (1969) who separated the different subunits of α-crystallin from adult chicks and from the lenses of 13 day and 18 day embryos. While the adult lens showed three differ­ent major types of subnit, the em­bryonic lenses showed only two.

We see, then, that during the development of the chick the com­position of the α-crystallin changes and that there is also evi­dence that the proportions of the different molecular populations of the α-crystallin aggregates may also undergo change. There is, however, no evidence of the ap­pearance of any different anti­genic determinants after the first appearance of α-crystallin anti­gens at about embryonic stage 25 (Truman et al., 1972). This implies that the different subunits are probably antigenically very simi­lar. Such immunological simila­rity is seen between the subunits of bovine α- crystallin, in which several subunits appear to be immunologically identical (Bjork, 1964), in agreement with the simi­larities of amino acid sequence re­ported by Corran and Waley (1969). It is probable that there is a gene­ral structural similarity between the α- crystallins of mammals and birds.

β-crystallins. The β-crystallin class in the chick is that with the lowest molecular weight and with a wide range of electrophoretic mobility. The class is so named because of their immunological cross-reaction with mammalian β-crystallins (Zwaan and Ikeda, 1968). In terms of molecular weight these proteins are heterogeneous, resolving by gel filtration into two main groups Truman, 1968) and probably ranging in molecular weight up to about 60,000 (Truman, Brown and Rao, 1971). When stud­ied by electrophoresis on polyacry­lamide gel as many as fifteen components have been found among the β-crystallins of the chick, with mobilities more extre­me then either the α-or β crystal­lins (Truman, 1968).

On treatment with 8M-urea the β-crystallins show a reduction in sedimentation coefficient from about 6.44S to 1.28S. and estimates of the molecular weight of the subunits by gel filtration give a value of about 16.500 (Truman et. al. 1971). This result. and the data of Rana and Maisel, (1970), con­firm the prediction of a subunit structure of the β-crystallins which had been proposed earlier from indirect evidence (Clayton and Truman, 1967).

Experiments in which β-crystal­lins separated by electrophoresis on polyacrylamide gels were then dissociated in urea solutions and re-electrophoresed have shown that there are several different subunits which form the aggre­gate β-crystallin molecules; that most β-crystallins contain more than one type of subunit and that the electrophoretic mobility of the aggregate is a function of the constituent subunits (Clayton, 1969). These experiments indi­cated the presence of six major types of subunit and five minor types among the β-crystallins. Subsequent experiments in which β-crystallin preparations have been subjected to isoelectric focussing on columns in the pre­sence of 7M-urea have also shown 6 major and 5 minor components with isoelectric points ranging from pH 7.72 to pH 5.50. A further confirmation has come from expe­riments in which the β-crystallins were first fractionated by isoelec­tric focussing in the absence of urea, to yield 9 different /3-crystal­iin preparatons. These were each dissociated and analysed; by poly­acrylamide electrophoresis in urea at pH 8.9. When allowance was made for any possible cross-conta­mination, two fractions were found to contain only one major type of subunit, others contained two types, while two fractions each contained three types of sub­unit. Some of the subunit types were found in more than one aggregate molecule.

In immunoelectrophoresis the β-crystallins of the chick give rise to an extended series of precipitin arcs, known sometimes as the long line' (Zwaan, 1963). This it­self implies that there are a series of immunologically related pro­teins which vary in their electro­phoretic mobility. The spurs which are found in the `long line' also indicate that some antigens are not distributed over the entire range of electrophoretic mobility of the β-crystallins. The pattern of occurrence of the various sub­units in these crystallins provides a basis for understanding the im­munoelectrophoretic pattern, since it can be shown that many of the subunits which differ in their mobility also differ in their anti­genicity. However we cannot say that `one subunit is equivalent to one antigen'. Firstly, no one sub­unit appears to occur throughout the entire electrophoretic range of the `long line', and yet immuno­electrophoresis indicates that one. and possibly two, antigens are common to the entire length of the precipitin line. Thus one determinant must be associated with more than one type of sub­unit. This has been confirmed by tests on isolated subunits. Such tests also show that a single sub­unit may possess more than one antigenic determinant (Truman and Clayton, in prepn.).

The correlation between sub­units and antigenic determinants is also to be seen as a result of studies of the β-crystallins during development, since the increase in complexity of the `long line' arcs is paralleled by an increase in the number of different types of sub­unit in the β-crystallins of the chick

A feature comparable to the 'long line' of the chick crystallins is seen in immunoelectrophoresis of crystallins of the toad Xenopus laevis (Campbell, Clayton and Truman, 1968).

stallin. That fraction of the chick crystallins intermediate in molecular weight between the α- and β-crystallins has been called γ-crystallin to emphasise that it is not immunologically cross-reactive with any of the major classes of crystallin of the mammalian lens (Zwaan and Ikeda, 1968). However, it should not be supposed that this group of proteins is analogous with the Y- crystallins of the fish (Bon, Swanborn, Ruttenberg and Dohrn, 1964). The molecular weight of these proteins has been estimated at about 150,000 by Hoenders (1965) and. Truman et al. (1971). but Zwaan 1968 has given evi­lins with molecular weights rang­dence of five classes of Y-crystal­ing from 155,000 to 460,000. This polydispersity in size may account for the thickness of the precipitin arc which is usually found when γ-crystallin is examined by im­munoelectrophoresis, and which has been resolved into at least three components (Truman et al., 1.967).

There is evidence that the γ-­crystallins are also composed of subunits, since treatment with urea reduces the molecular weight to about 25.600 . Six such subunits would give an aggregate molecule of weight 154,000 (Truman et al., 1971). Clayton (1969) carried out electrophoresis on polyacrylamide gel at pH 8.9 of purified γ-crystal­tin and detected five different components, while studies with starch gel in urea at pH 5.6 have shown at least four different sub­units. While the starch gel is less sensitive and its resolution at this pH is poorer than the acrylamide, the avoidance of alkaline urea diminishes the possibility of arte­facts caused by the carbamylation of proteins (Stark, Stein and Moore, 1960.)

A further method of fractiona­tion of crystallin subunits which is both sensitive and with a high degree of resolution is isoelectric focussing on polyacrylamide gels. In this system the gel stabilises the pH gradient set up by the ampholytes, but does not act as a molecular sieve (Wrigley, 1958). If total adult crystallin is incor­porated into gels containing 6M­urea a total of 14 major subunits can be detected using ampholytes with a pH range of 3 to 10. This agrees well with the results of other methods applied to total lens and with the sum of major subunits from the various crystallin fractions. The isoelectric focussing technique in gels is now being used to investigate the sub­unit composition of purified crys­tallins and its resolution is being increased by varying the pH range of ampholytes used.

By the use of these different methods we feel confident that the γ-crystallins of the chick are also aggregations of different type of polypeptide chains. Studies of the ontogeny of these crystallins indicate that the relative proportions of the differents subunits in the aggregate mole­cules may change during develop­ment, just as there are changes in the α-and β-crystallins.

Dissociation and reassociation.

The result described above have involved the dissociation of cryst­allin molecules by treatment with urea solutions which weaken hydrogen bonds between the con­stituent subunits of the aggregate molecule. When the urea is removed the subunits may reaggre­gate, at least partially. We showed that dissociated crystallin subunits sedimenting at about 1.9s formed, on removal of the urea, an aggre­gate sedimenting at 5.9s and which retained an immunological reaction with antiserum to crystallins. It was also possible to show that aggregate molecules could be for­med in this way which involved combinations of subunits not pre­sent among the native crystallins (Clayton and Truman, 1967).

Other hydrogen bond reagents. such as 6M-guanadinium chloride and formamide, can cause reversi­ble dissociation of crystallins, and there is also evidence of a reversi­ble dissociation to some extent in the absence of such reagents (Truman. Clayton and Campbell, 1968). The relationship between sedimentation rate and protein concentration provides an illustra­tion of this. Normally sedimenta­tion rate diminishes as protein concentration rises due to viscosity effects, but when proteins dissoci­ate upon dilution there is also a reduction in sedimentation rate at low concentrations. Both these effects are seen when the sedimen­tation velocity of the γ-crystallin peak is measured in preparations of total crystallins at various con­centrations, (Clayton, 1970). Fur­ther evidence of a spontaneous reversible dissociation comes from studies of the behaviour of β-cry­stallins during gel filtration (Tru­man et al., 1971), while there is also an indication that the subunits may reassociate in different com­binations by the formation of hybrid molecules between the γ-­crystallins of different species of birds (Clayton, 1969).

If dissociation and reaggregation can occur spontaneously in physio­logical saline solutions, the ques­tion arises of whether the crystalline aggregate molecules as we know them may not themselves be the result of changes involved in the isolation of proteins from the lens. Maisel and Perry (Personal com­munication) have shown that in the chick lens there are filamentous aggregates of what are apparently globular units of α-and β-crystal­lins, and that these filaments may retain their structure when isolated by very gentle procedures, where­ as crystallin preparations purified by even so mild a method as gel filtration no longer have this form. This work implies that in the lens there are strong interactions bet­ween protein molecules which must be broken in order to isolate the crystallin fractions which have for so long been the material used by protein chemists and immunologis­ts working on the lens. We must be aware that these crystallin pre­parations may be to some extent artefacts of isolation.

The immunological specificity of crystallins. The question of the organ specificity of the antigens of the lens has been discussed in some detail in an earlier paper (Clayton, Campbell and Truman, 1968. Us­ing a variety of tests and studying tissues from both the chick and from Xenopus laevis we found some cross-reaction with antisera to lens proteins in extracts both of eye tissues and of other organs such as brain, liver and kidney. By means of the Osserman test we were able to show that crystallins in all classes showed cross-reactions with some other tissues. Many of the cross-reactions involved were reactions of partial indentity or reactions with limited portions of the precipitin arcs of the crystal­lins. It seemed from these results that only certain subunits of the crystallins might have antigenic determinants which were common to other tissues.

A somewhat similar situation obtains when we study the species specificity of chick crystallins. Comparison of the chick crystallins with those of the mouse and of Xenopus laevis again show many reactions of partial identity. For example chick α- crystallin showed one antigen which was shared with the amphibian and with the mam­malian species and another antigen which was restricted to the chick. In studies made on isolated sub­units it has been found that some cross react with antisera to other species, while some do not. Six out of eleven subunits of chick β-cry­stallin reacted with antiserum to 4 amphibian lens while two subunits reacted with antiserum to mamma­lian lens (Clayton and Truman. 1968: Clayton, 1969). This limited cross-reaction can explain the pat­terns that we obtain in immuno­electrophoresis when chick crystal­lins are tested with an antiserum to Xenopus laevis lens. We find that the entire length of the precipitin arc of the chick β-cry­stallin reacts with the antiserum to the amphibian lens, but that no spurs are found of the type that occur when homologous antisera are used. We must conclude that the subunits which react show antigenic indentity and that they are distributed over the entire electrophoretic range of the β-cry­stallins.

Subunits and antigens during ontogeny. Since the subunits of chick crystallins appear to consist of single polypeptide chains, while the crystallins are aggregates of a number of different subunits, it follows that if we wish to consider the differentiation of the lens fibre as a system involving the activity of a number of different genes, then it is the synthesis of the in­dividual subunits which must be followed. The value of this appro­ach to the question of cell differ­entiation has been discussed by Clayton (1970). While there is evi­dence from work on bovine α- crys­tallin that the different subunits may not necessarily be the products of different genes since one can be converted to another, possibly by mechanisms such as the deamida­tion of glutamine and asparagine residues (Palmer and Papacenst­antinou, 1969; Delcour and Papa­constantinous, 1970, it is never­theless possible to discover whe­ther such conversions take place by following the kinetics of amino acid incorporation into the subunits. In the chick lens we have as yet no reason for supposing that each subunit is not the product of a se­parate gene. The value of immuno­logical techniques in these studies of differentiation lies in their ability to characterise and discri­minate between the different sub­units, and in certain techinques to actually participate in the separa­tion of individual subunits. The sensitivity and specificity of immunological methods has already been applied to studies of lens development in terms of the syn­thesis of specific subunits (Clayton, 1970; Truman et al., 1972). An example of the indispensabiity of the immunological approach to pro­blems of subunit ontogeny is to be found in the application of immunofluorescence to the studies of the localisation of subunits within the lens (Clayton, 1970).

Immunological methods of the identification of crystallin subunits can also be extended to the isola­tion of those polysomes involved in the synthesis of particular sub­units, making use of the presence of antigenic determinants on the polypeptide chain while it is still being synthesised (Clayton, Tru­man and Campbell, 1970). Such a fractionation technique can be used to yield a preparation of nolysomes with a particular role in protein biosynthesis, which will contain the messenger RNA speci­fic for that polypeptide to which the antibody is directed. We have already begun to apply this method to studies of the stability of the messenger RNA molecules which specify particular crystallin sub­units. and to the problem of the way in which this messenger RNA becomes stabilised during the diff­ereniation of the lens fibre (Clayton, Truman and Campbell, 1972).

We believe that the work which has been done to relate the struc­ture of the crystallins to their immunological properties can throw light on some of the immunological problems of the lens, and that, in turn, the imm­unologists' techniques can then be directed towards the elucidation of some of the molecular problems of cell differentiation.

Acknowledgments . We wish to thank Mrs. A. G. Brown, Mr. A. G. Gillies, Mrs. A. H. Hannah and Mrs. H. J. Mackenzie for their skilled technical assistance. We are grateful to Sterling Poultry Pro­ducts Ltd., and D. B. Marshall (Newbridge) Ltd. for their con­tinuing co-operation in the supply of material for our experiments. This work has been made possible by the financial support of the Medical Research Council, the Cancer Research Campaign and the W. H. Ross Foundation to Combat Blindness[29].


1Bjork, 1., 1964. Exp. Eye Res. 3, 239.
2Bloemendal, H., Bont, W.S., Jongkind, J.F. and Wisse, J.H. 1962 Exp. Eye Res. 1, 300.
3Bon, W.F., Swanborn, P.L. Ruttenberg. G. and Dohrn, A., 1964. Pubbl. staz. zoo]. (Napoli) 34, 59.
4Campbell, J.C. Clayton, R.M. and Truman, D.E.S., 1968. Exp. Eye Res. 7, 4.
5Corran, P.H. and Waley, S.G., 1969. Biochem. J. 115, 789.
6Clayton R.M., 1969. Exp. Eye Res. 8, 326.
7Clayton R.M. 1970. Current Topics in Dev. Biol. 5, 115.
8Clayton R.M. Campbell, J.C. and Truman, D.E S. 1968. Exp. Eye Res. 7, 11.
9Clayton R.M. and Truman D.E.S., 1967. Nature, Lond, 214, 1201.
10Clayton R.M. and Truman D.E.S. 1968. J. Physiol. 198, 72p.
11Clayton R.M. Truman D.E.S. and Campbell J. C. 1970. Biochem. J. 117. 63p.
12Clayton R.M. Truman, D.E.S. and Campbell J. C. 1972. Cell Differentiation 1, in press.
13Delcour, J., and Papaconstantinou, J. (1970). Bicchem. Biophys. Res. Comm. 41, 401.
14Hoenders, H .J. 1965. Uber die Proteinzu­samensetzung der Augenlinse des Huhns THESIS. Amsterdam.
15Hoenders, H.J., 1965. De Groot, K, Gerd­ing., J. J. T. and Bloemendal H. 1969, Biochim, Biophys. Acta 188, 162. Palmer W.G. and Papaconstantinou, J. 1968. Nature, Lond. 219, 170.
16Palmer W.G. and Papaconstantinou, J. 1968. Proc. Natl. Acad. Sci., U.S. 64, 404.
17Rana. M.W. and Maisel, H, 1969. Exp. Eye Res. 8, 216.
18Rana, M.W. and Maisel, H, 1970. Ophthal
19Res 1,156 Schoenmakers, J.G.G. and Bloemendal, H. 1968. Nature. Lond. 220, 790.
20Stark. G R. Ste:n, W.H. and Moore S. 1960. J. biol. Chem. 235, 3177.
21Truman. D.E.S., 1968. Exp. Eye Res. 7, 358.
22Truman, D.E,S., Brown, A.G. and Campbell. J.C., 1972. Exp. Eye Res. 13, 58.
23Truman D.E.S. Brown A.G. and Rae, K.V., 1971, Exp. Eye Res. 12, 304.
24Truman D.E.S. Clayton R.M. and Campbell J.C. 1967. Experimentia 23 502.
25Truman D.E.S. Clayton R.M. and Campbell J.C. 1968. Biochem. J. 108, 9P.
26Wrigley C. 1968. Science Tools 15, 17.
27Zwaan, J. 1963. Immunochemical Analysis of the Eye Lens during Development. Thesis, Amsterdam.
28Zwaan, J., 1968. Exp. Eye Res. 7, 461.
29Zwaan, J. and Ikeda, A.. 1968. Exp. Eye Res. 7, 301.