Intermolecular disulfide bonding of lens membrane proteins during human cataractogenesis

Transcription

1 Intermolecular disulfide bonding of lens membrane proteins during human cataractogenesis L.J. Takemoto andj. S. Hansen Twodimensional diagonal electrophoresis has been used to characterize intermolecular disulfide bonding of membrane proteins from lenses of cataractous and normal patients. A component of approximately 18,000 daltons, linked via intermolecidar disulfide bonding, was found in membrane preparations from 10 of 17 cataracts studied. In comparison, membrane from only 1 of 12 normal lenses of approximately the same age range was found to contain intermolecular disulfide bonding of a component of similar molecular weight. Treatment of normal lens with the oxidizing agents cupric sulfate and 1,10phenanthroline resulted in intermolecular disulfide bonding of the 18K component in a manner similar to that found in cataractous lenses. Together these results demonstrate that human cataractogenesis is many times accompanied by intermolecular disulfide bonding of a membrane component of 18K and. suggest that this intermolecular bonding may be the result of the previously reported oxidative insult of the lens during human cataract formation. (INVEST OPHTHALMOL VIS SCI 22:336342, 1982.) Key words: lens membrane, human cataract, disulfide bonding s the only structural entity separating the intracellular and extracellular spaces, the plasma membrane plays an important role in maintaining normal function in the intact lens fiber cell. In animal model systems, abnormal functioning of membrane components have been implicated in the formation of lens opacities. 1 " 3 Studies of human cataracts have reached less definitive conclusions principally because of the extensive variability in cataract morphology seen in patients of widely From the Division of Biology, Kansas State University, Manhattan, Kan. Supported by grant EY to L. J. T. from the National Institutes of Health. Publication No j from the Agricultural Experiment Station of Kansas State University. Submitted for publication March 23, Reprint requests: L. J. Takemoto, Division of Biology, Kansas State University, Manhattan, Kan varying ages. Nonetheless, among the most consistent findings has been the presence of increased amounts of high molecular weight protein aggregates in cataractous vs. normal lenses. 4 " 9 Some of these aggregates result from intermolecular covalent bonding of both a disulfide and nondisulfide nature 4 " 6 ' 9 and appear to originate at least partially from the increased oxidative phenomena present in the cataractous lens. Because of the obvious importance of the membrane in maintaining normal lens function, we wished to determine whether some of the intermolecular disulfide bonding occuring during cataractogenesis was localized to the fiber cell membrane. Twodimensional diagonal electrophoresis provides a method for determining the molecular weight of the component species involved in this type of bonding. 10 The results of this study demonstrate that human cataractogenesis is accompanied by intermolecular disulfide bonding /82/ S00.70/ Assoc. for Res. in Vis. and Ophthal., Inc.

2 Volume 22 Number 3 Disulfide bonding of cataract membranes 337 of an 18K component to the fiber cell membrane in some, but not all, cataractous lenses. Methods Cataractous lenses were stored at 20 within 30 min of their removal. Classification of these lenses was obtained from each patient's ophthalmologist as previously described." Normal lenses were obtained from eyes donated for corneal transplant and were processed and stored in the same manner as the cataractous lenses. To obtain fiber cell lens membrane, the decapsulated lens was homogenized in 4 ml of 2 mm EDTA, 0.1 mm phenylmethylsulfonylfluoride and 5mM Tris HC1, ph 7.9, containing 50 mm iodoacetamide. Before use, the homogenizing buffer was degassed with a vacuum pump. Immediately after homogenization, nitrogen gas was blown over the suspension. The suspension was then pelleted at 27,000 X g for 15 min in a Sorvall SS34 rotor. The homogenization and pelleting procedure was repeated four more times, and the resulting pellet was resuspended in 4 ml of degassed homogenizing buffer Containing 8M urea and 50 mm iodoacetamide. The solution was vortexed and diluted with an equal volume of homogenizing buffer not containing urea. After pelleting at 27,000 X g for 15 min, the urea wash was repeated and the pellet was washed in 4 ml of 0.01M sodium phosphate, ph 7.4. The final pellet was resuspended with vortexing in 1 ml of 0.01M sodium phosphate buffer and layered on top of a gradient containing 0.5 ml of 45% (w/w) sucrose, 1.5 ml of 35% (w/w) sucrose, and 1.5 ml of 30% (w/w) sucrose. All sucrose solutions contained 0.01M sodium phosphate, ph 7.4. The gradient was spun at 140,000 x g for 1V6 hr in an SW 50.1 rotor. Material at the bottom of the tube and fractions at each of the interfaces were taken for analysis. Twodimensional diagonal electrophoresis using 10% polyacrylamide gel in both dimensions was used according to the method of Takemoto et al. 10 Approximately 100 fig of protein, as determined by the Coomassie Blue assay, 12 was dissolved in sample buffer containing 50 mm IVethylmaleimide. In certain cases, samples were instead dissolved in the same solution containing 2% (v/v) 2mercaptoethanol. The and membrane fractions from both normal and cataractous lenses were completely solubilized in either of these sample buffers. In some cataractous lenses approximately 5% to 25% of the bottom pellet material was not solubilized under the same conditions. During twodimensional analysis the molecular weight of proteins in the second dimension was determined with human erythrocyte membrane proteins as molecular weight markers. 13 For crosslinking studies, normal lenses were decapsulated, refrozen, and sliced into 100 micrometer slices with a handoperated microtome (Spencer Lens Co.). Slices were immediately placed in a 1 ml solution of 0.15M NaCl and 0.01M sodium phosphate (ph 7.4) containing 0.5 mm CuSO 4 and 1.5 mm 1,10phenanthroline (CuP). After incubation at room temperature for 30 min the reaction mixture was diluted with a large excess (20 ml) of icecold 0.15M NaCl and 0.01M sodium phosphate (ph 7.4), followed immediately by pelleting at 27,000 x g for 15 min. Lens membrane was prepared from the resulting pellet as previously described, with the omission of sucrose density centrifugation. Results Homogenization in dilute buffer followed by homogenization in dilute buffer plus 7M to 8M urea has been previously used to obtain membrane from human lens fiber cells. 14 ' 15 We have followed essentially the same protocol but have included sucrose density centrifugation as an additional step to further separate the lens membrane into fractions of different density. On the basis of studies of the chick and bovine lens membrane, the lightest fraction () contains the material with the least amount of cytoplasmic material attached to the lipid bilayer (ref. 16; Takemoto, unpublished data). Generally, similar conclusions were obtained after transmission electron microscopy of cataractous and normal human lens material after fractionation by sucrose density centrifugation (results not shown). However, the and bottom pellet fractions also contained substantial amounts of membrane bilayer material and were therefore analyzed by twodimensional diagonal electrophoresis. This method has been previously used by Takemoto et al. 10 to identify membrane components linked via intermolecular disulfide bonding. In this procedure, proteins are solubilized in sodium dodecyl sulfate and sequentially resolved in two gel systems containing the same concentrations of polyacrylamide. Between the first and second dimen

3 Incest. Ophthalmol. Vis. Sci. March Takemoto and Hansen \ \ \ Fig. 1. Twodimensional diagonal electrophoresis of membrane proteins from normal and cataractous lenses. Membranes were prepared and analyzed as described in Methods, a, fractions from a 61yearold normal lens; b, bottom pellet fraction from a 65yearold cataractous lens with opacities in the nuclear and posterior subcapsular regions; c, same material as in b except preincnbated with 2mercaptoethanoI before electrophoresis. The 26K and 22K components have been designated by horizontal arrows. sions, protein disulfide bonds are reduced by treatment with 2mercaptoethanol. Components with no intermolecular disulfide bonding will migrate with the same relative mobility in both gel systems and will appear in the second dimensional slab as components of a diagonal. Components linked via intermolecular disulfide bonding will be cleaved by treatment with 2mercaptoethanol and will migrate with a greater relative mobility in the second dimension. These components will appear "offdiagonal" in the lower left side of the second dimensional gel. Fig. 1 illustrates representative results of diagonal electrophoresis of various membrane fractions from a cataractous and a normal lens. The normal lens membrane (Fig. 1, a) was characterized by a lack of offdiagonal components indicative of intermolecular disulfide bonding of membrane proteins. Present on the diagonal were the 26K and 22K polypeptides that have been previously identified as intrinsic components of the human lens membrane. 14 When compared with the twodimensional pattern from normal lens (Fig. 1, a), the pattern from cataractous lens showed lower levels of the 26K and 22K intrinsic polypeptides and higher levels of lower molecular weight components (Fig. 1, b). In addition, this latter pattern also showed the presence of offdiagonal complexes, all with a second dimensional molecular weight of approximately 18,000 daltons. Because proteolysis is a welldocumented trait of lens proteins, it is possible that the 18K offdiagonal component shown in Fig. 1, b, was the result of proteolysis that occurred during electrophoresis and subsequent incubation of the first dimensional gel. Two observations argue against this possibility. First, the results shown in Fig. 1, c, demonstrate that the 18K offdiagonal components were lost when the sample shown in Fig. 1, b, was first preincubated with 2mercaptoethanol. Second, incubation of first dimensional gels at room temperature for periods of 30 min to 4 hr did not change the amounts of 18K oftdiagonal components present in the second dimension (results not shown). Tables I and II indicate that oligomeric

4 Volume 22 Number 3 Disulfide bonding of cataract membranes 339 Table I. Presence of the 18K offdiagonal component in cataractous and normal lens membrane* Age (yr) Cataract or normal Gradient fraction Presence ofl8k offdiagonal component NSPSC NSCORT NS NSCORT Normal Normal NS = nuclear cataract; CORT = cortical cataract; PSC = posterior subcapsular cataract; = not determined because of lack of protein. 'Gradient fractions containing 100 /ig of protein were resolved as described in Methods. Any offdiagonal material with a second dimensional molecular weight of 18K was scored as positive.

5 340 Takemoto and Hansen Invest. Ophthalmol. Vis. Sci. March 1982 Table I. cont'd Age (yr) Cataract or normal 52 Normal 54 Normal 61 Normal 61 Normal Normal Normal 64 Normal 65 Normal 66 Normal 66 Normal Gradient fraction Presence oflsk offdiagonal component _ complexes of this 18K offdiagonal component were found in many, but not all, cataractous lenses studied. They were most prevalent in the heavier fractions, being present in most of the and bottom pellet fractions from cataractous lenses. In contrast, analyses of normal human lenses demonstrated that offdiagonal oligomers of the 18K component were not found in any of the 0/30 30/45 fractions and were found in only one of the bottom pellet fractions. Only four of 12 normal lenses provided enough material for twodimensional resolution of the bottom pellet fraction. In contrast, 12 of 17 cataractous lenses provided sufficient bottom pellet fraction for analysis. The apparent greater density of these bottom pellet fractions may be caused by increased associations of proteins such as 18K with the lipid bilayer (compare Fig. 1, a and b). The results shown in. Tables I and IL ;uggest that lens membrane from some, but not all, human cataracts are characterized by an 18K component that is involved in intermolecular disulfide bonding. It is possible that this crosslinking at the lens membrane could be the result of abnormally high levels of oxidizing agents that accompany human cataractogenesis. Implicit in this argument is the assumption that in the normal lens, the 18K component is in the correct nearestneighbor associations for intermolecular disulfide bonding to occur. To test this latter hypothesis, slices from normal lenses were incubated in CuP, a reagent that has been previously used to oxidize nearestneighbor proteins by intermolecular disulfide bonds. 17 The results shown in Fig. 2 demonstrate oxidation of the 18K component to form offdiagonal components in a manner similar to that seen in human cataracts (compare Fig. 1, b, with Fig. 2). Discussion Oxidation of lens proteins has been widely reported in human cataracts. 4 ' 18 ~ 20 Of recent interest have been reports that membrane proteins are involved in this phenome

6 Volume 22 Number 3 Disulfide bonding of cataract membranes Table II. Summary of 18K offdiagonal component in lens membrane from cataractous and normal lenses Lens type Gradient fraction Cataract Normal r 341 1st Presence of 18K offdiagonal component 3/17 6/11 9/12 0/ /3 1/4 non These observations raise the possibility that oxidation of membrane components may disrupt normal cell functioning to an extent sufficient to result in lens opacification. The objectives of this investigation were therefore to ascertain the formation of intermolecular bonding of membrane components during cataractogenesis. At the same time, efforts have been made to identify the membrane proteins involved in this event. We have developed an experimental protocol to investigate the presence of membrane disulfide bonding in individual human lenses. This method provides information concerning the relative frequency of occurrence of the 18K disulfide bonding and eliminates ambiguities involved in studies involving pooled cataractous lenses. Our results indicate that intermolecular disulfide bonding involving the 18K component occurs in many, but not all, cataracts that are characterized by opacification in the nuclear region, plus cortical or posterior subcapsular regions. In addition, sucrose density fractionation of individual cataractous and normal lenses indicates that cataractous lenses contain more of the denser material that pellets to the bottom of the gradient tube. This increased density may result from both covalent and noncovalent associations of cytoplasmic components with the lens membrane. One of these components may be the 18K polypeptide. Our protocol has also been designed to minimize the possibility of lens protein oxidation during membrane preparation. All homogenizations were conducted with buffer degassed by a vacuum pump. Furthermore, iodoacetamide was included in this buffer to Fig. 2. Twodimensional diagonal electrophoresis of membrane from a 61yearold normal human lens after treatment with the CuP reagent {see Methods for details). The 26K and 22K components have been designated by horizontal arrows. alkylate sulfhydryl groups. The almost complete absence of any offdiagonal components in our preparations of normal lens membrane supports the likelihood that our protocol did not produce any artifactual disulfide bonds of an intermolecular nature.. In contrast to the results of Spector et al., 21 our results do not indicate involvement of a disulfidelinked 43K component in human cataractogenesis. This membranebound polypeptide was found as a disulfidelinked component of high molecular weight aggregates in the range of 5 X 106 daltons.21 Even though these high molecular weights are not resolved by our 10% polyacrylamide gel system, it is possible that in vivo formation of such aggregates might be preceded by the formation of lower molecular weight aggregates that would be resolved by our gel system and would therefore appear as offdiagonal components after analysis by twodimensional diagonal electrophoresis. Twodimensional analysis of membrane from normal lenses demonstrated the presence of little, if any, offdiagonal components. This is surprising in view of the presence of extensive intermolecular disulfide bonding in other eukaryotic membrane sys

7 342 Takemoto and Hansen Invest. Ophthalmol. Vis. Sci. March 1982 tems. 10> 23> 24 The maintenance of this state in normal lens is probably dependent on a relatively low level of oxidizing agents. It therefore follows that during human cataractogenesis, the observed increase in the level of oxidizing agents could result in the crosslinking of components that are in the correct juxtaposition for nearestneighbor disulfide linkage. Evidence for such an occurrence comes from crosslinking of lens slices with the CuP oxidizing agent. The 18K polypeptide is crosslinked under these conditions, suggesting that nearestneighbor associations of the 18K component are already preexistent in normal lens. Preliminary peptide mapping studies have suggested that the 18K component is related in amino acid sequence to some of the human soluble crystallins. Further studies involving better resolution and identification of the human crystallins are needed before definitive identification of the 18K component can be made. REFERENCES 1. Kinoshita JH: Mechanism initiating cataract formation. INVEST OPHTHALMOL 13:713, Iwata S and Kinoshita JH: Mechanism of development of hereditary cataract in mice. INVEST OPHTHALMOL 10:504, Kador PF, Fukui HN, Fukushi S, Hernigan HM Jr, and Kinoshita JH: Philly mouse: a new model of hereditary cataract. Exp Eye Res 30:59, Buckingham R: Behavior of reduced proteins from normal and cataractous lenses in highly dissociating media: crosslinked protein in cataractous lenses. Exp Eye Res 14:123, Dilley KJ: The proportion of protein from the normal and cataractous human lens which exists as high molecular weight aggregates in vitro. Exp Eye Res 20:73, Harding J: Disulphide crosslinked protein of high molecular weight in human cataractous lens. Exp Eye Res 17:377, Jedziniak JA, Kinoshita JH, Yates EM, Hocker LO, and Benedek GB: On the presence and mechanism of formation of heavy molecular weight aggregates in human normal and cataractous lenses. Exp Eye Res 15:185, Jedziniak JA, Kinoshita JH, Yates EM, and Benedek GB: The concentration and localization of heavy molecular weight aggregates in aging normal and cataractous human lenses. Exp Eye Res 20:367, Takemoto LJ and Azari P: Isolation and characterization of covalently linked, high molecular weight proteins from human cataractous lens. Exp Eye Res 24:63, Takemoto LJ, Fox CF, Jensen FC, Elder JH, and Lerner RA: Nearestneighbor interactions of the major RNA tumor virus glycoprotein on murine cell surfaces. Proc Natl Acad Sci USA 75:3644, Takemoto LJ and Azari P: Amino acid composition of normal and cataractous human lens proteins. Exp Eye Res 23:1, Bradford MM: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of proteindye binding. Anal Biochem 72:248, Fairbanks G, Steck TL, and Wallach DF: Electrophoretic analysis of the major polypeptides of the human erythrocyte membrane. Biochemistry 10: 2606, Horwitz J, Robertson NP, Wong MM, Zigler JS, and Kinoshita JH: Some properties of lens plasma membrane polypeptides isolated from normal human lenses. Exp Eye Res 28:359, Alcala J, Valentine J, and Maizel H: Human lens fiber cell plasma membranes. I. Isolation, polypeptide composition and changes associated with aging. Exp Eye Res 30:659, Bloemendal H, Zweers A, Vermorken F, Dunia I, and Benedetti EL: The plasma membranes of eye lens fibres. Biochemical and structural characterization. Cell Differ 1:91, Murphy AJ: Crosslinking of the sarcoplasmic reticulum ATPase protein. Biochem Biophys Res Commun 70:160, Zigman S: Eye lens color: formation and function. Science 171:807, Takemoto LJ and Azari P: Amino acid composition of normal and cataractous human lens proteins. Exp Eye Res 23:1, Garner MH and Spector A: Selective oxidation of cysteine and methionine in normal and senile cataractous lenses. Proc Natl Acad Sci USA 77:1274, Spector A, Garner MH, Garner WH, Roy D, Farnsworth P, and Shyne S: An extrinsic membrane polypeptide associated with high molecular weight protein aggregates in human cataract. Science 204: 1323, Garner WH, Garner MH, and Spector A: Gamma crystallin, a major cytoplasmic polypeptide disulfide linked to membrane proteins in human cataract. Biochem Biophys Res Commun 98:439, Wang K and Richards FM: An approach to nearestneighbor analysis of membrane proteins: application to the human erythrocyte membrane of a method employing cleavable crosslinkages. J Biol Chem 249:8005, Takemoto LJ, Miyakawa T, and Fox CF: Analysis of membrane protein topography of Newcastle disease virus and cultured mammalian fibroblasts. In Cell Shape and Surface Architecture. New York, 1972, Alan R. Liss, Inc.