Exp. Eye RES. (1992)

54. 611-619

Progressive Modifications Cataracts Induced HAROLD

I. CALVIN*,

SUBHASH AND

Departments of Ophthalmology

of Mouse by Buthionine A. PATEL, JIA-PING S.-C. JOSEPH FU

Lens Crystallins Sulfoximine ZHANG,

in

MIN-YUAN

and of Biochemistry and Molecular Biology, UMD-NJ Newark, NJ 07103, U.S.A.

LI

Medical School,

(Received Bethesda 2 January 7991 and accepted in revised form 25 July 1991) L-buthionine-S,R-sulfoximine (BSO), a specific inhibitor of GSH biosynthesis, was administered four times daily to mouse pups on post-natal days 7 and 8, inducing initiation of opacification on day 9. The initial progression of the cataract (< 24 hr) was divided into four stages: (1) developing floriform: (2) mature floriform; (3) degenerate floriform; and (4) amorphous translucent cataract. Following this, dense corticonuclear opacities developed within several days. Two-dimensional gel electrophoresis of watersoluble whole lens extracts indicated that the most rapid early cataractous changes, occurring mainly during stage 2, were loss of the two major components of the heavy /J’-crystallin fraction, a 31-kDa basic polypeptide and an acidic component at 27 kDa, concomitant with the appearance of new species at 30 and 2 5 kDa. This was followed by more extensive modification of both CLand P-crystallins during stages 3 and 4 and the appearance of abnormal species at 26, 19 and 18 kDa, which were slightly more acidic than the major normal aA-crystallin polypeptide. The y-crystallin components, relatively unaffected at stage 4, were then lost rapidly as dense opacities ensued. By contrast with the water-soluble fraction, the normal day 9 urea-soluble fraction was deficient in y-crystailin polypeptides and enriched in anodic

componentswhose relative electrophoretic mobilities were similar to those reported previously for phosphorylatedbovineaA-crystallin and severalcytoskeietalpolypeptides.At stage4 of the cataract, the modificationsof normalCLandP-crystallin componentsin the urea-solublefraction paralleledthosein the water-solublefraction, but the productsseenweremorenumerous.In addition,the cytoskeletalproteins were no longerdetectable.Substantialincreasesin lensCa2+that precedeall of the abovechangesin lens polypeptidecompositionsuggestthat Ca2+-activated proteolysismay play a major role in developmentof BSOcataracts. Key words: mouse: lens: crystallins: buthionine sulfoximine; cataract: GSH depletion: 2D-gel electrophoresis

; proteolysis : calpain.

1. Introduction The mammalian crystalline lens possessesa high content of glutathione (GSH), whose concentration decreases during the formation of most types of cataracts (Reddy, 1990). The primary functions of GSH in the lens have not been clearly delineated, but are believed to involve the preservation of protein thiol groups and generalised protection against oxidation (Reddy, 1990). In recent years, the inhibitor of GSH biosynt.hesis,L-buthionine-S,R-sulfoximine (BSO), has been employed to test the response of the lens to depletion of GSH in vivo. At 2 weeks of age in the rat or the mouse, approximately 95 Y0 depletion of lens GSH by BSOis insufficient to produce cataracts (David and Shearer, 19 84 ; Calvin, Medvedovsky and Worgul, 1986). However, large, dense opacities have been induced by this drug in pre-weanling mice (Calvin et al., 1986) and in early post-natal rats and mice (Martensson et al., 1989). Cataract induction by BSO in pre-weanling mice * For correspondence at: Department of Biochemistry and Molecular Biology, UMD-NJ Medical School, 18 5 South Orange Ave., Newark, NJ 07103. U.S.A. 00144835/92/040611+09

$03.00/O

entails rapid deterioration of lens fibers (Calvin et al., 1991). preceded by significant changes in lens mineral content (Calvin et al., 1992). To gain additional information about the biochemical correlates of BSO-induced cortical opacification in preweanling mice, we have administered a seriesof eight injections of BSO on post-natal days 7 and 8 and observed changes in the two-dimensional electrophoretic pattern of lens polypeptides in water-soluble and urea-soluble whole lens extracts during early cataract formation.

2. Materials and Methods Generation of Cataracts Swiss-Webster mice were bred continuously. Births usually occurred between 1800 and 0800 hr. The first morning on which litters were observed was scored as post-natal day 1. Prior to the initiation of treatment at 0900-1000 hr on post-natal day 7 (t,), the size of litters was reduced, if necessary, to eliminate runts and limit the total number of pups to $ 12. Cataracts were generated by four subcutaneous injections of BSO per day on days 7 and 8, administered at intervals 0 1992 Academic

Press Limited

l-l I. CALVIN

612

of 2.5 hr. The dosage of each injection was 4 pmol g-’ body weight (20 ~1 g-l of a 0.2 M solution of BSO, prepared in 0.10 M NaCl). BSO was purchased from Schweizerhall (South Plainfield, NJ). In some litters, pups were left uninjected or administered 0.3 M NaCl (20 ~1 g-l), in order to establish the normal electrophoretic pattern of mouse lens polypeptides in agematched controls. The mice were killed by decapitation at 2, 3 and 7 days after t,. Their eyes were enucleated and placed in 0.02 M Hepes (pH 7*4)-0.13 M saline (HBS). Lenses were removed in HBS by a posterior approach, trimmed free of extraneous tissue and examined under a dissecting microscope. After entrainment of excess medium, the lenses were weighed and stored at - 70°C. Evaluation and Photography of Lenses Prior to weighing, lenses were staged under a dissecting microscope (American Optical, Model AO5 70) by transillumination with diiusely reflected light. The criteria used for staging are described below (see Results and legend to Fig. 3). Photographic records of cataract development were obtained under the above conditions, with the aid of a Polaroid Model MF-10 photomicrographic camera kit and Type 107C film. Preparation of Protein Fractions The water-soluble (WS) fraction was prepared by homogenization of whole lensesin 20-25 volumes of 0.05 M sodium phosphate buffer (pH 6.8)-l mM EDTA (PE). A smaller volume of medium (lo-12 volumes) was employed for the preparation of WS protein from lenses with dense cataracts isolated at 7 days after initiation of BSO treatment. Insoluble material was separated by centrifugation at 17 000 g for 10 min. To prepare the urea-soluble (US) fraction, the PEinsoluble pelletsfrom at least 20 lenseswere suspended in PE buffer (50 ~1 mg-’ original lens wet weight) and recentrifuged. The supernatant, which contained < 8 % of the protein in the WS fraction, was discarded. The washed residue was suspended in 6 M urea-PE (4 ~1 mg-’ original lens weight) and the suspension was centrifuged for 10 min at 17000 g. The supernatant was analysed as the US fraction. Four WS crystallin subfractions (a, &, & y) were prepared from a pool of 40 lensesisolated from day 10 mice’by chromatography of WS lens extracts on Sephadex G200 sf (Ocken et al., 19 77) and concentration of selected fractions by ultrafiltration with Amicon Centriprep-10. The pattern of elution is depicted in Fig. 1 (A). All protein preparations were carried out at approximately 4°C. Electrophoresis The concentration of protein in samplessubmitted

ET AL.

to electrophoresis was determined with bicinchoninic acid (Smith et al., 1985). Two-dimensional (2D) electrophoresis of WS proteins was carried out by procedures similar to those reported previously (Huang, Zhang and Fu, 1990). Standardprocedure. The first dimension consisted of non-equilibrium pH gel electrophoresis (NEPHGE) in a Multiphor horizontal electrophoresis cell (LKB Instruments, Bromma, Sweden). The 0.5-mm slab gels, cast on a Gel-Bond film template, contained 5.2 5% polyacrylamide, 6 M urea and 3 Y03.5-10 Ampholines (LKB). Samples (approximately 40 ,ug protein in 6 M urea-PE) were applied on l-cm applicator squares, positioned 20 mm from the cathodic end of the gel, and were electrophoresed for a total of 47004800 V hr. During the first 45 min of electrophoresis, the gel was subjected to a constant current of 15 mA until the voltage had reached 500 V. The voltage was increased to 1200 V over a 1-hr period and electrophoresis was continued at 1200 V for an additional 3 hr. At the conclusion of the run, the current had decreasedto 6-7 mA. This residual current indicated that the system was not yet in equilibrium. The firstdimension gel was fixed in 10% trichloroacetic acid (TCA)-5 % sulfosalicylic acid (SSA) for 20 min, washed with water, dried and cut into 1.5-cm wide strips. These were incubated for 8 min in 0.1% sodium dodecyl sulfate (SDS)-0.12 5 M Tris-HCI-5 % mercaptoethanol, and were embedded in a layer of 0.8% agarose-O.l ‘% SDS-O.12 5 M Tris-HCl, above l-mm thick, 5 x 8 cm 12.5 % polyacrylamide gels, prior to SDS-polyacrylamide gel electrophoresis (SDS-PAGE) (Huang et al., 1990). Modified procedure. The procedure described above was modified slightly during analysis of US fractions, by: (1) applying the sample to the first-dimension gel 10 mm from the basic end ; and (2) replacing the acid fixation step at the end of the run with a brief water wash (5-10 min) prior to drying. Both modifications were found to enhance detection of acidic cytoskeletal polypeptides. WS protein samples were also electrophoresed in this way, for comparison with their behavior in the standard procedure (above).

3. Results Normal Pattern of Water-soluble(WS) Crgstaflins in Day 9-11 Mouse Lenses The results of 2D electrophoresis (NEPHGE/SDSPAGE) of a WS extract of 20 pooled day 9 mouse lenses are presented in Fig. 2 (A). A total of 19 components have been numbered in order of increasing acidity and, where necessary, in order of decreasing molecular weight. Components l-6 have been assignedto the y-crystallin fraction, on the basis of chromatographic fractionation of day 10 mouse crystallins on Sephadex G-200. The assignment of 7 and 9 is uncertain, although 9 may correspond to

MOUSE

CRYSTALLINS

IN BSO

CATARACTS

613

Fraction

no.

FIG. 1. Resolution and analysis of WS crystallin subfractions. A, El&ion pattern of WS crystallins on Sephadex G-200 sf, showing separation of a, &, /I, and y-crystallins. The central portion of each of the four major peaks was selected for pooling of fractions, as indicated by the vertical lines. B. Electrophoretic analysis of a. showing location of 8 and 18. A minor component, 14, may represent aINS. The figure also shows analyses of /3, (C) and of p,, (D). Arrowhead in (D) indicates faint spot, which may represent component 20 [Fig. 2 (B)]. The applicator square for the first dimension was placed 10 mm from the cathode. Gel was fixed in TCA-SSA prior to preparation for SDS-PAGE. [See Fig. 5 (A) for comparison of pattern and Fig. 2 (A) for identification of individual spots.]

the bovine polypeptide currently designated as ys (Bloemendal, Piatigorsky and Spector, 19 8 9). Numbers 8 and 18 were the major components in the a-crystallin chromatographic fraction [Fig. 1 (B)] and therefore must be the two major a-crystallin polypeptides. aB2 and aA2, respectively (Ramaekers et al., 1982), more currently designated as aB and aA (Bloemendal et al., 1989). A minor component, approximately coincident with 14, was also seen in the a-crystallin chromatographic fraction and may represent aAins (Garber and Gold, 1982). The trace component, 19, was not detected in this chromatographic fraction. However, its position in the ZD-gel, reiative to 18 (aA2), suggestscorrespondence to rat aA (Ramaekers et al., 1982), a phosphorylated derivative of aA (Voorter et al., 1986). The two major components found in the ,&crystallin fraction were 10 and 16, with an additional trace component at 15 [Fig. 1 (C)l. Those present in the P,-crystallin fraction included 11, 12, 13, 15, 16 and 17 [Fig. 1 (D)]. Thus, only 15 and 16 were common to both fractions. The exclusive assignment of the largest P-polypeptide, 10, to PH confirms similar Endings with mature rat and calf lens WS protein fractions (Ramaekers et al., 1982: Huang et al., 1990).

However, both of these previous studies noted a relatively large number of components in PH and a considerable overlap in the contents of the PH and pL fractions, in contrast to the distribution in the day 10 mouse lens (Fig. 1). The systematic designation of P-crystallin polypeptidesin the mouse lens is not yet available. Possible analogies with bovine and rat polypeptides have been deduced on the basis of relative positions on 2D gels. Number 10 is tentatively designated as /3Bla, by analogy with the highest molecular weight basic /3crystallin component in rat and bovine lenses (Ramaekers et al., 1982). The most prominent WS /?crystallin polypeptide in day V-l 1 mouse lenseswas 11. This component may be analogous to PBl b in rat and bovine lenses(Ramaekers et al., 1982). PBlb is believed to be derived from flla by a post-translational proteolytic modification (Ramaekers et al., 1982). A neighboring polypeptide of slightly lower molecular weight and p1, 12, was found to increase rapidly during the next several days (seebelow). This developmental increment is characteristic of /3Bp (or fl2), the principal basic polypeptide in mature mouse lenses (Nakamura et al., 1988). Number 13 is tentatively designated as pB3 (Berbers et al., 1984).

614

FIG. 2. Two-dimensional electrophoretic patterns of normal WS mouselenspolypeptides.Spotsare numberedby an arbitrary system(seetext). The bandsat the left represent the following standardsrun in parallel: phosphorylaseb (97 kDa). bovine serum albumin (67 kDa), ovalbumin (45 kDa), carbonic anhydrase (30 kDa), soybean trypsin inhibitor (21 kDa). and lysozyme (14 kDa). A, Post-natal day 9 pattern, showingat least 19 significantcomponents. B, Post-natalday 11 pattern, showingthe presenceof a new component,20, and increasedconcentration of 12.

Number 15 probably included the acidic P-crystallin designated as mouse p23 (Inana et al., 1983) whose amino acid sequencebears a 9 5 ‘%homology to that of bovine PA1 (Berbers et al., 1984). The mouse equivalent of bovine PA2 has not yet been reported, but may also have been present in 15. Components 16 and 17 may correspond to bovine PA3 and /3A4, respectively, (Berbers et al., 1984). Under an earlier nomenclature, these were designated as PA2 and PA3 (Ramaekers et al, 1982). The rapidly increasing concentration of 12 on postnatal days 9-11 can be appreciated by comparison of a day 11 WS extract [Fig. 2 (B)] with that from a pool of day 9 lenses[Fig. 2 (A)]. In addition, this comparison reveals the presenceof a new component (20) which was barely detectable in the day 9 WS fraction and may be identical with cataract-associated component a [Figs 4(A)-(C)]. Changesin WS Polypeptidesinduced by BSO The earliest changes in lens WS polypeptides were detected between 48 and 72 hr after the commencement of BSO injections on day 7 (t,). These coincided

H. I. CALVIN

ET AL.

with a rapid transition from clear lensesto those with pervasive changes readily visible with the naked eye. Examination of lenses with a dissecting microscope established that altered polypeptide patterns were never derived from lenses that were normal in appearance. Instead, the first detectable abnormalities of polypeptide composition were invariably correlated with abnormal refractile patterns, readily observable with the dissecting microscope. The sequence of opacification was very reproducible and is illustrated in Fig. 3. In stage la (early developing floriform). radial swelling aberrations emanating from the superficial central anterior cortex extended in the direction of the equator [Fig. 3 (A)]. These then bent downwards into the posterior cortex (not shown). The radia1 abberrations increased in number and appeared to fuse [stage 1 b, Fig. 3 (B)]. Eventually, a completed floriform cataract was seen [stage 2, Fig. 3 (C)l. This transformation coincided with decreasein the size of the normal nuclear cataract seen at room temperature. Increasingly amorphous swelling patterns followed, in the absenceof the nuclear cataract [stage 3, degenerate floriform, Fig. 3(D)], until the boundaries of the lens nucleus were no longer apparent and dense opacification was commencing [stage 4, amorphous translucent, Fig. 3 (E)]. The mature cataract, seen at 1 week after initiation of injections [stage 5. Fig. 3 (F)] was a dense corticonuclear opacity with a clear equatorial region. The loci of new or more intense electrophoretic spots seen in successive stages of the cataract have been labeled alphabetically in approximate order of their appearance (Fig. 4). In some cases, the new components appeared as groups of two or more spots. These have been designated by a letter with an opencircle superscript (b”, d”, h”. i” j”). In general, little or no change in the electrophoretic pattern of WS crystallins was detected in lenses characterized as stage 1. The earliest readily detectable change was the appearance of an abnormal spot (a) at approximately 25 kDa, possibly equivalent to normal trace component 20 [cf. Fig. 2 (B)]. This abnormality, first discernible late in stage 1, is illustrated in Fig. 4 (A). Meanwhile, a second abnormal species(b) was barely detectable (arrow). In stage 2 cataracts, the following additional changes appeared [Fig. 4 (B)] : reduction of 16 and 10, marked increase of a, b, and 7 (labeled here as abnormal species c). and initial appearance of d. During stage 3 [Fig. 4(C)], 10 and 16 virtually disappeared. 8, 17 and 18 decreased, N, d, e. f, and g became more prominent, and a new series of spots appeared (h”). Degradation of 18 to d was now quite strongly indicated, and b appeared to undergo further degradation (redesignated as group b”). During stage 4 [Figs 4(D) and (E)], the above modifications of 18 and b continued, proceeding nearly to completion. Additional absorption was seen at e, f”. g and h”, and new spotsappeared (i, j, k. 1). Component j appeared to

MOUSE

CRYSTALLINS

IN BSO

CATARACTS

615

FIG. 3. Whole lens photographs, illustrating progression of BSO cataract. A. Stage la, cataractous lens (anterior view), showing the normal reversible nuclear cataract seen at room temperature and two spatulate radial aberrations in an otherwise normally refractile cortex. These defects were initiated in the anterior cortex and were not correlated with detectable abnormalities of crystallin polypeptide composition. B, Stage 1 b. anterior view, with larger and more pronounced disturbances. Arrowhead at left indicates barely detectable border of radial aberration. Two arrowheads at right define boundaries of a large sector of floriform swelling, which signaled the initial appearance of detectable polypeptide degradation. C, Stage 2. anterior view, illustrating well-defined floriform cortical cataract. Reversible nuclear cataract was smaller than in stage 1, and perinuclear demarcation was present. Posterior view of this lens presented a similar pattern (not shown). D, Stage 3. posterior view, depicting breakdown of the floriform pattern, disappearance of nuclear cataract. Individual cloud-like zones were still apparent. E. Late stage 4 lens, anterior view, isolated 72 hr after initial BSO injection, displaying amorphous pattern of light scattering and incipient dense opacification. F, Stage 5 lens, side view, anterior surface facing upwards. Mature BSO cataract, isolated 7 days after initiation of experiment, displaying dense corticonuclear opacity with clear zone at equator. x 15.

be similar in pI to 19, but was of slightly lower molecular weight [cf. Figs 2 (A) and 4 (D)]. Meanwhile, 8, which was very low in early stage 4 [t, plus 2 days, Fig. 4(D)], was relatively more prominent by the next day [t, plus 3 days, Fig. 4 (E)], and c became undetectable. The doublet, j”, may reflect further modification of j as stage 4 progressed. The electrophoretic pattern of WS protein obtained

Urea-soluble (US) fractions of normal and cataractous day 9 lenses were examined electrophoretically by a modified procedure, in which the sample for pH gel electrophoresis was applied 10 mm rather than

the detection of acidic cytoskeletal proteins. The consequences of modifying the normal procedure upon resolution of normal WS components can be appreciated by comparison of Fig. 5 (A) with Fig. 2 (A). The major effect of applying the sample closer to the cathode was retarded migration of basic species 14 and 10 ; this resulted in trailing of these components and ambiguous elution patterns (arrows). Nevertheless, neutral and acidic components focused very well under these conditions. The influence of sample positioning on the electrophoretic behavior of basic rat crystallin polypeptides has been noted previously (Huang et al., 1990). The results of NEPHGE/SDS-PAGE performed on a pooled US fraction of normal day 9 mouse lenses are shown in Fig. S(B). By comparison with the WS fraction [Fig. 5 (A)], the US fraction appeared to consist mainly of a: and ,&crystallin polypeptides, with a relative deficiency of y-crystallins. In addition, the US fraction possessed a prominent spot apparently coincident with 19 [cf. Fig. 2 (A)], which was only a trace component in the WS fraction, suggesting that 19 is essentially confined to the US fraction in the day 9 mouse lens. The presence of other components below 45 kDa that were not seen in the WS fraction is

20 mm from the basic end of the gel and acid fixation

indicated. Prominent among these were possible

was omitted at the end of the run. Normal and cataractous WS fractions were treated similarly, as controls for the altered procedure, which enhanced

charge isomers of 16 and 17 (16’ and 17’). In addition, the US fraction contained a very high concentration of component 13. as well as higher

from lenses with dense mature cataracts

[stage 5, Figs

3 (F) and 4 (F)] suggests further modification of a and /3-crystallins, increased prominence of anodic species, and nearly complete loss of y-crystallins. Two of the abnormal components seen in Fig. 4(D), d and j. possibly both derived from 18, were replaced by groups designated as d” and j”. The relative concentrations of these and of e, f” g, h”, and 1 were very high. In addition, 8, 12, 15 and 18 were now the major surviving normal crystallin components. Changes in US Polypeptides

616

H. I. CALVIN

ET AL.

FIG. 4. Progression of BSO-induced modifications in electrophoretic pattern of WS lens protein fraction. Lenses were removed at t, plus 2 days, unless otherwise stated. A, Initial changes in late stage 1 cataract, showing abnormal spot at position [I as well as a trace abnormality at b (arrow). B. Stage 2 cataract pattern, in which the two abnormal spots first detected in late stage 1, a and b. are now prominent, coincident with marked decreases in 16 and 10. New or relatively increased staining is also evident at c and just below 18 (d). Trace amounts of e, f and g are present. C, Stage 3 pattern, showing marked degradation of 18 to d, prominent spotsat e,f and g, and trace componentsat h”. D, Early stage4 pattern. Increasedprominenceof e.J g, decreasein c, circled cluster of abnormalspots(ho). new spotsi and j and a trace componentat I (arrow). E, Late stage4 pattern (to plus 3 days). All of the new abnormalitiesseenin early stage4 are now more prominent and new spots,k and I, are now evident. F, Mature stage5 cataract (to plus 7 days). Preponderanceof componentsof lower M, and lower PI, aswell as lossof y-crystallin components. molecular weight polypeptides which were absent in the WS fraction. The latter included a group of acidic proteins at approximately 45-55 kDa (circled area) known to be of cytoskeletal origin (Voorter et al., 1986). All other novel components have been labeled with arrows. In general, the US fraction tended to be relatively enriched in components of lower p1. The lossesof a: and B-crystallin polypeptides seenin the WS fraction of early stage 4 cataracts [Figs 4(D) and S(C)] were also evident in the US fraction [Fig.

5(D)]. Thus, by early stage 4, the US fraction was entirely deficient in ,+polypeptides 10 and 16 and had lost most of the major normal crA-crystallin component, 18, apparently by conversion to d”. Other abnormal species included relatively low concentrations of a and b”, a high concentration offD and a doublet at about 30-31 kDa (m”) not seen in the WS fraction. The abnormal speciesin the US fraction were more numerous than those in the corresponding WS fraction and tended to occur in clusters. Also note-

MOUSE

CRYSTALLINS

IN

BSO

CATARACTS

617

@-@ID I 2D

c

FIG. 5. Electrophoretic patterns of normal and cataractous day 9 lens WS and US extracts. Procedure for electrophoresis involved omission of acid fixation of first-dimension gel and more cathodic positioning of samples, as described in text. A, Normal WS pattern. Basic polypeptides 14 and 10 did not focus well under these conditions [cf. Fig. 2 (A)]. Arrows indicate the presence of either streaking artifacts or possible additional components not evident in Fig. 2(A). In the absence of acid fixation, components 9 and 12 [cf. Fig. 2 (A)] are barely visible here and 13 and 19 are absent. B, Normal US pattern, showing deficiency of y-crystallins. Components 13 and 19, 16’ and 17’ are prominent in US (B), although undetectable in WS (A). Arrows indicate other new or enriched spots in US, relative to WS pattern. Cluster of putative cytoskeletal components is encircled. C, Early stage 4 WS pattern (t, plus 2 days). D. Corresponding US pattern. Loss or decrease of normal polypeptides in (D) resembles that seen in (C), but the relative quantities off (or f”) and a are reversed and more degradation products are evident in (D), including a prominent doublet at 30 kDa (m’) and multiple modifications of 17 and 18. worthy was the absence of the acidic cytoskeletal proteins in the stage 4 US fraction [cf. circled area in Fig. 5 (B)]. These proteins, which include vimentin and actin (Voorter et al., 1986), were approximately normal in content during stage 1 (data not shown), became increasingly difficult to detect during stages 2 and 3, and had totally disappeared by early stage 4. 4. Discussion Cataract induced by BSO in l-week-old mice is an outstanding example of a rapid-onset experimental cataract. Initial refractile changes in the lens were observed with a dissecting microscope at approximately 48 hr after t, [stage 1 a, Fig. 3(A)]. These proceeded rapidly [Figs 3 (B)-(D)] and resulted in a large, diffuse opacity within the next 24 hr [Fig. 3 (E)]. The first barely detectable modifications of crystallin polypeptides [Fig. 4 (A)] occurred in lenses with readily observable refractile aberrations [stage 1 b, Fig. 3 (B)]. Evidence reported elsewhere (Calvin et al., 1992)

indicates that initial changes in lens transparency and crystallin polypeptide composition in BSO cataract are preceded by increases in lens Na+ and Ca2+ and decreases in K+. Elevation of Ca*+, although less consistent than that of Nat in late pre-cataractous lenses (stage 0), is remarkable in stage 1 cataractous lenses, whose Ca2+ content is typically increased by ten-fold. These high Ca2+ levels and the observed alterations in 2D electrophoretic patterns which followed them during stages 24 may be related. Calpain II, a Ca”+activated protease, is believed to be implicated in opacification in a variety of cataracts (Shearer et al., 1991). The proximity of the new spots to normal components of slightly higher molecular weight that decreased as the new spots developed [Figs 4(B) and (C)] suggests strongly that limited proteolysis was occurring in the crystallins of lenses with BSO cataracts. On this assumption, b (30 kDa) could have been derived from 10 (3 1 kDa), d (19 kDa) from 18 (20 kDa), and a (25 kDa) from 16 (27 kDa). These ERR54

618

proposed modifications each resemble specific examples of limited proteolytic degradations of rat crystallins catalysed by calpain II (David and Shearer, 1986; David, Dickey and Shearer, 1987). The report that aA [Fig. 2(A), S)] is also a substrate for degradation by calpain (Yoshida, Murachi and Tsukahara, 1985) may be relevant to the decrease in this polypeptide seen in Figs 4(C) and (D). During incubations with rat lens protein fractions, calpain II has been found to degrade selectively p- and acrystallin components, while sparing those in the yfraction (David and Shearer, 1986). This would be consistent with the observation that y-crystallins appear at first relatively unaffected by the cataract during stages 14 [Fig. 4(A)-(E)], but then decrease remarkably as the mature, dense cataract forms [Fig. 4(F)], possibly as a result of protein leakage from the lens (Piatigorsky, Fukui and Kinoshita. 1978). Since other proteolytic activities in lens could be responsible for degradation of crystallin polypeptides in cataracts, further studies showing that calpain II can transform selected purified mouse polypeptides to their proposed degradation products are needed, in order to confirm hypothetical reactions catalysed by this enzyme during BSO cataract development. In addition, it would be desirable to demonstrate immunochemically or by enzymic digestion that the putative precursors and products are closely related. The early loss of polypeptides 10 and 16 in BSO cataracts may be analogous to changes observed in three other rapidly developing cortical cataracts : hereditary Nakano mouse cataract (Kobayashi, Kasuya and Itoi, 1989), rat galactose cataract (Huang et al., 1990) and in vitro mouse hypoglycemic cataract (Hu, Russell and Kinoshita, 1982). Gel chromatography on Sephadex G200 has revealed that these two components are the major constituents of the /3”crystallin fraction in the day 10 mouse lens [Fig. 1 (C)l. Previous investigations have demonstrated loss of a heavy /$crystallin fraction in rat galactose (Fu et al., 1980), CatFRAsER mouse (Garber, Stirk and Gold, 19 8 3) and in vitro mouse hypoglycemic cataracts (Hu et al., 1982), suggesting that this type of crystallin complex is especially labile. In addition to proteolysis, the results in Fig. 4 suggest post-translational charge modification of polypeptides, possibly by phosphorylation. This is indicated by the appearance of new species more acidic than 18. The position of these acidic components in the pH gradient was approximately the same as that of 19 [cf. Fig. 2 (A)]. Such anodic poiypeptides have been observed by others in cataractous lenses (Kobayashi et al., 1989; Huang et al., 1990). The concerted appearance of the discrete species at i and j was a consistent phenomenon, anticipated by development of a new spot at d and decrease of 18 [Figs 4(B)-(E)]. This reproducible and rapid sequence implies that the changes were brought about by enzymatic reactions (e.g. proteolysis and phosphorylation), rather than

l-l. I. CALVIN

ET AL.

spontaneous non-enzymatic transformations (e.g. oxidations). By analogy with the known enzymatic phosphorylation of the major aA-crystallin polypeptide to a product of slightly more acidic p1 in rats and mice (Spector et al., 1985: Voorter et al., 1986), it may be postulated that 18 was converted to 19 in this way in normal mouse lenses and that, similarly, tl gave rise to j in those with cataracts [cf. Fig. 2 (A) and Fig. 4(D)]. On the other hand, the data presently available in normal lenses (Kleiman et al.. 1988) do not provide as clear a precedent for the origin of component i by phosphorylation. If abnormal phosphorylations of proteins did occur in the cataractous lenses, this poses the interesting question of whether a high Ca’+ environment could have favored these transformations. Normally, phosphorylation of crystallins appears to be regulated by CAMP rather than Ca2+ (Spector et al., 198 5 : Voorter et al., 1986). However, phosphorylation of membrane proteins in lens is Ca2+ and phospholipid-dependent (i.e. catalysed by protein kinase C) (Lampe et al.. 1986). In the context of a massive wave of fiber cell lysis in the BSO cataract (Calvin et al., 1991). membrane-bound protein kinase C activity could easily have been released into close contact with crystallins. Since it is also well-known that proteolysis by calpain can activate protein kinase C (Kajikawa et al., 1983). a possible link between proteolytic events in cataractous lenses and subsequent novel phosphorylations of crystallins is suggested. The destruction of lens fiber cytoarchitecture in BSO cataracts is completed within a span of approximately 24 hr (Calvin et al., 1991), and thus it is not surprising that major higher molecular weight cytoskeletal polypeptides are undetectable in the US fraction of early stage 4 cataracts [Fig. 5(D)]. Most noticeable is the disappearance of acidic polypeptides at about 50 kDa that are readily detectable in the normal US fraction. These may include vimentin (Voorter et al., 1986), the intermediate filament protein shown to be susceptible to Ca2+-activated proteolysis in lens (Roy, Chiesa and Spector, 1983: Ireland and Maisel, 1984). Recently, it has been reported that incubation of rat lenses for 14 hr with 5 mm Ca2+ in the presence of a sulfhydryl reagent completely degrades spectrin, vimentin and an unidentified 1 lo-kDa polypeptide (Truscott et al., 1990). Increase in the relative size of the IJS fraction was also noted in this system, a phenomenon associated with other cataracts in which Ca”+ is known to be elevated (David et al., 1987: Kobayashi et al., 1989). The 2D electrophoretic procedures described above have provided a sensitive and precise evaluation of early polypeptide modifications during BSO cataract development. However, none of the polypeptide aberrations in WS and US lens fractions were detected prior to changes in lens transparency and electrolyte composition. Subtle alterations of lens membrane proteins localized in the urea-insoluble (UI) fraction

MOUSE

CRYSTALLINS

IN BSO

CATARACTS

may play a role in generating these earlier symptoms of BSO-induced pathology. Since BSO is a specific inhibitor of GSH biosynthesis, it is also reasonable to assume that BSO cataracts are caused by GSH depletion and may therefore depend on alteration of the normal thiol/disulfide status of lens proteins, as well as generalized oxidative disturbances. These possibilities will be the subject of further investigations on the role of protein changes in BSO-induced cataracts.

Acknowledgements This work was supported in part by NIH research grants EY-07355 (H.C.) and EY-01156 (S.X.J.F.), a grant from Research to Prevent Blindness, Inc. to the Department of Ophthalmology, UMD-NJ Medical School, and by NIH biomedical research support grant RR-05393 to UMD-NJ Medical School. The authors wish to thank Mr Werner Broede for preparation of photographs.

References Berbers, G. A. M.. Hoekman, W. A., Bloemendal, H.. defong, W. W., Kleinschmidt, T. and Braunitzer, G. (1984). Homology between the primary structures of the major bovine &crystallin chains. Eur. J. Biochem. 139, 467-75. Bloemendal, H., Piatigorsky, J. and Spector, A. (1989). Recommendations for crystallin nomenclature. Exp. Eye Res. 48, 465-6. Calvin, H. I., Medvedovsky, C., David, J. C., Broglio, T. M., Hess, J. L., Fu, S.-C. J. and Worgul, B. V. (1991). Rapid deterioration of lens fibers in GSH-depleted mouse pups. Invest. Ophthafmol. Vis. Sci. 32, 1916-24. Calvin, H. I., Medvedovsky, C. and Worgul, B. V. (1986). Near-total glutathione depletion and age-specific cataracts induced by buthionine sulfoximine in mice. Science 233, 553-5. Calvin. H. I., Von Hagen, S., Hess, J. L., Patel, S. A. and Fu, S.-C. J. (1992). Lens GSH depletion and electrolyte changes preceding cataracts induced by buthionine sulfoximine in suckling mice. Exp. Eye Res. 54, 621-6. David, L. L., Dickey. B. M. and Shearer, T. R. (1987). Origin of urea-soluble protein in the selenite cataract. Invest. Ophthaimol. Vis. Sci. 28, 1148-56. David, L. L. and Shearer, T. R. (1984). State of sulfhydryl in selenite cataract. Toxicol. Appl. Pharmacol. 74, 109-l 5. David, L. L. and Shearer, T. R. (1986). Purification of calpain II from rat lens and determination of endogenous substrates. Exp. Eye Res. 42, 227-38. Fu, S.-C. J., Wagner, B. J., Su, S. W. and Lai, C. Y. (1980). Characterization of lens proteins. III, A group of labile lens proteins involved in cataracts. Exp. Eye Res. 30, 455-67. Garber, A. T. and Gold, R. J. M. (1982). Comparative twodimensional electrophoretic analysis of water-soluble proteins from bovine and murine lenses. Exp. Eye Res. 35, 585-96. Garber, A. T., Stirk, L. and Gold, R. J. M. (1983). Abnormalities of crystallins in the lens of the CatFRAsER mouse. Exp. Eye Res. 36. 165-9. Hu. T., Russell, P. and Kinoshita. J. H. (1982). In vitro incubation paralleling changes occurring during mouse cataract formation. Exp. Eye Res. 35, 521-33. Huang, W.-Q., Zhang, J.-P. and Fu, S.-C. (1990). Differential

619

effects of galactose-induced cataractogenesis on the soluble crystallins of rat lens. Exp. Eye Res. 51, 79-8 5. Inana, G., Piatigorsky. J., Norman, B., Slingsby. C. and Blundell, T. (1983). Gene and protein structure of a pcrystallin polypeptide in murine lens: relationship of exons and structural motifs. Nature. 302, 310-l 5. Ireland, M. and Maisel, H. (1984). Evidence for a calcium activated protease specific for lens intermediate filaments. Curr. Eye Res. 3, 423-9. Kajikawa. N., Kishimoto. A., Shiota, M. and Nishizuka, Y. (1983). Ca2+-dependent neutral protease and proteolytic activation of Ca2+ -activated. phospholipid-dependent protein kinase.Methods Enzymol. 102, 279-90. Kleiman, N. S., Chiesa, R., Gawinowicz, M. A. and Spector, A. (1988). Phosphorylation of P-crystallin B2 (@p) in the bovine lens. J. Biol. Chem. 263, 14978-83. Kobayashi, S.. Kasuya, M. and Itoi, M. (1989). Changes in lens proteins induced at the early stage of cataractogenesis in cat (Nakano) mice. Exp. Eye Res. 49, 555-9. Lampe, P. D., Bazzi, M. D., Nelsestuen, G. L. and Johnson, R. G. (1986). Phosphorylation of lens intrinsic membrane proteins by protein kinase C. Exp. Eye Res. 45, 2 15-25. Martensson, J.. Steinherz. R.. Jain, A. and Meister, A. (1989). Glutathione ester prevents buthionine sulfoximineinduced cataracts and lens epithelial cell damage. Proc. Natl. Acad. Sci. U.S.A. 86, 8727-31. Nakamura. M., Russell, P., Carper, D. A., Inana, G. and Kinoshita, J. H. (1988). Alteration of a developmentally regulated, heat-stable polypeptide in the lens of the Philly mouse. 1. Biol. Chem. 263, 19218-21. Ocken, P. R.. Fu, S.-C. J., Hart, R.. White, J. H., Wagner. B. J. and Lewis, K. E. (1977). Characterization of lens proteins. I. Identification of additional soluble fractions in rat lenses. Exp. Eye Res. 24, 355-67. Piatigorsky, J., Fukui, H. N. and Kinoshita. J. H. (1978). Differential metabolism and leakage of protein in an inherited cataract and a normal lens cultured with ouabain. Nature 274. 558-62. Ramaekers. F., Dodemont. H., Vorstenbosch, P. and Bloemendal, H. (1982). Classification of rat lens crystallins and identification of proteins encoded by rat lens mRNA. Eur. J. Biochem. 128, 503-8. Reddy, V. (1990). Glutathione and its function in the lens an overview. Exp. Eye Res. 50, 771-8. Roy, D., Chiesa, R. and Spector, A. (1983). Lens calcium activated proteinase: degradation of vimentin. Biochem. Biophys. Res. Commun. 116, 204-9. Shearer, T. R., Azuma, M., David, L. L. and Murachi, T. (1991). Amelioration of cataracts and proteolysis in cultured lenses by cysteine protease inhibitor E64. Invest. Ophthalmol. Vis. Sci. 32, 533-40. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K.. Gartner. F. H., Provenzano, M. D., Fujimoto, E. K., Goeke. N. M., Olson, B. J. and Klenk, D. C. (198 5). Measurement of protein using bicinchoninic acid. Anal. Biochem. 150, 76-85. Spector, A., Chiesa. R.. Sredy, J. and Garner, W. H. (1985). CAMP-dependent phosphorylation of bovine lens alpha crystallin. Proc. Natl. Acad. Sci. U.S.A. 82, 4712-16. Truscott, R. J. W.. Marcantonio, J. M., Tomlinson. J. and Duncan, G. (1990). Calcium-induced opacification and proteolysis in the intact rat lens. Invest. Ophthalmol. Vis. Sci. 31, 2405-11. Voorter, C. E. M., Mulders. J. W. M., Bloemendal, H. and De Jong. W. W. (1986). Some aspects of the phosphorylation of a-crystallin A. Eur. J. Biochem. 160, 203-10. Yoshida, H., Murachi, T. and Tsukahara, I. (1985). Agerelated changes of calpain II and a-crystallin in the lens of hereditary cataract (Nakano) mouse. Curr. Eye Res. 4, 983-8.