identification of origin of two polypeptides of 4 and 5 kd isolated

Identification of Origin of Two Polypeptidesof 4 and 5 kD Isolated from Human Lenses

Om P. Srivastava*^ K. Srivastava,* and C. Silney*

Purpose. To purify crystallin fragments (degraded polypeptides molecular weight < 18 kD) andidentify their parent crystallins.

Methods. The purification of polypeptides with apparent molecular weights of 4 and 5 kD wascarried out using three sequential steps: Sephadex G-50 chromatography under denaturingconditions, preparative sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and high-performance liquid chromatography using a C-18 column. The parent crystallins of the twopolypeptides were identified by the Western blotting method using polyclonal antibodiesraised against individual 4 and 5 kD polypeptides and by comparing N-terminal amino acidsequences of the polypeptides with crystallins.

Results. Two polypeptides of 4 and 5 kD were purified by the three sequential steps as de-scribed from water-soluble proteins of lenses from 60-80-year-old donors. Both purified poly-peptides showed a single major peak during high-performance liquid chromatography on aC-18 column and also a single band during sodium dodecyl sulfate-polyacrylamide gel electro-phoresis.

The Western blot analyses showed maximum immunoreactivity of the anti-4 kD polypep-tide antibody to a 22 kD species of /8-crystallin, whereas the anti-5 kD polypeptide antibodyshowed maximum reactivity to only the aB crystallin. These results were further confirmedduring comparison of the N-terminal amino acid sequences of the two polypeptides withcrystallins. Such comparison showed that the 4 kD polypeptide originated from j8A3/Al crys-tallin after cleavage at His187-His,88 bond. Further, the 5 kD polypeptide was a fragment of aBcrystallin that originated after cleavage at Val145-Asn146 bond.

Conclusion. These results showed that specific bonds of j8A3/Al and aB crystallins are post-translationally cleaved in vivo to produce 4 kD and 5 kD polypeptides, respectively. InvestOphthalmol Vis Sci. 1994;35:207-214

.Lens contains several posttranslationally derivedcrystallin fragments, ie, degraded polypeptides. Themajor polypeptides with molecular weights of 9-10 kDseem to derive from a, |8, and/or 7-crystallins.1"6 Al-though it is presently unknown if these polypeptidesderive as a result of an enzymatic (proteolytic) or a non-enzymatic (Fenton reaction),7 process or both, thepolypeptides seem to exist in normal lenses of allspecies of different ages and also in cataractouslenses.1>8~12

From the *Missouri Lions Eye Research Foundation, and the ̂ Department ofOphthalmology, University of Missouri, Columbia, Missouri.Supported by grants from Retirement Research Foundation, Inc. and NIH grantEY06400.Submitted for publication: December 22, 1992; revised June 1, 1993; accepted June•7, 1993.Proprietary interest category: N.Reprint requests: Om P. Srivastava, Department of Physiological Optics, VisionScience Research Center, Worrell Building, University of Alabama, 924 S. 18thStreet, Birmingham, AL 35294-4390.

The degraded polypeptides seem to play an im-portant role in the heavy molecular weight proteinformation by virtue of their property of cross-linkingvia nondisulfide bonding10 and cleavage.11 Such ag-gregation and cross-linking of these polypeptides havebeen observed in vitro and immunologic results sug-gest that such cross-linked species also exist in increas-ing quantities in vivo with aging.10

In an attempt to identify the parent crystallins ofdegraded polypeptides that may be involved in the invivo cross-linking process and characterize the proper-ties of the individual posttranslationally derived de-graded polypeptides, we undertook to isolate, purify,and determine the parent crystallins of the major de-graded polypeptides. We previously identified 9 kD7D-crystallin and 8 kD 7S-crystallin fragments in hu-man lenses.513 In this study we purified two polypep-tides of 4 and 5 kD from human lenses and identifiedtheir parent crystallins.

Investigative Ophthalmology & Visual Science, January 1994, Vol. 35, No. 1Copyright © Association for Research in Vision and Ophthalmology 207

Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933177/ on 04/07/2018

208 Investigative Ophthalmology & Visual Science, January 1994, Vol. 35, No. 1

MATERIALS AND METHODS

Materials

Normal human lenses with no apparent opacity from60-80-year-old donors were obtained from the LionsEye Tissue Bank of the Missouri Lions Eye ResearchFoundation. The lenses were retrieved within 24 hoursof death and stored in Medium-199 without Phenolred at -20°C until used.

The prestained and unstained protein molecularweight markers were from Bethesda Research Labora-tory (Grand Island, NY) and Pharmacia Fine Chemi-cals (Piscataway, NJ), respectively. The sources ofvarious additional materials are noted throughout thetext. Unless otherwise indicated, all other chemicalsused in this study were from Sigma Chemical Com-pany (St. Louis, MO). The polyclonal and-/? A 3/A1-crystallin antibodies generated to N (residues no. 37-68) and C (residues no. 204-215) terminal regionswere donated by Drs. J. N. Hope and J. F. Hejtmancikof National Institute of Eye, Bethesda, Maryland.

Purification of the 4 and 5 kD PolypeptidesThe 4 and 5 kD polypeptides were purified fromwater-soluble protein fraction isolated from humanlenses of 60-80-year-old-donors. The lens water-solu-ble protein fraction was prepared as described previ-ously.10 The water-soluble proteins were freeze-dried,dissolved in buffer A (50 mM Tris-HCl, pH 7.9, con-taining 7 M urea and 5 mM /?-mercaptoethanol). Thisprotein sample was subjected to a Sephadex G-50(Pharmacia) column chromatography (two sequen-tially attached columns, each 2.5 X 70 cm) and elutedwith buffer A at a rate of 1 ml/min. The column frac-tions were analyzed by sodium dodecyl sulfate-poly-acrylamide gel electrophoresis (SDS-PAGE) for poly-peptides, using 15% acrylamide gels, and two fractionscontaining primarily either 4 or 5 kD polypeptideswere pooled separately, dialyzed against 0.05 M Tris-HCl, pH 7.9, for 72 hours and then concentrated bylyophilization. The 4 kD polypeptide containing frac-tion was subjected to a preparative SDS-PAGE (15%acrylamide gel) using the BioRad (Hercules, CA) PrepCell (Model 491) by the Laemmli14 method. Variouseluted proteins were collected in different fractions ontheir exit from the gel. The fractions containing the 4kD polypeptide were identified by SDS-PAGE, pooled,dialyzed against 4000 volume excess of 0.05 M Tris,pH 7.9, at 4°C and concentrated by lyophilization anddialyzed again as described earlier. The 4 kD polypep-tide was further purified by high-performance liquidchromatography (HPLC) using a 0.8 X 25 cm C-18column (Whatman, Clifton, NJ). The elution wascarried out by a gradient of 0 to 100% acetonitrilecontaining 0.1% trifluoroacetic acid. The 5 kD poly-

peptide was further purified as described for the 4 kDpolypeptide; ie, by the preparative SDS-PAGE methodusing the BioRad Prep Cell followed by HPLC using aC-18 column. During all dialysis steps, the 3500 molec-ular weight cutoff dialysis tubing (Spectrum, Gardena,CA) was used.

Determination of Purity of the 4 and 5 kDPolypeptides

The purity of the 4 and 5 kD polypeptides were testedby an HPLC method using a C-18 HPLC column(Whatman). The polypeptides were eluted with a gra-dient of 0 to 100% acetonitrile containing 0.1% tri-fluoroacetic acid.

Determination of Parent Crystallins of the 4and 5 kD Polypeptides

The parent crystallins of each of the polypeptides wereidentified by two methods: by immunologic Westernblotting method and by comparing N-terminal aminoacid sequences of polypeptides and crystallins.

Individual polyclonal antibodies to the purified 4or 5 kD polypeptides were raised in rabbits as de-scribed previously.10 The Western blot analysis wasperformed by the procedure of Towbin, Stahelin, andGordon.15 The antibody was used at a dilution of1:100. The immunoreactive proteins were visualizedwith a second antibody, anti-rabbit immunoglobulin G(raised in goat) conjugated to peroxidase. As a control,various proteins were treated with rabbit preimmuneserum and then reacted with the second antibody inthe Western blot method. The purified 4 and 5 kDpolypeptides were subjected to N-terminal amino acidsequence analysis at the protein core facilities of theUniversity of Missouri, Columbia. The deduced se-quences of each polypeptide were compared to re-ported sequences of various crystallins.

Miscellaneous Methods

Protein was determined by a modified Lowrymethod.16 SDS-PAGE was performed by the methodof Laemmli.14

RESULTS

Purification of 4 and 5 kD Polypeptides

The polypeptides were purified from water-solubleprotein fractions in three sequential steps: SephadexG-50 chromatography, preparative SDS-PAGE separa-tion, and HPLC using a C-18 column.

When the water-soluble proteins, isolated fromlenses of 60-80-year old donors, were fractionated


Polypeptides from Human Lenses 209

through a Sephadex G-50 column under denaturingconditions, some of the degraded polypeptides wereeluted in separate fractions away from crystallins.During such chromatography, several polypeptidesremained associated with crystallins. However, threepolypeptides of 4, 5, and 9 kD (identified by arrows inFig. 1) were found in fractions that contained verylittle amounts of crystallins. A 4 kD polypeptide wasfound to elute last in fractions 94 to 106 during thechromatography (Fig. 1) (identified by an arrow). Inaddition, a 5 kD polypeptide, along with a 9 kD poly-peptide, eluted in fractions 74 to 90 (identified byarrows). The two fractions containing either 4 or 5 kDpolypeptides were individually pooled, dialyzed, con-centrated by lyophilizadon, and further separated by apreparative SDS-PAGE method using the BioRadModel 491 Prep Cell. On examination of the Prep

(a) Sephadex G-50 Chromatography

20 30 40 SO 60 70Fraction Number

t l (b) SDS-PAGE o« Column Fractions

s 62 66 70 74 78 82 B6 90 98 102 106

FIGURE l. Sephadex G-50 chromatogram of water-solubleproteins isolated from lenses of 60-80-year-old donors. Thewater-soluble proteins (200 mg protein) were freeze-dried,dissolved in buffer A, and chromatographed through Sepha-dex G-50 column as described in Materials and Methods.Two columns (2.5 X 70 cm each) were sequentially con-nected. Equilibration and elution were carried out by bufferA at room temperature, (a) Elution profile at 280 nm of thewater-soluble proteins fractionated through the Sephadexcolumn; (b) SDS-PAGE analysis of column fractions. Thecolumn fractions are identified by their numbers at the topof the gel. The 4, 5, and 9 kD polypeptides are identified byarrows. Two fractions consisting of fractions 74 to 90 and 94to 106 were pooled separately and were further fractionatedas described in Figure 2.

FIGURE 2. Purification of 4 kD and 5 kD polypeptides by apreparative SDS-PAGE method using the BioRad 491 PrepCell, (a) Fractionation of 4 kD polypeptide-containing frac-tion (column fractions 94 to 106 of Fig. lb) and (b) fraction-ation of 5 kD polypeptide-containing fraction (column frac-tions 74 to 90 of Fig. lb). Both the fractions in (a) and (b)were subjected to preparative SDS-PAGE by an identicalprocedure using 15% acrylamide gels. The preparative SDS-PAGE was carried out according to manufacturer instruc-tions.

Cell-eluted fractions, the 4 kD polypeptide was foundin fractions 8 to 15 whereas a contaminating 9 kDpolypeptide eluted later in fractions 17 to 25 (Fig. 2a).A similar preparative SDS-PAGE separation of thepooled fractions 74 to 90 (as shown in Fig. 1) sepa-rated the 5 kD polypeptide (fractions 8-10) from the 9kD polypeptide (Fig. 2b).

Further purification of the 4 and 5 kD polypep-tides was carried out by an HPLC method using a C-18column as shown in Figure 3. The 4 kD polypeptideeluted as a broad peak at 16 to 18 minutes with ~ 6 0 %acetonitrile during elution (Fig. 3a). The 4 kD polypep-tide was identified in the peak fractions by SDS-PAGEanalysis. The 4 kD polypeptide peak fractions wereconcentrated by freeze-drying and dialyzed. On a sec-ond HPLC chromatography, a single peak of the 4 kDpolypeptide eluting at 21 minutes with 68% acetoni-trile was observed that showed a single protein bandon SDS-PAGE (Fig. 3, b and e, lane 2). A shift in theretention time of the 4 kD polypeptide during the sec-ond chromatography could be attributable to absenceof SDS that was present in trace amounts during thefirst chromatography. On an identical C-18 column-HPLC of the 5 kD polypeptide, a single major peakwas also observed at 22 minutes (Fig. 3c). The 5 kDpolypeptide containing fraction, after second identicalHPLC, showed a single band on SDS-PAGE (Fig. 3e,



2 4 6 8 10 12 14 16 IS 20 22 24 26 2B 30 2 4 6 8 10 t2 14 16 18 20 22 24 26 28 30

2 4 6 8 10 12 14 16 IB 20 22 24 26 28 36 2 4 6 8 10 12 14 16 18 20 22 24 26 2B 30

Time (mln)

FIGURE 3. Purification of 4 and 5 kD polypeptides by HPLC using a C-18 column (Whatman).The individual 4 or 5 kD polypeptide-containing fractions, recovered after preparative SDS-PAGE as described in Figure 2 were dialyzed against 0.05 M Tris-HCl, pH 7.9 at 4° for 48hours and then further fractionated by HPLC. (a) HPLC of 4 kD polypeptide containingfraction recovered after preparative SDS-PAGE; (b) rechromatography of the purified 4 kDpolypeptide recovered in (a); (c) HPLC of 5 kD polypeptide recovered after preparativeSDS-PAGE; (d) rechromatography of 5 kD polypeptide recovered as described in (c) and (e)SDS-PAGE analysis of 4 kD (lane 2) and 5 kD (lane 3) polypeptides isolated by the HPLCmethods as described in (b) and (d).

lane 3). The 5 kD polypeptide was eluted with about68% of acetonitrile during both the HPLC separations(Fig. 3, c and d).

Purity of the 4 and 5 kD PolypeptidesThe purity of the 4 and 5 kD polypeptide was deter-mined by an HPLC method using a C-18 column(Whatman). As stated earlier and shown in Figure 3B,the purified 4 kD polypeptide showed a single majorpeak on a C-l 8 column, which, when pooled and exam-ined by SDS-PAGE, showed a single 4 kD band duringSDS-PAGE (Fig. 3e, lane 2). Similarly, the 5 kD poly-peptide also showed a single protein peak on a C-18HPLC analysis and a single band during SDS-PAGE(Fig. 3, d and e, lane 3).

Determination of Parent Cry stall ins of 4 and 5kD Polypeptides

The parent crystallins of each of the polypeptides weredetermined by examining the immunoreactivity of theanti-4 kD or anti-5 kD polypeptides antibodies to chro-

matographically separated individual crystallins by theWestern blot method. For this analysis, the a, /3, andy-crystallins of lenses from a 21-month-old donorwere separated by an HPLC method using a TSKG-4000 PWXL column. Among the crystallin species, a22 kD /3-crystallin species reacted most strongly withthe anti-4 kD polypeptide antibody (Fig. 4b, lanes7-8). A similar Western blot analysis showed the immu-noreactivity of the anti-5 kD polypeptide antibody tothe a-crystallin isolated from lenses of 21-month-olddonors (Fig. 4c, lanes 3-4). Together the Western blotresults showed that the 4 kD and 5 kD polypeptideswere fragments of the jS- and a-crystallins, respec-tively. On the Western blot analysis of the 4 kD poly-peptide with two site-specific and-/5A 3/A1 antibodies,one to a peptide representing the N-terminal region(residues 37 to 68), and the other to a peptide repre-senting C-terminal region (residues 204 to 215), thepolypeptide showed immunoreactivity to the antibodydirected to the C-terminal region (Fig. 5, a and b). Asshown later, the N-terminal amino acid sequence de-



(a)

20-

14.4-

P + Y

143.7-12&9-I

118.4-1114.7-115.5-1

(b)1 2 3 4 5 6 7 8 9 10

i

43.7-28.9 - *\

18.4 -214.7" *

5.5-2.9-

(c)1 2 3 4 5 6 7 8 9 10

2.9-

FIGURE 4. Western blot analysis of a, /?, and 7-crystallins of lenses from a 21-month-old donorusing anti-4 kD and anti-5 kD poiypeptide antibodies. The crystallins were fractionated fromwater-soluble proteins by HPLC using a TSK G-4000 PWXL column, (a) Coomassie bluestaining of protein species of each crystallin fraction; (b) an immunoblot after reactivity of •crystallins with anti-4 kD poiypeptide antibody; and (c) an immunoblot after reactivity of thecrystallins with anti-5 kD poiypeptide antibody.

termination showed that the 4 kD poiypeptide was in-deed a fragment of the /?A 3/Al crystallin and its C-terminal end was intact. Although the band in the blot(Fig. 5b) appeared to be a doublet, on careful examina-tion of an additional identical experiment it was foundto be a single band. To ascertain the origin of the 5 kD

poiypeptide further from either aA or aB species, thelatter two species were isolated by the preparativeSDS-PAGE method and immunoreacted with the anti-5 kD poiypeptide antibody. As shown in Figure 5 (cand d), only aB showed immunoreactivity to the anti-body.

(a) (b) (c)12 3 4 12 34

FIGURE 5. Western blot analysis of the 4 kD poiypeptide (a and b) and aA- and aB-crystallin (cand d) using anti-C-terminal jS A 3/Al crystallin antibody and anti-5 kD poiypeptide anti-body, respectively, (a) and (b), lane 2: Coomassie blue stained purified 4 kD poiypeptide (4Mg) and an immunoblot, respectively, (c) and (d): Coomassie blue stained a crystallin speciesand its immunoblot, respectively. Both aA and aB species were purified from lenses of19-year-old donors by a preparative SDS-PAGE method, (c) and (d), lane 2: aA crystallin (3fig); lane 3: aA and aB crystallin (10 Mg). a nd lane 4: aB crystallin (8 ^g).



To ascertain the exact cleavage site in the identi-fied crystallins to generate the respective polypeptidefragments, the N-terminal amino acid sequences ofeach polypeptide was determined. As shown in Figure6, the partial N-terminal sequence of the 4 kD polypep-tide matched to that of human lens j8 A 3/Al crystallin.The cleavage occurred at His187-His188 bond produc-ing a polypeptide with 28 residues. The calculated mo-lecular weight of this polypeptide from the publishedsequence would be 3405 daltons, which was close toour determined molecular weight of ~4000 daltonsbased on SDS-PAGE analysis. As stated earlier, theC-terminal amino acid residues of this polypeptide re-mained intact after cleavage as ascertained by the site-specific C-terminus antibody.

The partial N-terminal amino acid sequence analy-sis of the 5 kD polypeptide showed that its N-terminalsequence matched with that of aB-crystallin with cleav-age at Val145-Asn146. Based on the human aB aminoacid sequence,17 cleavage would produce a fragmentof 3273 daltons, which was shorter than our estimatedmolecular weight of 5 kD for this polypeptide.

(a)

Human Ions 4 kDa Polypeptide: X - E - G - D - Y - K - H - X - R - E

Human Lens P A 3/1:

Bovine Lens p A 3/1:

Residue no.:

H - E - G - D - Y - K - H - W - R - E

H - G - G - D - Y - K - H - W - R - E

188- 189 - 190 -191 • 192 -193 • 194 - 195 - 196 -197

Human lens 5 kDa polypeptide:

Human lens aB crystallin:

Residue no.:

N - G - P - R - L - V

N - G

146 • 147 • 148 - 149 - 150 - 151

FIGURE 6. Comparison of partial N-terminal amino acid se-quences of 4 and 5 kD polypeptides with that of /3 A 3/A1-and aB-crystallins. (a) Comparison of the N-terminal aminoacid sequence of 4 kD polypeptide with /3 A 3/Al crystallinfrom bovine and human lenses, and (b) Comparison of theN-terminal amino acid sequence of 5 kD polypeptide withthe human lens aB crystallin.

DISCUSSION

Posttranslational modification as well as cleavage oflens crystallins during aging and cataractogenesis iswell documented.1"4'8'9 Among the crystallin frag-ments, a 10 kD species has been extensively studiedand is believed to be a mixture of fragments of a, (3,and/or 7-crystallins.1'2'4 Recently we identified 9 kDand 8 kD fragments of 7D and 7S-crystallins, respec-tively, in human lenses.513 Together these results sug-gest that a mixture of crystallin fragments exists inhuman lenses and careful identification of their parentcrystallins and exact cleavage sites are needed. Theidentified cleavage sites may provide clues about theidentity and specificity of endogenous proteinase thatmay be responsible for such cleavage. These polypep-tides may play a potential role in the formation ofheavy molecular weight proteins. Such a role is specu-lated because these fragments: (1) are present at in-creasing levels in water insoluble protein frac-tions1'2'410; (2) are able to aggregate and cross-link invitro per se to species with molecular weight to 1.5 X106 daltons or higher10; (3) possibly exist in vivo incovalently cross-linked form in water insoluble pro-teins;10 (4) are present in opaque but not in clear por-tion of the same cataractous lens18: (5) become insolu-ble in aqueous solution19; and (6) of /3L crystallin onCalpain treatment became insoluble and turbid. Suchan insolubilization could be inhibited by addition of aCalpain inhibitor E64.11 For these aforementionedreasons, it is important to identify and characterize themajor crystallin fragments of human lenses.

Results presented in this study show that two poly-peptides of 4 and 5 kD could be separated by gel filtra-tion chromatography from older human lenses of 60-80-year-old donors. However, such separation waspossible only under denaturing conditions, ie, in thepresence of urea. Under nondenaturing conditions,all the three polypeptides remained associated withcrystallins after gel filtration chromatography (resultsnot shown).

After Sephadex G-50 gel filtration chromatogra-phy of water-soluble proteins, the SDS-PAGE analysisof column fractions showed three major polypeptidesof 4, 5, and 9 kD. Additional polypeptides with molecu-lar weights of higher than 9 kD were also observed butthey eluted with crystallins. The 4 and 5 kD polypep-tides were the focus of our current studies because oftheir relatively easier separation from crystallins. How-ever, additional major polypeptides with molecularweights higher than 9 kD that remained associatedwith crystallins during the chromatography (Fig. 1)were also purified in our laboratory by preparativeSDS-PAGE method and their N-terminal amino acidsequence are presently under investigation.

Purification of individual polypeptides by a pre-



parative SDS-PAGE method using BioRad Prep Cellmodel 491 proved to be an easy and efficient method.The 4 kD polypeptide could easily be separated fromthe 9 kD polypeptide by this method (Fig. 2a). A simi-lar separation of the 5 kD polypeptide from 9 kD poly-peptide was also achieved by this procedure (Fig. 2b).Further purification of the 4 and 5 kD polypeptides bya C-18 HPLC column showed a major peak with fewminor peaks. This suggested that both 4 and 5 kD poly-peptide species were the major polypeptides in thepurified preparations. A few additional comigratingspecies comprising less than 20% of the major polypep-tide also existed in these fractions as determined byHPLC profiles. The HPLC-purified 4 and 5 kD wereused for developing polyclonal antibodies and for N-terminal amino acid sequence analysis. Further, basedon the absorbance at 280 nm and protein determina-tion results, the 4 and 5 kD polypeptide comprised lessthan 1% of total water-soluble protein of lenses from60-80-year-old donors. An estimated 5-10 ng of eachpolypeptide per lens was recovered.

Both immunologic (Western blot) and partial N-terminal amino acid sequence analysis results showedthat the 4 kD and 5 kD polypeptides were fragments of/? A 3/A1 and aB-crystallins, respectively. This was fur-ther confirmed as the 5 kD polypeptide showed noimmunoreactivity to anti-aA crystallin antibody, whichwas donated by Dr. Paul Fitzgerald (University of Cali-fornia, Davis). Similarly, the antibody directed to theC-terminal peptide of 12 residues of/? A 3/A1 crystal-lin showed immunoreactivity with the 4 kD poly-peptide.

The N-terminal sequence analyses data showedthat it was the in vivo cleavage of His187-Hisi88 bond ofhuman fi A 3/A1 that generated the 4 kD polypeptidewhereas the Vali45-Asn146 of aB was cleaved to gener-ate the 5 kD polypeptide. Our four attempts to deter-mine the N-terminal amino acids of 5 kD polypeptidebeyond the five amino acid residues yielded mixed re-sults. This could be due to modifications of certainamino acids of this fragment in the older lenses. Wetherefore reported only the five N-terminal aminoacid residues in Figure 6 that could be reproduced.Further, because of very close retention time for Leuand Lys, 27.16 and 27.24 minutes, respectively, on aPTH-C-18 column, the identity of the fifth residuescould not definitely be established.

Our previous results demonstrated that Gly-Ser of7D-crystallin and Ser-Gly of 7S-crystallins werecleaved in vivo to generate the 9 kD5 and 8 kD polypep-tides,13 respectively. Currently, neither the identity ofa lens endogenous proteinase with such a specificityfor amino acids nor enzymatic or nonenzymatic natureof such cleavage sites are known.

Results of this report have shown that the /5 A3/A1 crystallin is cleaved in vivo in human lenses. The

jS-crystallin is most heterogenous among crystallinsand apparently plays a vital role in maintaining lenstransparency. Indeed, in the congenital cataractmodel of Philly mouse, the abnormal /3B2, the princi-pal (8 crystallin of the mouse lens, possibly leads toinherited cataract development.20 The opacificationwas possibly caused by the deletion of four amino acidsat the carboxyl terminal that may be necessary for sta-bility of Philly mouse j8B2. The breakdown of a and/3-crystallins during selenite cataract of mouse hasbeen reported.2122 Further, in the selenite cataract,the major portion of water-insoluble proteins presum-ably responsible for opacification, was composed ofpartially degraded /3-crystallin polypeptides.22 Simi-larly, the in vitro incubation of water-soluble proteinswith Calpain resulted in proteolysis of )8L crystallin andturbidity.11 Together, these results suggest that pro-teolysis of crystallins may lead to lens opacification.However, the exact mechanism leading to such opaci-fication is yet unknown.

It is interesting that in the acidic /?-crystallin, thereis insertion of an additional four amino acids com-pared to other ^-crystallins23 and the identified cleav-age site, ie, His187-His188, is located among these fouramino acid residues. Further, it is apparent that theseinserted amino acid residues are possibly surface ex-posed because they are adjacent to residue 154, whichhas been shown to be surface exposed.23 This may re-sult in a preferential cleavage in these sites by an enzy-matic or nonenzymatic process.

Our results also demonstrated the cleavage of aBin vivo in human lenses. Of aA and aB, the latter isfound in many nonocular tissues.24 Furthermore, aBgene expression occurs in various diseases cells; eg,astrocytes of patients with Alexander's disease25 al-though its function/dysfunction under pathologicalconditions is not known. However, it is definitelyproven that aB is a lens structural protein and alsoexhibits the properties of a heat shock protein.26 Acleavage in aB in vivo may cause changes in its tertiarystructure and possibly its function as a heat shock pro-tein. However, any such effects on properties of theaB-crystallin must await further investigation regard-ing a definite function of a crystallin in vivo. Further,the enzymatic or nonenzymatic nature of the cleavageof crystallins must be known, which is the focus of ourcurrent investigation.

Key Words

crystallins, lens, human, polypeptides, protein

A cknowledgmen t

The authors thank Ms. Janet Hussey for her expert typing.



References

1. Roy D, Spector A: Human insoluble lens protein II.Isolation and characterization of a 9600 dalton poly-peptide. Exp Eye Res. 1978;26:445-459.

2. Garner WH, Garner MH, Spector A: Comparison of10,000 and 43,000 dalton polypeptide population iso-lated from the water soluble and insoluble fractions ofhuman cataractous lenses. Exp Eye Res. 1979; 29:257-276.

3. Zigler S, Horwitz J, Kinoshita J: Studies on the lowmolecular weight proteins of human lens. Exp Eye Res.1981;32:21-30.

4. Takemoto LJ, Straatsma B, Horwitz J: Immunochemi-cal characterization of the major low molecular weightpolypeptide (10 K) from human cataractous lenses.Exp Eye Res. 1989;48:261-270.

5. Srivastava OP, McEntire JE, Srivastava K: Identifica-tion of a 9 kDa-7-crystallin fragment in human lenses.Exp Eye Res. 1992;54:893-901.

6. Emmons T, Takemoto L: Age-dependent loss of theC-terminal amino acid from alpha crystallin. Exp EyeRes. 1992;55:551-554.

7. Kim K, Rhee SG, Stadman EF: Non-enzymatic cleav-age of proteins by reactive oxygen species generatedby dithiothreitol and iron. / Biol Chem. 1985; 26:15394-15397.

8. de Jong WW, van Kleef FSM, Bloemendal H: Intracel-lular carboxyl-terminal degradation of aA chain of a-crystallin. EurJBiochem. 1974;48:271-276.

9. Roy D, Spector A: High molecular weight proteinsfrom human lenses. Exp Eye Res. 1976;22:273-279.

10. Srivastava OP: Age-related increase in concentrationand aggregation of degraded polypeptides in humanlenses. Exp Eye Res. 1988;47:525-543.

11. Shearer TR, David LR, Anderson RS, Azuma M: Re-view of selenite cataract. Invest Ophthalmol Vis Sci.1992; 11:357-369.

12. Siezen R, BindelsJ, Henders H: The interrelationshipbetween monomeric, oligomeric and polymeric alphacrystallin in calf lens nucleus. Exp Eye Res.1979;28:551-567.

13. Srivastava OP, Srivastava K, Silney C: Identification ofa 7S-crystallin fragment in human lenses. Exp Eye Res.1993;56:367-369.

14. Laemmli UK: Cleavage of structural proteins duringthe assembly of head of bacteriophage T4. Nature(London) 1970; 227:680-685.

15. Towbin H, Stahelin T, Gordon T: Electrophoretictransfer of proteins from polyacrylamide gels to nitro-c e l l u l o s e s h e e t s . Proc Natl Acad Sci U S A .1979;79:4350-4354.

16. Peterson GL: A simplification of the protein assaymethod of Lowry et al which is more generally applica-ble. Anal Biochem. 1977;83:346-356.

17. Kramps JA, de Man BM, de Jong WW: The primarystructure of the B2 chain of human a-crystallin. FEBSLett. 1977; 74:82-84.

18. Horwitz J, Hansen JS, Cheung C-C, Ding L-L,Straatsma BR, Lightfoot DO, Takemoto LJ: Presenceof low molecular weight polypeptides in human bru-nescent cataracts. Biochem Biophys Res Commun.1983; 113:65-71.

19. Yoshida H, Yumoto N, Tsukahara I, Murachi T: Thedegradation of a-crystallin at its carboxyl terminalportion by calpain in bovine lens. Invest Ophthalmol VisSci. 1986; 27:1269-1273.

20. Chambers C, Russell P: Deletion mutation in an eyelens /3-crystallin, an animal model for inherited cata-ract. J Biol Chem. 1981; 266:6742-6746.

21. David LL, Dickey BM, Shearer TR: Origin of urea-soluble protein in selenite cataract. Invest OphthalmolVis Sci. 1987; 23:1148-1156.

22. David LL, Shearer TR: Calcium-activated proteolysisin lens nucleus during selenite cataractogenesis. InvestOphthalmol Vis Sci. 1984;25:1275-1283.

23. Slingsby C, Driessen HPC, Mahadevan D, Bax B,Blundell TL: Evolutionary and functional relationshipbetween basic and acidic /3-crystallin. Exp Eye Res.1988;46:375-403.

24. Bhat SP, Nagineni CN: a B subunit of lens-specificprotein a-crystallin is present in other ocular and non-ocular tissues. Biochem Biophys Res Commun.1989; 158:319-325.

25. Iwaki T, Kume-Iwaki A, Liem RKH, Goldman JE: a/3-crystallin is expressed in non-lenticular tissues and ac-cumulates in Alexander's Disease brain. Cell.1989;57:71-78.

26. Klemenz R, Frohli E, Steinger R H, Schafer R,Aoyama A: aB-crystallin is a small heat shock protein.Proc Natl Acad Sci USA. 1991; 88:3652-3656.


identification of origin of two polypeptides of 4 and 5 kd isolated

Documents