2011...

Effects of Ultraviolet-A and Riboflavin on the Interaction ofCollagen and Proteoglycans during Corneal Cross-linking*SReceived for publication, July 28, 2010, and in revised form, January 31, 2011 Published, JBC Papers in Press, February 18, 2011, DOI 10.1074/jbc.M110.169813

Yuntao Zhang1, Abigail H. Conrad, and Gary W. ConradFrom the Division of Biology, Kansas State University, Manhattan, Kansas 66506-4901

Corneal cross-linking using riboflavin and ultraviolet-A(RFUVA) is a clinical treatment targeting the stroma in progres-sive keratoconus. The stroma contains keratocan, lumican,mimecan, and decorin, core proteins of major proteoglycans(PGs) that bind collagen fibrils, playing important roles in stro-mal transparency. Here, amodel reaction system using purified,non-glycosylated PG core proteins in solution in vitro has beencomparedwith reactions inside an intact cornea, ex vivo, reveal-ing effects of RFUVA on interactions between PGs and collagencross-linking. Irradiation with UVA and riboflavin cross-linkscollagen and chains into larger polymers. In addition,RFUVA cross-links PG core proteins, forming highermolecularweight polymers.When collagen type I is mixed with individualpurified, non-glycosylated PG core proteins in solution in vitroand subjected to RFUVA, both keratocan and lumican stronglyinhibit collagen cross-linking. However, mimecan and decorindo not inhibit but instead form cross-links with collagen, form-ing new high molecular weight polymers. In contrast, cornealglycosaminoglycans, keratan sulfate and chondroitin sulfate, inisolation from their core proteins, are not cross-linked byRFUVAanddonot formcross-linkswith collagen. Significantly,when RFUVA is conducted on intact corneas ex vivo, both ker-atocan and lumican, in their natively glycosylated form, do formcross-links with collagen. Thus, RFUVA causes cross-linking ofcollagen molecules among themselves and PG core proteinsamong themselves, together with limited linkages between col-lagen andkeratocan, lumican,mimecan, anddecorin. RFUVAasa diagnostic tool reveals that keratocan and lumican core pro-teins interact with collagen very differently than do mimecanand decorin.

The cornea is the transparent, dome-shaped tissue coveringthe front of the eye. It is a powerful refracting surface, providing6575% of the eyes focusing power, and is made up of threedistinct layers: epithelium, stroma, and endothelium. Thestroma comprises about 90% of cornea thickness in humans (1).

Although it is very highly innervated, it does not contain anyblood vessels (24). Collagen gives the cornea its strength, elas-ticity, and form (5). The unique molecular shape, paracrystal-line arrangement, and very regular fine diameter of the evenlyspaced collagen fibrils are essential in producing a transparentcornea (6, 7).In the corneal stromal extracellular matrix, glycosaminogly-

can (GAG)2 polysaccharides are the most abundant negativelycharged macromolecules and are classified on the basis of theirrepeating disaccharide structures into four main groups: kera-tan sulfate (KS), chondroitin sulfate/dermatan sulfate (CS/DS),heparan sulfate and heparin, and hyaluronan. Glycosaminogly-cans, normally covalently attached to proteoglycan (PG) coreproteins, play important roles in corneal transparency, nervegrowth cone guidance, and cell adhesion, largely dependent ontheir patterns of sulfation or lack of it (8, 9). The corneal stromais composed primarily of collagen fibrils, each of which consistsof a core of type V collagen coated with type I collagen (10, 11),coated in turn by two classes of PGs (12): keratan sulfate pro-teoglycan and chondroitin sulfate/dermatan sulfate pro-teoglycan. Through N-linked oligosaccharides, KS chainsare attached to three core proteins, lumican, keratocan, andmimecan, to formkeratan sulfate proteoglycans (1315). Thesethree core proteins belong to a class of proteins known as smallleucine-rich repeat proteins (1618). Through O-linked oligo-saccharide, CS/DS chains are attached to core protein decorin(19, 20). In the case of decorin, a single CS/DS linkage site ispresent near the amino terminus of the core protein (21, 22),whereas lumican and keratocan possess four or five potentialKS attachment sites in the central leucine-rich repeat region ofeach core protein molecule (13, 23, 24), and mimecan has twopotential KS attachment sites (25, 26). Currentmolecularmod-els of the corneal stroma suggest that these proteoglycan coreproteins wrap themselves laterally around the collagen fibrils ina manner that folds their hydrophobic domains inside, againstthe collagen fibrils (27). In contrast, the highly sulfated GAGchains (together with their associated water molecules ofhydration) are thought to stick out laterally away from the sidesof the collagen fibrils, forming an exterior hydrophilic shell.The thickness of that shell matches the thickness of the shellsurrounding adjoining fibrils, producing a very precise center-to-center spacing between the collagen fibrils characteristic of

* Thisworkwas supported, inwholeor inpart, byNational Institutes ofHealthGrant R01EY000952 (to G. W. C.). This work was also supported by theResearch Career Development Core (Brychta) in the Division of Biology atKansas State University GOBO000657 (to G. W. C.), by National ScienceFoundation Major Research Instrumentation Program Grant 0521587 (toKansas State University), and by the Terry C. Johnson Center for Basal Can-cer Research at Kansas State University.

S The on-line version of this article (available at http://www.jbc.org) containssupplemental Fig. S1.

1 To whom correspondence should be addressed: Division of Biology, AckertHall, Kansas State University, Manhattan, KS 66506-4901. Tel.: 785-532-6553; Fax: 785-532-6653; E-mail: [email protected].

2 The abbreviations used are: GAG, glycosaminoglycan; RF, riboflavin; UVA,ultraviolet-A; PG, proteoglycan; KER, keratocan; LUM, lumican;MIM,mime-can; DEC, decorin; Ab, antibody; Coll, collagen; KS, keratan sulfate; CS,chondroitin sulfate; GHCl, guanidine HCl; DS, dermatan sulfate; NMWL,normal molecular weight limit; RFUVA, riboflavin and ultraviolet-A; rcf, rel-ative centrifuge force.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 286, NO. 15, pp. 1301113022, April 15, 2011 2011 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

APRIL 15, 2011VOLUME 286NUMBER 15 JOURNAL OF BIOLOGICAL CHEMISTRY 13011

by guest, on April 19, 2012w

ww

.jbc.orgD

ownloaded from

http://www.jbc.org/content/suppl/2011/02/18/M110.169813.DC1.html Supplemental Material can be found at:

the corneal stroma and necessary for its transparency (28).Through this interaction with collagen (mostly with type I),PGs play important biological roles in collagen fibrillogenesisand matrix assembly (29). They also may serve as binders ofother proteins, some of whichmay be neurorepellants and neu-roattractants (30).Keratoconus is a vision disorder of unknown molecular eti-

ology but with some genetic basis that occurs when the nor-mally round cornea begins to bulge outward. This abnormalshape, arising as the central region becomes thinner, preventsthe light entering the eye from being focused correctly on theretina and causes distortion of vision (31). Keratoconus mayprogress for 1020 years and then slow in its progression. Eacheye may be affected differently. Keratoconus affects 1 in 2000people (32). Treatment options include conservative ap-proaches, aimed at maintaining visual acuity, such as rigid con-tact lenses placed to straighten corneal aberrations (33). Alter-natively, more invasive methods are applied using intrastromalcorneal implants (34) or performing anterior lamellar kerato-plasty or penetrating keratoplasty in extreme cases (35, 36).However, all of these techniques only correct the refractiveerrors of keratoconus and do not treat the cause underlying thecorneal ectasia and therefore do not stop the progression ofkeratoconus.Recently, a new technique of corneal cross-linking was

devised that directly improves the biomechanical rigidity of thecorneal stroma (37). This approach consists of irradiation withultraviolet-A (UVA) in the presence of the photosensitizer,riboflavin (RF), as a chromophore, to stop progression of thekeratoconus syndrome. The basic principle of corneal cross-linking is thought to be that in the presence of UVA at 370 nm,riboflavin is excited into its triplet state, generating singlet oxy-gen, which can react further with various molecules, inducingchemical covalent bonds bridging amino groups of collagenfibrils (38). In brief, the central 7 mm of the corneal epitheliumis removed to allow better diffusion of riboflavin into thestroma. Before irradiation with UVA, a 0.1% riboflavin solution(10 mg of riboflavin 5-phosphate in 10 ml of dextran 20% solu-tion) is applied onto the debrided central cornea every 5min for30 min. Then, as the application of riboflavin continues, theirradiation is performed from a 1-cm distance for 30 min withUVA at 370 nm and irradiance of 3 milliwatt/cm2 (5.4 J/cm2)(39). Clinical studies have shown that the progression of kera-toconus is effectively stopped (40), and the biomechanicalstrength of a keratoconus cornea is significantly increased by70300% (41, 42). In the anterior stroma, average collagen fibrildiameter in the treated cornea is significantly increased by12.2%, and in the posterior stroma, it is increased by 4.6% (43).Currently, laboratory studies aremainly focused on document-ing biomechanical effects (44), thermomechanical effects (45),morphological changes (46), effects on keratocytes (47), local-ization of cross-linking (48), and effects on collagenase resis-tance (49, 50). However, the chemical mechanisms of collagenand PG interactions catalyzed by riboflavin UVA treatmentduring corneal cross-linking have not been elucidated except inrecent work demonstrating that singlet oxygen and carbonylgroups are required (51). Some additional mechanisms arereported here.

EXPERIMENTAL PROCEDURES

MaterialsBovine eyeballs were freshly collected from AltaVista Locker (Alta Vista, KS). Collagen type I from bovine skinwas purchased from INAMED (Fremont, CA). Collagen type IIIand type IV from human placenta and RF 5-monophosphatesodium salt were purchased from Sigma (St. Louis, MO). KSfrom bovine cornea was purchased from Seikagaku America(East Falmouth,MA) andwas purified further by chromatogra-phy and chondroitinase ABC treatment (sodium salt), asdescribed in detail below. The average molecular mass of KSwas 15 kDa. CS from bovine trachea was purchased fromSigma. CS was purified by chemical treatment, as described indetail below. Recombinant human keratocan and human lumi-can proteins were purchased from Abnova Corp. (Taipei, Tai-wan). Recombinant mouse mimecan and recombinant humandecorin proteins were purchased from R&D Systems (Minne-apolis, MN). Anti-human keratocan polyclonal antibody waspurchased from Abnova Corp. Anti-human lumican antibody,anti-mouse mimecan antibody, and anti-human decorin anti-body were purchased from R&D Systems (Minneapolis, MN).Rabbit polyclonal antibody to collagen type I (ab34710) waspurchased from Abcam Inc. (Cambridge, MA). NuPAGENovex BisTris 412% precast gels (8 cm 8 cm 1.5 mm),NuPAGE Novex Tris acetate 38% precast gels (8 cm 8cm 1.5 mm), and chromogenic Western blot kits were pur-chased from Invitrogen (Carlsbad, CA).Preparation of Free KS Polysaccharide ChainsBased on the

manufacturers protocol for isolation of KS, there is residualcore peptide at the reducing terminus. N-Glycanase wasemployed to release N-linked KS chains from the short pep-tides. Following the procedure suggested by Prozyme, the mix-ture of KS and N-Glycanase was incubated overnight at 37 C.Enzymatic digestions were terminated by heating for 10 min inboiling water and then cooled to room temperature. Subse-quently, 1 volume of 100% ice-cold trichloroacetic acid (TCA)was added to 4 volumes of the digestion sample, mixed, andkept at 4 C overnight. The digestion sample was centrifuged at14,000 rpmand at 4 C for 30min to remove precipitatedN-gly-canase. The supernatant, containing KS, was transferred toUltra free-MC centrifugal filter units (5,000NMWL;Millipore)and washed extensively with water to remove salt, and then theretentate was lyophilized.Preparation of Free CS Polysaccharide ChainsToobtainCS

chains free of residual core protein peptides for cross-linkinganalysis, an alkaline -elimination reaction was carried out inthe presence of NaBH4. The CS sample was adjusted to 1.0 MNaBH4, 0.05 MNaOH, incubated at 45 C for 24 h (52), neutral-ized with 1 M acetic acid, washed extensively with water usingUltra free-MC centrifugal filter units (5,000NMWL;Millipore)to remove salt, and then the retentate, containing CS, waslyophilized.Cross-linking Model SystemPurified collagen type I was

used to assess the mechanisms of cross-linking induced by thephotosensitizer RF and UVA under conditions that resemblethose used for clinical treatment of progressive keratoconus.Themixture of purified collagen type I (1g/l in PBS) and RF(0.1% in PBS), in solution, was irradiated with UVA of 370 nm

Collagen and Proteoglycans in Corneal Cross-linking

13012 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286NUMBER 15APRIL 15, 2011


ww

.jbc.orgD

ownloaded from

for 30 min at a distance of 5 cm from the light source (UV-XRadiation System for Corneal Cross-linking CXL; IROCMedi-cal, Zurich, Switzerland). The volume of reaction solution was20 l in 0.5-ml plastic centrifugal tubes. The sample solutionswere irradiated with the UVA shining down directly onto thesurface of the solutions, mixed two times by vortexing at10-min intervals while it was being irradiated.Cross-linking Treatment on Intact Whole Corneas ex Vivo

The cross-linking procedure was performed as described pre-viously (51). In brief, the bovine corneal epithelium wasmechanically removed using a blunt knife, and riboflavin 0.1%(w/v) solution (RF PBS) was applied beginning 30min beforeirradiation and continuing every 5 min during irradiation. Thecorneas were irradiated with UVA of 370 nm for 30 min at adistance of 5 cm from the light source.Guanidine HCl Extraction of Corneal TissueThe de-epi-

thelialized bovine corneas (0.6 g wet weight), treated oruntreated with RFUVA, were frozen by liquid nitrogen, pulver-ized, homogenized further in 4 M guanidine HCl (GHCl) con-taining protease inhibitors (53), and then incubated for 24 h at04 C with gentle agitation. The tissue residue was removedby centrifugation at 10,000 g for 30 min, and the supernatantwas retained as the extract. The tissue residue pellet was re-ex-tracted for a second 24 hwith fresh 4 MGHCl solution. The twoextracts were combined together and then applied to an Ami-conUltra centrifugal filter (regenerated cellulose, 3,000Mr cut-off; Millipore), centrifuged to desalt, and concentrated to one-fifth of the original volume. The retentates that did not passthrough the filter were used for collagen cross-linking evalua-tion without any further processing.Pepsin Extraction of Corneal TissueFor pepsin extraction,

the frozen, pulverized, and homogenized corneal tissue (0.6 gwet weight) was extracted for 48 h in 20 volumes of 0.5 M aceticacid with 10 mg/ml pepsin (8002,500 units/mg, EC 3.4.23.1;Sigma) with gentle stirring movement at 04 C (54). The tis-sue residue was removed by centrifugation at 10,000 g for 30min. The supernatant was collected and neutralized by theaddition of NaOH and then desalted and concentrated asdescribed above. These retentates were used for collagen cross-linking evaluation.SDS-PAGEFor analysis of collagen and PG cross-linking, 5

l of NuPAGE LDS sample buffer (Invitrogen) and 2 l ofNUPAGE reducing agent (Invitrogen) were added into eachsample solution, heated at 70 C for 10 min, and then loadedontoNuPAGENovexBisTris 412%gels (8 cm 8 cm 1.5mm precast gel; Invitrogen) and subjected to electrophoresis(100mA/gel for 60min) under reducing conditions. After elec-trophoresis, the gels were stained with 0.1% (w/v) CoomassieBrilliant Blue R-250.Western Blot Analysis of Collagen Interaction with Proteogly-

can Core ProteinsFor detection of possible cross-linkingbetween collagen and proteoglycan core proteins, sampleswereloaded onto NuPAGE Novex Tris acetate 38% polyacryl-amide gels (8 cm 8 cm 1.5 mm precast gel; Invitrogen) andsubjected to electrophoresis (40 mA/gel for 60 min) underreducing conditions. Following electrophoresis, proteins weretransferred to nitrocellulose (Fisher) by electroblotting (Bio-Trans (Ann Arbor, MI) semidry electrophoretic transfer unit).

Protein transfer buffer was prepared following instructions ofBioTrans: pH 8.4, 48 mM Tris base, 39 mM glycine hydrochlo-ride, and 1.3mMSDS in 20%methanol. Voltagewas set at 150V.Transfer time was 45 min. Subsequently, chromogenic immu-nodetection was performed following the kit protocol for smallmembranes (Invitrogen). Collagen type I and core proteinswere identified using anti-collagen type I antibody (ab34710),anti-keratocan antibody (Abnova), anti-lumican antibody(AF2846), anti-mimecan antibody (AF2949), and anti-decorinantibody (AF143), respectively. The specificity of each of thesefour PG core protein antibodies was confirmed by Westernblotting of electrophoretic gels of all four core proteins (supple-mental Fig. S1).Analysis of the Interaction of GAGs and Collagen Type I with

Mass SpectrometryWhen gel electrophoresis was finished,the areas containing cross-linked collagen (the area from 150kDa up to the base of the gel sample well) and KS/CS (the areafrom 100 kDa down to the bottom of the gel) were each cut offand eluted from unstained SDS gels with electrophoretic elu-tion (model 422 Electro-Eluter, Bio-Rad). Eluted moleculeswere centrifuged through an Ultra free-MC centrifugal filterunit (5,000 NMWL;Millipore) to remove buffer salts. For anal-ysis of KS-sulfated disaccharides, the retentates were recoveredin 100l of 0.1 M ammonium acetate buffer (pH 6.0) containing0.1milliunits/l keratanase II and digested for 24 h at 37 C. Foranalysis of CS-sulfated disaccharides, the retentates insteadwere recovered in 100 l of 50 mM ammonium acetate buffer(pH 8.0) containing 1 milliunit/l chondroitinase ABC anddigested for 24 h at 37 C. The enzymeswere then inactivated at100 C for 5 min. Digest solutions (10 l) of each set werediluted by adding 70 l of MeOH, 5 l of 5 mM ammoniumacetate buffer (pH 7.5), 5 l of 2 mM (NH4)2SO4, and 10 l ofwater. The mixtures were centrifuged at 3,800 rcf for 30 min at4 C (Ultra free-MC centrifugal filter unit, 5,000 NMWL; Mil-lipore) with subsequent analysis of the retentates by electro-spray ionization-MS/MS (55, 56). Mass spectra were obtainedusing an electrospray ionization source on a quadrupole iontrap instrument (Bruker Daltonics Esquire 3000). Parametersused for analysis of all internal standards and corneal stromaltissue digests were as follows: spray voltage, 3.5 kV; dry gas(nitrogen), flow, 5.0 liters/min; drying temperature, 180 C. Themass range scanned was m/z 501000. Data acquisition soft-ware used was Bruker Daltonics Data Analysis version 3.0.Evaluation of Collagen Cross-linking inWhole CorneasAn-

tibodies to collagen type I and to each of the PG core proteinswere used to assess the cross-linking of collagen and PG coreproteins. Cleavage of collagen chains and non-collagen pro-teins at methionine residues with cyanogen bromide (CNBr)was performed in 70% formic acid under argon covered at roomtemperature as described previously (57). Samples were treatedwithN-glycanase orO-glycanase according to the manufactur-ers protocols (ProZyme), to remove intact chains of KS orCS/DS, respectively. Briefly, 2l ofN-glycanase orO-glycanasewas added to the reaction mixture (50 l) and incubated for24 h at 37 C. Other samples were digested with keratanase II(100 milliunits/ml in 0.1 M ammonium acetate buffer, pH 6.0)or with chondroitinase ABC (1000 milliunits/ml in 50 mMammoniumacetate buffer, pH8.0) for 24 h at 37 C, respectively




ww

.jbc.orgD

ownloaded from

(56) to depolymerize chains of KS or CS/DS while leaving theirlinkage region sugars still attached to the core protein. Enzy-matic digestions were terminated by heating for 10min in boil-ing water and then lyophilized. The same proportions of eachdigested sample were subjected to electrophoresis on 412%SDS gels and then toWestern blot analysis, as described above.

RESULTS

Establishment of Cross-linking Model SystemThree puri-fied collagens, type I, type III, and type IV, were chosen to assaythe reactivity of collagen cross-linking under the clinical condi-tions of treatment of keratoconus patients with UVA and ribo-flavin. The reaction was monitored by SDS-PAGE. The resultsshowed that collagen types I, III, and IV each display a bandingpattern distinct from the other two. For collagen type I, RFUVAcauses (a) 1, 2, and chains to almost disappear; (b) chainsto increase slightly in intensity; and (c) protein staining at thebase of the well to increase (Fig. 1, lane 3). For collagen types IIIand IV, similar results were observed, as shown for the RFUVAtreatment in Fig. 1, lane 5 (Type III) and lane 7 (Type IV). Theseresults suggest that collagen cross-linking definitely happenswhen irradiating with UVA in the presence of riboflavin. Theanterior corneal stroma is the location of the major effects ofriboflavinUVA treatment, and the stromal fibrils are primar-ily composed of collagen type I. This collagen type thereforewas chosen for the following experiments. Based on the initial

experiment shown in Fig. 1, we used the following standardreaction conditions in the rest of this study (unless stated oth-erwise): collagen type I (1 g/l) in 0.1 M PBS (pH 7.2), 0.1%riboflavin in 0.1 M PBS (pH 7.2), and UVA at 370 nm for 30minat a distance of 5 cm from the light source. This reactionmodelsystem was in a 0.5-ml plastic centrifuge tube and mixed twotimes by vortexing at 10-min intervals while it was being irra-diated from directly above the sample solution.Core Protein Cross-linkingUnder experimental conditions

described above for Fig. 1, core proteins keratocan, lumican,mimecan, and decorin (all at a concentration of 0.2g/l) eachmigrate as a monomer band. However, after exposure toRFUVA, each suchmonomer band virtually disappears (Fig. 2).For keratocan, lumican, andmimecan, RFUVAdoes not cause adistinct higher molecular mass band to form but rather causesformation of polymers of a very wide range of molecular massvalues from 100/150 kDa up to the bottom of the gel samplewell (Fig. 2, lanes 3, 5, and 7). In contrast, in the case of decorin,RFUVA causes an entirely new band to form of approximatelytwice the size of the monomer band (Fig. 2, lane 9).Interaction of Collagen and Core ProteinsTo detect possi-

ble collagen interactions with core proteins during RFUVA,collagen type I in solution, in the presence or absence of PGcore proteins, was irradiated byUVA in the presence of RF. Therelative intensity profiles of individual bands of SDS gels (Figs. 3and 4) are summarized in Table 1 as a percentage of the totalprotein in each scanned sample. When collagen type I in solu-tion (Fig. 3, lane 2) was exposed to RFUVA (Fig. 3, lane 3), therelative intensity of collagen chains and chains significantlydecreased, and the band at the very top of the gel increased.However, when the same reaction was performed in the pres-ence of keratocan (Fig. 3, lane 5) or lumican (Fig. 3, lane 7), the

FIGURE 1. SDS-PAGE patterns of cross-linking collagens. Collagen types I(lane 2), III (lane 4), and IV (lane 6) each displayed a banding pattern distinctfrom the other two. RFUVA causes collagen type I (lane 3) and chains toalmost disappear, chains to increase slightly in intensity, and protein stain-ing at the base of the gelwell to increase. A similar pattern of banddisappear-ance is seen also with type III collagen (lane 5) and type IV collagen (lane 7),with increased staining at the top of the gel (base of the sample well). Lane 1,molecular mass standards. Gels were stained with Coomassie Blue.

FIGURE 2. SDS-PAGE patterns of cross-linking core proteins. RFUVAcaused the monomer band of each PG to virtually disappear. For keratocan,lumican, andmimecan, RFUVA PG did not cause a higherMr band to form,instead forming new higher Mr polymers of a wide range of sizes from 100/150 kDa to the base of the gel well (lanes 3, 5, and 7). For decorin, RFUVAcaused an entirely new band to form of approximately twice the size of themonomer band (lane 9). Gels were stained with Coomassie Blue.




ww

.jbc.orgD

ownloaded from

relative intensity of collagen chains and chains did notdecrease as significantly as in their absence, suggesting thatkeratocan and lumican strongly inhibit the effect of RFUVA. In

other words, keratocan and lumican both appear to inhibit col-lagen cross-linking with other collagen molecules. However, incontrast to keratocan and lumican, in the presence ofmimecan,the intensity of collagen and chains significantly decreasedafter RFUVA, as if mimecan were not present (Fig. 3, comparelane 3 versus lane 9), indicating thatmimecandoes not interferewith collagen cross-linking (Fig. 3, compare lane 3with lane 9),resulting in40%decrease in the relative intensities of chainsand chains, whereas the relative intensity of chains (Fig. 3,lane 3 versus lane 9) was increased 20%. Similarly to mime-can, decorin does not interfere with collagen cross-linking inthe presence of RFUVA (Fig. 4, lane 3 versus lane 7), suggestingthat neither mimecan nor decorin interferes with RFUVA col-lagen cross-linking.Western Blot Analysis of Collagen Interaction with Core

ProteinsIn the SDS-PAGE experiments, the core protein-to-collagen ratio was 1:5. These same ratios were used in allWest-ern blot experiments. In case of keratocan (Fig. 5A), keratocanpartly inhibits RFUVA cross-linking of collagen I (lane 2 versuslane 4), yet keratocan does not appear to become bound tocollagen I by RFUVA (lane 5 versus lane 7 versus lane 8). Inter-estingly, if RFUVA is applied to lumican in the presence ofcollagen I (Fig. 5B, lanes 4 and 8), lumican displays a pattern ofinteraction with collagen similar to that of keratocan in thatRFUVA cross-linking of collagen I is partly inhibited (Fig. 5B,lane 2 versus lane 4). Thus, keratocan and lumican interact withcollagen I and respond to RFUVA in ways that correspondclosely with one another. In sharp contrast with that pattern,mimecan does not prevent RFUVA from causing collagen I toundergo cross-linking (Fig. 5C, lane 2 versus lane 4). Con-versely, however, the presence of collagen does affect the cross-linking response of mimecan to RFUVA (Fig. 5C, lane 7 versuslane 8). Thus, in the absence of collagen (Fig. 5C, lane 7), mime-can in the presence of RFUVA forms higher molecular weightregions, with an especially strong one at the bottom of the sam-ple well. However, in the presence of collagen I (Fig. 5C, lane 8),mimecan is found in two higher molecular weight regions,especially in the upper region of the gel, including in bands inthe same position as chains of collagen (arrowhead). In com-parison with mimecan, the responses of decorin to RFUVA areeven more dramatic in the new banding patterns created whencollagen is also present, especially in the appearance not only ofthe apparent dimer form seen in Fig. 4 (lane 4 versus lanes 5 and7) but also of three even higher molecular weight forms (Fig.5D, lane 7). The presence of collagen and RFUVA causesdecorin to migrate even less as its single monomer band, leavesthe dimer band approximately the same, but instead, causestwo of its four higher molecular weight polymers to displaygreatly diminished staining intensity, and causes a new group ofbands to form near the top of the gel co-migrating with the chains of collagen (arrowhead) (Fig. 5D, lane 8). Thus, althoughmimecan and decorin do not inhibit RFUVA-induced cross-linking of collagen I with itself as do keratocan and lumican,both mimecan and decorin do respond significantly to RFUVAand to the simultaneous presence of collagen I in ways thatcorrespond closely with one another and contrast greatly withthe pattern shown by keratocan and lumican. Thus, the patternof interaction of keratocan and lumican with collagen appears

FIGURE 3. Interaction of core proteins keratocan, lumican, and mimecanwith type I collagen. Presence of keratocan or lumican prevented RFUVA fromcausingdiminishedstainingofchainsandchainsofcollagen(lanes5and7). Incontrast, mimecan did not interfere with collagen cross-linking, thus allowing chains and chains to be cross-linked to formhighMr polymers near chains insize and toward the top rangeof thegel, and staining intensity increased sharplyat the base of the samplewell (lane 9). Gels were stainedwith Coomassie Blue.

FIGURE 4. Interaction of core protein decorin with type I collagen. In thepresence of RFUVA (lane 7), decorin did not interfere with collagen cross-linking, as evidenced by the virtual disappearance of and chains, theincreased density of chains, and the same increased density of staining inthe upper region of the gel as in the absence of decorin (lane 3). Gels werestained with Coomassie Blue.




ww

.jbc.orgD

ownloaded from

to be very different from the pattern of interaction of mimecanand decorin with collagen.Fig. 6 shows that keratocan and lumican not only interfere

with RFUVA cross-linking of collagen but also interfere withRFUVA cross-linking of both mimecan and decorin. In theabsence of keratocan and lumican, but even in the simultaneouspresence of both mimecan and decorin, collagen undergoescross-linking ( and chains almost disappear) (Fig. 6A, lane 1versus lanes 2 and 3). Decorin undergoes cross-linking in thepresence (Fig. 6, lane 5) or absence (Fig. 6, lane 4) of collagen I,but in its presence, decorin also occurs in a band in the sameposition as collagen I chains (arrowhead), suggesting thatsome decorin is bound to the chain, three-chain polymerform of collagen I. Similar to decorin, mimecan also occurs in aband in the same position as collagen I chains (arrowheads),

suggesting that some mimecan is bound to the polymer formof collagen I (Fig. 6A, lane 7). In the presence of keratocan, aswell as mimecan and decorin, collagen I cross-linking byRFUVA is partly inhibited (Fig. 6B, lane 3 versus lane 2). More-over, for both decorin and mimecan, RFUVA-induced cross-linking in the presence or absence of collagen I is inhibited bythe presence of keratocan (Fig. 6B, lane 4 versus lane 5 and lane6 versus lane 7 in comparison with those same lanes in Fig. 6A).Thus, the presence of mimecan or decorin in the same positionas collagen I chains is not seen in Fig. 6B, lanes 5 and 7 (arrow-heads), in comparison with those same lanes in Fig. 6A (arrow-heads), suggesting that keratocan can prevent both decorin andmimecan from cross-linking with themselves and also can pre-vent their interaction with collagen. Importantly, lumican dis-plays the same inhibitory ability as keratocan (Fig. 6C, lane 4versus lane 5 and lane 6 versus lane 7, arrowheads).Interaction of Collagen with KS and CSTo determine if KS

or CS might be involved in RFUVA-induced collagen cross-linking, mass spectrometry was used to analyze the SDS gelsamples. As can be seen from Fig. 7,AD, a strong signal inten-sity detected at m/z 462.0 corresponds to KS-monosulfateddisaccharide released by keratanase II from gel slices (area from100 kDa down to the bottom of the gel), whereas no signal atm/z 462.0 (Fig. 7, EH) was observed with the corresponding

TABLE 1Scanning density analysis of SDS-PAGE bands using Image Quantity Software TL (band intensity, percentage of total detected protein stain ineach sample, n 3)The density of each band is comparedwith the integrated total of all the bands (and diffuse staining) in that individual sample. For any individual sample, the sum of stainingfrom the 1, 2, , and bands will be less than 100%; the remainder represents the diffuse staining in otherMr regions of that sample.

2 1

Collagen I 10.75 0.39 24.37 0.43 27.36 0.60 23.06 0.16Collagen I RFUVA 5.89 0.41 10.99 0.39 15.34 0.61 44.47 1.07Collagen I KER RFUVA 7.71 0.16 17.63 0.29 32.56 0.76 42.34 0.79Collagen I LUM RFUVA 7.58 0.19 19.13 0.41 28.21 0.23 34.83 0.78Collagen IMIM RFUVA 3.22 0.22 7.28 0.20 11.76 0.58 54.07 0.61Collagen I DEC RFUVA 2.25 0.28 5.11 0.22 15.11 0.35 59.86 1.69

FIGURE 5.Westernblot of proteoglycan coreprotein interactionwith type Icollagen induced by RFUVA.Western blot is shown. A, the presence of kerato-canpreventedRFUVAfromcausingcollagentype I toundergocross-linking (lane4 versus lane 2). The intensities ofand chains (lane 4)were almost the sameasthe corresponding bands in lane 1. B, lumican inhibited collagen cross-linking(lane 4 versus lane 2). C, mimecan did not inhibit RFUVA-induced collagen cross-linking (lane2 versus lane4),whereasmimecan formedanewbandnear chainsand in the top range of the gel (lane 8 versus lane 7). D, decorin did not inhibitRFUVA-induced collagen cross-linking (lane 2 versus lane 4), whereas decorinformed a new band near chains (lane 8 versus lane 7). Decorin dimers andoligomers induced by RFUVAwere observed (lane 7).

FIGURE 6. Effect of keratocan or lumican on RFUVA cross-linking of colla-gen in the simultaneouspresenceofbothmimecananddecorin.Westernblot is shown. A, in the simultaneous presence of bothmimecan and decorin,collagen cross-linking in response to RFUVA is not inhibited, and bothmime-can and decorin each undergo their respective patterns of cross-linking. B, incontrast, in the presence of keratocan, as well as mimecan and decorin, col-lagen I cross-linking by RFUVA was partly inhibited (lane 3 versus lane 2). Inaddition, RFUVA-induced cross-linking of both decorin and mimecan in thepresence or absence of collagen I are inhibited (lane 4 versus lane 5 and lane 6versus lane 7). C, similarly, in the presence of lumican, collagen I cross-linkingby RFUVA was partly inhibited (lane 3 versus lane 2). In addition, RFUVA-in-duced cross-linking of both decorin andmimecan in the presence or absenceof collagen I is inhibited (lane 4 versus lane 5 and lane 6 versus lane 7).




ww

.jbc.orgD

ownloaded from

gel slice area from 150 kDa up to the base of the gel sample well,suggesting an absence of higher molecular polymer formsinvolving KS. As can be seen from Fig. 8, AD, in the case ofCS, a strong signal intensity was detected at m/z 458.0 for CS-monosulfated disaccharide (4S/6S) from gel slices (area from100 kDa down to the bottom of the gel), whereas no signal atm/z 458.0 (Fig. 8, EH) was observed with the correspondinggel slice area from 150 kDa up to the base of the gel sample well,suggesting an absence of higher molecular polymer formsinvolving CS. These results indicate that neither KS nor CS isinvolved in RFUVA-induced corneal cross-linking.Collagen and Core Proteins Cross-linking inWhole Cornea ex

VivoAntibodies to collagen type I and to each of the PG coreproteins were used to assay for possible cross-linking betweencollagen and PG core proteins in native cornea with RFUVA.Fig. 9A shows the electrophoreticmigration pattern of collagentype I fromwhole cornea treated by RFUVA and then extractedwith either GHCl or with pepsin; the latter, by cleaving only

their telopeptide domains, leaves the triple-chain helicaldomains of collagen molecules intact. Compared with controlsamples (Fig. 9A, lanes 1 and 3), the staining intensities of 1chains, 2 chains, and chains in the presence of RFUVAweredecreased, whereas the staining intensities of chains weresignificantly increased (Fig. 9A, lanes 2 and 4), thus indicatingthat collagen molecules were undergoing cross-linking even inthe presence of the normal mixture of all four natively GAG-glycosylated PGs ex vivo.Fig. 9C indicates that, in the absence of RFUVA treatment,

incubation of samples of purified collagen with CNBr causescollagen type I to be cleaved at methionine residues, generatingthe collagen type I-specific pattern of lowermolecular size pep-tides expected (Fig. 9C, lane 3). However, CNBr incubation ofcollagen type I that was first treated with RFUVAdemonstratesthat all domains of the collagen type I molecule are involved incross-linking into higher Mr molecules, as evidenced by theabsence of low Mr collagen peptides and, instead, the appear-

FIGURE 7. Analysis to detect possible interaction of KS and collagen type I, using mass spectrometry. Peaks at m/z 462.0 (AD) correspond to thecharacteristic KS-monosulfateddisaccharide releasedbykeratanase II fromgel slices (area from100kDadown to thebottomof theSDSgel). There arenopeaksatm/z 462.0 (EH) observed with gel slices in the area from 150 kDa up to the base of the gel sample well.




ww

.jbc.orgD

ownloaded from

ance of a broad range of higherMr molecules (Fig. 9C, lane 4).Fig. 9B indicates that regardless of whether whole corneal sam-ples were extracted from corneas using the GHCl protocolstandard for extracting PGs (53) or the pepsin protocol standardfor extracting the triple-helical domains of collagens (54), whensuch samples are derived from corneas treated with RFUVA,chemical cleavage with CNBr reveals a range of collagen peptidesthatmigrate in a highMWrange (Fig. 9B, lanes 2 and 4), just as dothe CNBr peptides from collagen type I (Fig. 9C, lane 4).Fig. 10, AD, displays the profiles of each individual PG core

protein extracted with GHCl fromwhole corneas with or with-out exposure to RFUVA and with or without treatments toremove GAG chains from their core proteins. Exposure ofwhole corneas to RFUVA causes some proportions of all pro-teoglycan core proteins to migrate in the same band positionsas collagen type I chains (Fig. 9), shown here as arrowheads,

regardless of whether glycosaminoglycan chains of KS wereintact (untreated lanes) or were removed with N-glycanase orwith keratanase II, consistent with PGs in solutions containingtype I collagen also co-migrating with collagen chains (Figs. 5and 6). Moreover, in corneas exposed to RFUVA, even aftercleavage of core proteins to small peptides with CNBr, a rangeof reactivity was detected in the high Mr region (arrowheads),whereas none of the core proteinswere detected in that highMrregion in samples from control corneas treated with CNBr.Thus, these results suggest that RFUVA causes some cross-linking to occur between PG core proteins and collagen I inwhole cornea, ex vivo.

DISCUSSION

Corneal cross-linking by RFUVA represents a new methodfor treatment of progressive keratoconus in Europe (40) and

FIGURE 8. Analysis to detect possible interaction of CS and collagen type I, using mass spectrometry. Peaks at m/z 458.0 (AD) correspond to thecharacteristic CS monosulfated disaccharide (4S/6S) released by chondroitinase ABC from gel slices (area from 100 kDa down to the bottom of the SDS gel).There are no peaks atm/z 458.0 (EH) observed with gel slices in the area from 150 kDa to the base of the gel sample well.




ww

.jbc.orgD

ownloaded from

currently is under clinical study in the United States. Here, wehave used both an in vitro and an ex vivomodel reaction systemto investigate the effects of RFUVA specifically on the interac-tion of collagen and PGs, an interaction very likely to occurduring corneal cross-linking. Our data demonstrate that irradi-ation with UVA in the presence of riboflavin cross-links colla-gen and chains into high molecular weight polymers. Inaddition, RFUVA can induce purified, non-glycosylated PGcore proteins themselves to cross-link to form higherMr poly-mers. When both collagen and PG core proteins are presenttogether in solution, both of the corneal PG core proteins, ker-atocan and lumican, inhibit collagen cross-linking. However, insharp contrast, two other corneal PG core proteins, mimicanand decorin, not only do not interfere with collagen cross-link-ing, they in fact form cross-links with collagen to form newbands in high molecular weight regions of electrophoretic gels.When studied in isolation from their normal PG core proteins,the corneal GAGs, KS and CS, are not induced by RFUVA tocross-link among themselves or to form cross-links with colla-gen. Thus, based on interactions between these purified pro-teins in solution in vitro, RFUVA mainly appears to involvecross-linking of collagen molecules with themselves, PG coreproteins with themselves, and collagen molecules with the twospecific PG core proteins, mimecan and decorin. These resultssuggest that in the absence of RFUVA, collagen interacts non-covalentlywith keratocan and lumican very differently from theway interacts with mimecan and decorin. Data from ex vivoexposure of whole corneas to RFUVA lead to these same con-clusions. This reveals two sharply distinct styles of molecularinteractions between collagen and PGs.Type I collagen, themost abundant collagen in cornea (11), is

composed of three polypeptide chains and two types of singlechains, 1 and 2, both of which are accessible to polymeriza-tion through intra- and intermolecular bonds (58). Dimers of chains are called -components (composed of 1 2 or 1 1chains). Trimers of three chains are called -components. Inthis study, RFUVAmainly causes 1, 2, 11, and 12 chains to

cross-link into macromolecules at least as large as -compo-nents, and larger (Fig. 1, lane 3) and thus to disappear from theirnormal and chain positions on electrophoretic gels. Inter-microfibrillar cross-links significantly affect the mechanicalproperties of the corneal collagen tissue (59). The strength ofthe collagen fibers depends on the formation of covalent cross-links between the telopeptide terminal domains and adjacenthelical domains of collagen molecules (60). Riboflavin-sensi-tized photodynamic modification of collagen causes formationof cross-linked molecules, in which an energy transfer betweenthe excited riboflavin and molecular oxygen produces singletoxygen that then interacts with the collagen fibrils in an oxida-tion reaction to formphysical, covalent cross-links (61). Resultspresented here provide new and direct evidence that covalentcross-links between collagenmolecules, especially involving1,2, 11, and 12 chains, are initiated by RFUVA and are consis-tent with earlier data indicating that singlet oxygen plays a pri-mary role in corneal cross-linking (51).

FIGURE 9. Effect of RFUVA on collagen type I cross-linking in whole cor-nea ex vivo versus collagen type I in solution. A, patterns of collagen type Ifrom whole cornea treated by RFUVA. Lanes 1 and 2, GHCl extraction; lanes 3and 4, pepsin extraction. B, patterns of collagen type I from whole corneatreated by RFUVA and then digested with CNBr. Lanes 1 and 2, GHCl extrac-tion; lanes 3 and 4, pepsin extraction. C, patterns of electrophoreticmigrationof purified collagen type I in solution, treated by RFUVA, and digested withCNBr. All peptide domains of collagen type I participate in cross-linking toform a range of highMr molecules. A and B, Western blot; C, Coomassie Bluestaining.

FIGURE 10. Patterns of cross-linking collagens and PG core proteins inwhole cornea ex vivo.Western blot is shown.A, interaction of keratocan andcollagen type I from corneas (with or without RFUVA treatment) treated byCNBr (lanes 4 and 5), N-glycanase (NGase) (lanes 6 and 7), and keratanase II(KSase) (lanes 8 and 9). B, interaction of lumican and collagen type I fromcorneas (with or without RFUVA treatment) treated by CNBr (lanes 4 and 5),N-glycanase (lanes 6 and 7), and keratanase II (lanes 8 and 9). C, interaction ofmimecan and collagen type I from corneas (with or without RFUVA treat-ment) treated by CNBr (lanes 4 and 5), N-glycanase (lanes 6 and 7), and kera-tanase II (lanes 8 and 9). D, interaction of decorin and collagen type I fromcorneas (with or without RFUVA treatment) treated by CNBr (lanes 4 and 5),O-glycanase (OGase) (lanes 6 and 7), and chondroitinaseABC (ABCase) (lanes 8and 9). All four PG core proteins in the cornea show some molecules thatco-migrate in the same region as collagen chains (arrowheads).




ww

.jbc.orgD

ownloaded from

Data presented here indicate that collagen types I, III, and IV,in isolation from one another and from PGs, undergo cross-linking reactions that all share some characteristics (Fig. 1).Both and chains virtually disappear, presumably because oftheir covalent cross-linking into higher molecular weight poly-mers that eithermigrate as chains (300 kDa) to some extentor as even larger aggregates, some of which barely enter the gelat the bottomof the samplewell (increased density at topof lane3 in Fig. 1). In response to RFUVA, normal corneal PG coreproteins, keratocan, lumican, mimecan, and decorin (but non-glycosylated (i.e. lacking their normal GAG chains)), similarlyappear to undergo cross-linking reactions with themselves thatcause their virtual disappearance as monomer molecules (Fig.2). One PG core protein, decorin, appears to form distinctdimermolecules (80 kDa) (Fig. 2, lane 9).When collagen typeI is subjected to RFUVA in the simultaneous presence of thesePG core proteins (Fig. 3), collagen cross-linking is drasticallyinhibited if either keratocan or lumican is present (Fig. 3, lanes5 and 7). In sharp contrast, however, the presence of mimecandoes not inhibit collagen cross-linking (Fig. 3, lane 9). In thosesame solutions containing collagen type I, PG core proteinskeratocan and lumican appear to undergo almost normal cross-linking with themselves (as indicated by a diminution of theirmonomer bands) as a result of RFUVA (Fig. 3, lanes 5 and 7), asdoes mimecan (Fig. 3, lane 9). These results suggest that ker-atocan and lumican can block those sites on collagen type I thatnormally allow collagen to cross-link with itself. However, evenwhen any one of those three PG core proteins is present,RFUVA treatment still produces large molecular aggregatesthat are present at the bottom of the sample wells (Fig. 3, lanes5, 7, and 9), as it does in their absence (Fig. 3, lane 3), perhapsrepresenting some of the cross-linked collagen seen in Fig. 1.Thus, none of the core proteins appears to produce molecularaggregates with collagen type I, at least in the region in whichmolecules can enter the gel (with the possible exception ofmimecan (Fig. 3, lane 9)). Decorin (Fig. 4), like mimecan, doesnot inhibit the cross-linking of collagen type I (Fig. 4, lane 7) inresponse to RFUVA; likewise, DEC undergoes its normal cross-linking with itself in response to RFUVA, forming a prominentdimer molecule in the absence (lane 5) and presence (Fig. 4,lane 7) of collagen type I. Also similar to mimecan, decorinappears to form a new macromolecule (300 kDa) whenexposed to RFUVA in the presence of collagen type I (Fig. 4,lane 7). Altogether, the results in Figs. 14 indicate that kerato-can and lumican interact with collagen type I in solution verydifferently than do mimecan and decorin.In cornea and in other tissues, any of these core proteinsmay

exist either in forms lacking glycosylation or with glycosylationconsisting of GAG chains. Such glycosaminoglycan polysac-charides (which do not appear to participate in RFUVA cross-linking, as shown here in Figs. 79) may sterically hold PG coreproteins at sufficient distances from collagenmolecules in adja-cent fibrils as to prevent formation of cross-links between themin response to RFUVA in vivo. However, such PG core proteins,even when heavily glycosylated with GAG chains, are modeledas bindingwith their leucine-rich domains directly to the side ofcollagen fibrils (62, 63), physical proximity that apparently doesnot facilitate RFUVA cross-linking, at least between collagen

andpurified, non-glycosylated keratocan or lumican in solutionin vitro. The mechanism by which keratocan and lumican sodrastically inhibit collagen cross-linking with itself by RFUVAtherefore suggests amechanism of steric blockage that does notput collagen molecules in reactive juxtaposition with eithernon-glycosylated PG core protein. In contrast, although bothmimecan and decorin fail to inhibit cross-linking of collagenwith itself, they each produce a new band at300 kDa (Fig. 5,C(lane 8) andD (lane 8)) that appears to represent co-polymers ofcollagen-PG core protein that have not been describedpreviously.Binding antibodies to collagen type I and to each of the PG

core proteins (in both glycosylated and non-glycosylatedforms) allowed an independent way to assess degrees of cross-linking between collagen and PG core proteins. In the absenceof core proteins, collagen type I in solution in vitro forms highermolecular weight polymers, especially leading both to the dis-appearance ofmost collagen chains and chains and to someenhancement of triple-chain polymers ( polymers; 300 kDa)(Fig. 5, lane 2 in each gel). In contrast, the simultaneous pres-ence of purified non-glycosylated keratocan or lumican in solu-tion in vitro virtually eliminated such collagen cross-linkingwith itself (Fig. 5, A and B, lanes 4). In those same solutions,keratocan and lumican do not undergo cross-linking in theabsence of RFUVA (lanes 5), but they do appear in a range ofhigher molecular weight polymers in response to RFUVA, inthe absence (lanes 7) or presence of collagen type I (lanes 8).Thus, no purified, non-glycosylated keratocan or lumicanappears to be bound to collagen during RFUVA in solution invitro, although those core proteins form cross-links with them-selves. In sharp contrast, although neither mimecan nordecorin inhibit the ability of collagen molecules to cross-linkwith themselves in response to RFUVA, both of those purified,non-glycosylated core proteins in the simultaneous presence ofcollagen form higher molecular weight polymer forms withthemselves and even larger ones in the presence of collagen(Fig. 5,C andD, lanes 8) than in its absence (lanes 7), suggestingthat they cross-link to collagen type I.When collagen type I was exposed to RFUVA in solution in

vitro in the presence of both mimecan and decorin, togetherwith either keratocan or mimecan, (all as purified, non-glyco-sylated proteins), collagen cross-linking with itself was inhib-ited (as expected, because of the presence of keratocan andlumican) (Fig. 6, A versus B or A versus C, lanes 3) but not asmuch as if mimecan and decorin had been absent. In theabsence of keratocan or mimecan, both mimecan and decorinepitopesaredetected inmacromolecules (equal toand300 kDa)in the presence of collagen (Fig. 6A, lanes 5 and 7). However, ifeither keratocan (gel B) or lumican (gel C) is present in thatmixture in solution in vitro, the amount of mimecan anddecorin detected in the very high molecular size molecules isreduced although not eliminated. Thus, in a complete cornea invivowith abundant collagen type I and all four PG core proteinsin glycosylated form, some collagen molecules would undergocross-linking with mimecan and decorin and might contributesubstantively to tissue strength, even if such collagen-PG coreprotein cross-links did not represent a large proportion of themolar reaction yield of product.




ww

.jbc.orgD

ownloaded from

Evidence presented in Fig. 8 suggests that the GAG compo-nents of native PG molecules do not themselves participate inRFUVA cross-linking. The smallest molecules were catego-rized as those in a size range arbitrarily below 100 kDa (Figs. 8(AD) and 9 (AD)), whereas the largest were in a size range of150 kDa and above (to the bottom of the sample well). Theseanalyses probed for evidence of cross-linking to keratan sulfate(Fig. 8, AH), chondroitin sulfate (Fig. 9, AH), or any protein,but no such new indicatormolecules were detected in responseto exposure to RFUVA, consistent with an apparent absence ofRFUVA causing visible cross-linking of corneal GAGs withthemselves or with collagen.Data presented in Fig. 9 indicate that collagen type I, present

in the corneal stroma and examined here ex vivo, displaysRFUVA cross-linking in the presence of the normal mixture ofall four native PGs (i.e. PG core proteins with their GAG chainsattached). Although the presence of the purified, non-glycosy-lated core proteins, keratocan and lumican, suppresses collagencross-linkingwhen bothmolecules are in solution in vitro (Figs.1, 3, and 4), when those same molecules are present in thepolymerized gel of the corneal stroma and in their native, gly-cosylated form, cross-linking of these PG core proteins to col-lagen is detected (Fig. 10). Future work will be needed to deter-mine if this difference in the ability of both keratocan andlumican to form RFUVA cross-links with collagen type I arisesbecause the presence of KS chains affects the conformation ofthese two core proteins as they bind to collagen (enhancingtheir RFUVA cross-linking to collagen) or because the higherdensities of molecules in the native stroma bring collagen andthese two core proteins close enough together to undergoRFUVA cross-linking. In contrast to the behavior of both ker-atocan and lumican (in solution in vitro versus in native glyco-sylated form ex vivo), bothmimecan and decorin appear able toform RFUVA cross-links with collagen in solution in vitro (Fig.5,C (lane 8) andD (lane 8)) and also in their native glycosylatedform ex vivo (Fig. 10, C and D, lane 9). These results strikinglydemonstrate that, as revealed here by using RFUVA as a type ofmolecular diagnostic probe, both keratocan and lumican inter-act with collagen type I very differently than do both mimecanand decorin. Elucidating the exact nature of those two styles ofmolecular interaction with collagen type I will be the subject offuture experiments.Protocols involving treatment of corneas with RFUVA are

being used for an ever wider set of clinical conditions (40, 64,65). It therefore is essential to elucidate as completely as possi-ble the precise molecular effects of that treatment, not only onthe cellular populations exposed (66) but also on the compo-nent molecules of the extracellular matrix. That has been thepurpose of the experiments performed here. Many other inter-actions catalyzed by RFUVA in the cornea remain to be char-acterized in this same manner (e.g. possible interactionsbetween collagens type I and type V within individual collagenfibrils and interactions between the collagen type IV of the epi-thelial basal lamina and laminin and the stromal proteogly-cans). Using RFUVA reactivity as a molecular diagnostic probemay therefore prove useful in formulating new hypothesesabout the native conformations of molecules in the extracellu-lar matrix in general.

REFERENCES1. Beuerman, R. W., and Pedroza, L. (1996)Microsc. Res. Tech. 33, 3203352. Marfurt, C. F., Cox, J., Deek, S., and Dvorscak, L. (2010) Exp. Eye Res. 90,

4784923. Al-Aqaba, M. A., Fares, U., Suleman, H., Lowe, J., and Dua, H. S. (2010)

Br. J. Ophthalmol. 94, 7847894. Muller, L. J., Pels, L., andVrensen, G. F. (1996) Invest. Ophthalmol. Vis. Sci.

37, 4764885. Worthington, C. R. (1984) Q. Rev. Biophys. 17, 4234516. Robert, L., Legeais, J.M., Robert, A.M., and Renard, G. (2001) Pathol. Biol.

49, 3533637. Hay, E. D. (1985) J. Cell. Biochem. 27, 1431568. Cardin, A. D., and Weintraub, H. J. (1989) Arteriosclerosis 9, 21329. Ruoslahti, E., and Engvall, E. (1980) Biochim. Biophys. Acta 631, 35035810. Hassell, J. R., and Birk, D. E. (2010) Exp. Eye Res. 91, 32633511. Newsome,D. A., Gross, J., andHassell, J. R. (1982) Invest. Ophthalmol. Vis.

Sci. 22, 37678112. Scott, J. E. (1991) Biochem. Soc. Trans. 19, 87788113. Funderburgh, J. L., Funderburgh,M. L., Brown, S. J., Vergnes, J. P., Hassell,

J. R., Mann, M. M., and Conrad, G. W. (1993) J. Biol. Chem. 268,1187411880

14. Funderburgh, J. L., Corpuz, L. M., Roth, M. R., Funderburgh, M. L., Ta-sheva, E. S., and Conrad, G. W. (1997) J. Biol. Chem. 272, 2808928095

15. Corpuz, L. M., Funderburgh, J. L., Funderburgh, M. L., Bottomley, G. S.,Prakash, S., and Conrad, G. W. (1996) J. Biol. Chem. 271, 97599763

16. Funderburgh, J. L., Funderburgh, M. L., Mann, M. M., and Conrad, G. W.(1991) Biochem. Soc. Trans. 19, 871876

17. Kao, W. W., and Liu, C. Y. (2002) Glycoconj. J. 19, 27528518. Carlson, E. C., Liu, C. Y., Chikama, T., Hayashi, Y., Kao, C. W., Birk, D. E.,

Funderburgh, J. L., Jester, J. V., and Kao, W. W. (2005) J. Biol. Chem. 280,2554125547

19. Li, W., Vergnes, J. P., Cornuet, P. K., and Hassell, J. R. (1992) Arch.Biochem. Biophys. 296, 190197

20. Bianco, P., Fisher, L. W., Young, M. F., Termine, J. D., and Robey, P. G.(1990) J. Histochem. Cytochem. 38, 15491563

21. Kjellen, L., and Lindahl, U. (1991) Annu. Rev. Biochem. 60, 44347522. Spiro, R. C., Freeze, H. H., Sampath, D., andGarcia, J. A. (1991) J. Cell Biol.

115, 1463147323. Corpuz, L. M., Dunlevy, J. R., Hassell, J. R., Conrad, A. H., and Conrad,

G. W. (2000)Matrix Biol. 19, 69970424. Dunlevy, J. R., Neame, P. J., Vergnes, J. P., and Hassell, J. R. (1998) J. Biol.

Chem. 273, 9615962125. Corpuz, L. M., Dunlevy, J. R., Hassell, J. R., Conrad, A. H., and Conrad,

G. W. (2000)Matrix Biol. 19, 69369826. Chen, R., Jiang, X., Sun, D., Han, G., Wang, F., Ye, M., Wang, L., and Zou,

H. (2009) J. Proteome Res. 8, 65166127. Iozzo, R. V. (1999) J. Biol. Chem. 274, 188431884628. Meek, K. M., Quantock, A. J., Boote, C., Liu, C. Y., and Kao, W.W. (2003)

Matrix Biol. 22, 46747529. Christiansen, D. L., Huang, E. K., and Silver, F. H. (2000)Matrix Biol. 19,

40942030. Conrad, A. H., Zhang, Y., Tasheva, E. S., and Conrad, G. W. (2010) Invest.

Ophthalmol. Vis. Sci. 51, 4500451531. Edrington, T. B., Zadnik, K., and Barr, J. T. (1995) Optom. Clin. 4, 657332. Romero-Jimenez, M., Santodomingo-Rubido, J., and Wolffsohn, J. S.

(2010) Cont. Lens Anterior Eye 33, 15716633. Garcia-Lledo,M., Feinbaum,C., andAlio, J. L. (2006)Compr. Ophthalmol.

Update 7, 475234. Ertan, A., and Colin, J. (2007) J. Cataract Refract. Surg. 33, 1303131435. Pramanik, S., Musch, D. C., Sutphin, J. E., and Farjo, A. A. (2006)Ophthal-

mology 113, 1633163836. Tan, D. T., and Por, Y. M. (2007) Curr. Opin. Ophthalmol. 18, 28428937. Wollensak, G., Sporl, E., and Seiler, T. (2003)Ophthalmologe 100, 444938. Krishna, C. M., Uppuluri, S., Riesz, P., Zigler, J. S., Jr., and Balasubrama-

nian, D. (1991) Photochem. Photobiol. 54, 515839. Seiler, T., and Hafezi, F. (2006) Cornea 25, 1057105940. Wollensak, G., Spoerl, E., and Seiler, T. (2003) Am. J. Ophthalmol. 135,




ww

.jbc.orgD

ownloaded from

62062741. Wollensak, G., and Spoerl, E. (2004) J. Cataract Refract. Surg. 30, 68969542. Wollensak, G., Spoerl, E., and Seiler, T. (2003) J. Cataract Refract. Surg. 29,

1780178543. Wollensak, G., Wilsch, M., Spoerl, E., and Seiler, T. (2004) Cornea 23,

50350744. Wollensak, G., and Iomdina, E. (2009) J. Cataract Refract. Surg. 35,

54054645. Spoerl, E., Wollensak, G., Dittert, D. D., and Seiler, T. (2004)Ophthalmo-

logica 218, 13614046. Dhaliwal, J. S., and Kaufman, S. C. (2009) Cornea 28, 626747. Mencucci, R., Marini, M., Paladini, I., Sarchielli, E., Sgambati, E.,

Menchini, U., andVannelli, G. B. (2010)Clin. Exp. Ophthalmol. 38, 495648. Kohlhaas, M., Spoerl, E., Schilde, T., Unger, G., Wittig, C., and Pillunat,

L. E. (2004) J. Cataract Refract. Surg. 32, 27928349. Spoerl, E., Wollensak, G., and Seiler, T. (2004) Curr. Eye Res. 29, 354050. Schilde, T., Kohlhaas,M., Spoerl, E., and Pillunat, L. E. (2008)Ophthalmo-

loge 105, 16516951. McCall, A. S., Kraft, S., Edelhauser, H. F., Kidder, G. W., Lundquist, R. R.,

Bradshaw, H. E., Dedeic, Z., Dionne, M. J., Clement, E. M., and Conrad,G. W. (2010) Invest. Ophthalmol. Vis. Sci. 51, 129138

52. Isemura, M., and Ikenaka, T. (1975) Biochim. Biophys. Acta 411, 112153. Conrad, G. W., Ager-Johnson, P., and Woo, M. L. (1982) J. Biol. Chem.

257, 46447154. Newsome,D. A., Gross, J., andHassell, J. R. (1982) Invest. Ophthalmol. Vis.

Sci. 22, 37638155. Zhang, Y., Kariya, Y., Conrad, A. H., Tasheva, E. S., and Conrad, G. W.

(2005) Anal. Chem. 77, 90291056. Zhang, Y., Conrad, A. H., Tasheva, E. S., An, K., Corpuz, L. M., Kariya, Y.,

Suzuki, K., and Conrad, G. W. (2005) Invest. Ophthalmol. Vis. Sci. 46,16041614

57. Epstein, E. H., Jr., Scott, R. D., Miller, E. J., and Piez, K. A. (1971) J. Biol.Chem. 246, 17181724

58. Makareeva, E., Mertz, E. L., Kuznetsova, N. V., Sutter, M. B., DeRidder,A. M., Cabral, W. A., Barnes, A. M., McBride, D. J., Marini, J. C., andLeikin, S. (2008) J. Biol. Chem. 283, 47874798

59. Parry, D. A. (1988) Biophys. Chem. 29, 19520960. Sung, H. W., Chang, W. H., Ma, C. Y., and Lee, M. H. (2003) J. Biomed.

Mater. Res. A 64, 42743861. de La Rochette, A., Birlouez-Aragon, I., Silva, E., and Morlie`re, P. (2003)

Biochim. Biophys. Acta 1621, 23524162. Funderburgh, J. L. (2000) Glycobiology 10, 95195863. Kao, W. W., Funderburgh, J. L., Xia, Y., Liu, C. Y., and Conrad, G. W.

(2006) Exp. Eye Res. 82, 3464. Kampik, D., Ralla, B., Keller, S., Hirschberg,M., Friedl, P., and Geerling, G.

(2010) Invest. Ophthalmol. Vis. Sci. 51, 3929393465. Snibson, G. R. (2010) Clin. Exp. Ophthalmol. 38, 14115366. Wollensak, G., Sporl, E., Reber, F., Pillunat, L., and Funk, R. (2003) Oph-

thalmic Res. 35, 324328




ww

.jbc.orgD

ownloaded from

2011...

Documents

pg core proteinsamong

evenlyspaced collagen

glycosylated form

lumican core proteins

corneal crosslinking

heparan sulfate

keratan sulfate

ing effects of rfuva