Chondroitin sulfate proteoglycans of bovine cornea: structural characterization and assessment for the adherence of Plasmodium falciparum-infected erythrocytes

Biochimica et Biophysica Acta 1701 (2004) 109 – 119 www.bba-direct.com

Chondroitin sulfate proteoglycans of bovine cornea: structural characterization and assessment for the adherence of Plasmodium falciparum-infected erythrocytes Rajeshwara N. Achura, Arivalagan Muthusamya, Subbarao V. Madhunapantulaa, Veer P. Bhavanandana, Clement Seudieub, D. Channe Gowdaa,* a

Department of Biochemistry and Molecular Biology, Pennsylvania State University College of Medicine, 500 University Drive, Hershey, PA 17033, USA b Department of Biochemistry and Molecular Biology, Georgetown University Medical Center, Washington, DC 20007, USA Received 4 February 2004; received in revised form 27 May 2004; accepted 17 June 2004 Available online 14 July 2004

Abstract The structures of the bovine corneal chondroitin sulfate (CS) chains and the nature of core proteins to which these chains are attached have not been studied in detail. In this study, we show that structurally diverse CS chains are present in bovine cornea and that they are mainly linked to decorin core protein. DEAE-Sephacel chromatography fractionated the corneal chondroitin sulfate proteoglycans (CSPGs) into three distinct fractions, CSPG-I, CSPG-II, and CSPG-III. These CSPGs markedly differ in their CS and dermatan sulfate (DS) contents, and in particular the CS structure—the overall sulfate content and 4- to 6-sulfate ratio. In general, the CS chains of the corneal CSPGs have low to moderate levels (15–64%) of sulfated disaccharides and 0–30% DS content. Structural analysis indicated that the DS disaccharide units in the CS chains are segregated as large blocks. We have also assessed the suitability of the corneal CSPGs as an alternative to placental CSPG or the widely used bovine tracheal chondroitin sulfate A (CSA) for studying the structural interactions involved in the adherence of Plasmodium falciparum-infected red blood cells (IRBCs) to chondroitin 4-sulfate. The data demonstrate that the corneal CSPGs efficiently bind IRBCs, and that the binding strength is either comparable or significantly higher than the placental CSPG. In contrast, the IRBC binding strength of bovine tracheal CSA is markedly lower than the human placental and bovine corneal CSPGs. Thus, our data demonstrate that the bovine corneal CSPG but not tracheal CSA is suitable for studying structural interactions involved in IRBC-C4S binding. D 2004 Elsevier B.V. All rights reserved. Keywords: Bovine cornea; Chondroitin sulfate proteoglycan; Decorin; Characterization; Plasmodium falciparum binding

1. Introduction Abbreviations: IRBCs, infected red blood cells; GAG, glycosaminoglycan; CS, chondroitin sulfate; PG, proteoglycan; CSPG, chondroitin sulfate proteoglycan; C4S, chondroitin 4-sulfate; DS, dermatan sulfate; DSPG, dermatan sulfate proteoglycan; BSA, bovine serum albumin; HexN, hexosamine; GalN, galactosamine; GlcN, glucosamine; PVDF, polyvinyledene difluoride; DEAE, diethylaminoethyl; GdnHCl, guanidine hydrochloride; NEM, N-ethylmaleimide; PBS, phosphate buffered saline; PMSF, phenylmethylsulfonyl fluoride; TLCK, N-a-tosyl-l-lysine chloromethyl ketone; TPCK, N-tosyl-l-phenylalanine chloromethyl ketone; PAGE, polyacrylamide gel electrophoresis; CAPS, 3-(cyclohexylamino)propanesulfonic acid * Corresponding author. Tel.: +1 717 531 0992; fax: +1 717 531 7072. E-mail address: [email protected] (D. Channe Gowda). 1570-9639/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.bbapap.2004.06.007

The proteoglycans of the corneal stroma have been reported to comprise chondroitin sulfate/dermatan sulfate proteoglycans (CS/DSPGs) and keratan sulfate proteoglycans (KSPGs) [1,2]. Approximately 99% of the uronic-acidcontaining proteoglycan of the bovine cornea has been reported to represent decorin and the remainder biglycan [1,3–5]. Decorin accounts for ~40% of the total proteoglycan in the cornea [3]. Although decorins from various species and different tissues widely differ with respect to the overall proteoglycan structures due to variation in the structural features of their glycosaminoglycan (GAG) chains, all share

110

R.N. Achur et al. / Biochimica et Biophysica Acta 1701 (2004) 109–119

a common ~39-kDa core protein. Another common feature of decorin, regardless of its source, is the presence of a single GAG chain, CS, DS, or CS/DS copolymer, attached to the conserved Ser4 residue at the NH2-terminal end [6]. The decorin core protein consists of 10 leucine-rich tandem repeats flanked by two highly conserved cysteine disulfide loops. Each tandem repeat is about 25 amino acids in length and has the characteristic pattern LXXLXLXXNX(LX) [7]. Decorins are thought to have important roles in fibrillogenesis, tissue repair, inhibition of cell proliferation, induction of matrix metalloproteinase expression, and binding and regulation of TGF-h [7,8]. Decorin occurs in bone and in all soft connective tissues as a component of tissue matrices. While the GAG chains of bone decorin are chondroitin sulfate, those of the soft connective tissue decorins of various animals have been described as DS or CS/DS chains. Although bovine cornea has been reported to be a decorin CS/DSPG [1,2], information on the structural details of the GAG chains is lacking. Further, two early studies have reported the presence of chondroitin and low sulfated chondroitin in the bovine cornea [9,10]. However, whether the low sulfated chondroitin exist as free GAGs or as proteoglycans has not been studied. In this study, comprehensive structural analysis revealed that bovine cornea contains structurally diverse chondroitin sulfate proteoglycans (CSPGs), bearing CS with very low to moderate sulfate contents. The CS chains of the corneal CSPGs differ markedly in the levels of 4- and 6-sulfate groups and in DS content. The DS in the corneal CSPG is present as large segregated blocks. Our results also reveal that the previously reported chondroitin-like chains are exclusively linked to decorin core protein, forming unusually low sulfated decorin CSPGs. In this study, we have also investigated the usefulness of the corneal CSPGs for studying the structural interactions involved in the adherence of Plasmodium falciparum-infected red blood cells (IRBCs) to chondroitin 4-sulfate (C4S). Infection with P. falciparum during pregnancy leads to the selective adherence of IRBCs in the placenta, causing placental malaria [11–15]. The IRBC adherence in the placenta is mediated by unusually low sulfated CSPGs [16–18]. A cytoadherence assay employing CSPGs from various sources has been used to define the C4S structural requirements for IRBC adherence and to measure the C4S-IRBC adhesion inhibitory antibodies in pregnant women exposed to malaria [11,16– 19]. The CSPGs used in those studies include a commercially available bovine tracheal chondroitin sulfate A (CSA), a recombinant human thrombomodulin-CSPG (TM-CSPG) expressed in human kidney 293 cells, and CSPGs purified from human placenta. While placental CSPGs are the best for studying the structural interactions involved in C4S-IRBC adhesion, it is rather tedious to purify these CSPGs and access to human placenta is limited compared to animal tissues.

Because of this limitation and ready availability of bovine tracheal CSA commercially, this CSA has been used widely for IRBC adherence assays. Although this material supports the adherence of IRBCs under both static and flow conditions and inhibits adhesion of IRBCs to placental tissue, it is not the best material for studying the C4S structural interactions with IRBCs. The major disadvantage of bovine tracheal CSA is the presence of high 6-sulfate content, which is known to adversely affect IRBC adhesion [18,20,21]. Moreover, bovine tracheal CSA poorly coats onto plastic because it represents CSA chains linked to a single amino acid or short peptides [20]. Consequently, the IRBC binding capacity of bovine tracheal CSA is markedly lower compared to the natural receptor [20]. Recently, we found that the CS chains of the recombinant TM-CSPG has 87% 4-sulfate and its IRBCbinding strength is ~10-fold lower than that of the placental CSPG (Muthusamy et al., unpublished results). Further, the recombinant TM-CSPG is not readily available. In this regard, we reasoned that the bovine corneal CSPGs bearing the CS chains with low 4-sulfate content, comparable to the placental CSPG, might exhibit IRBCbinding characteristics similar to that of the placental CSPG. Furthermore, bovine cornea is readily available and the purification of CSPGs is relatively easy compared to human placental CSPGs. Our data demonstrate that the corneal CSPGs can efficiently bind IRBCs, and that the binding strength is either comparable or higher than that of the placental CSPGs. In contrast, the IRBC binding strength of tracheal CSA is markedly lower than the placental CSPGs. Thus, the bovine corneal CSPGs are better alternatives to the widely used bovine tracheal CSA for studying IRBC structural interactions.

2. Materials and methods 2.1. Materials Bovine eyes were obtained from a local slaughterhouse immediately after sacrifice, cornea were dissected and stored frozen at 80 8C until used. Protease-free chondroitinase ABC (Proteus vulgaris), chondroitinase AC-II (Arthrobacter aurescens), and chondroitinase B (Flavobacterium heparinum) were purchased from Seikagaku America (Falmouth, MA). Bovine tracheal CSA, PMSF, NEM, benzamidine, protein molecular weight standards for gel filtration were from Sigma. TLCK and TPCK were from Roche Molecular Biochemicals. Sepharose CL-4B, Sepharose CL-6B, DEAE-Sephacel, and blue dextran were from Amersham Pharmacia Biotech. Bio-Gel P-6, 4–15% gradient Tris–HCl polyacrylamide minigels, and protein molecular weight standards for SDS-PAGE were from Bio-Rad. HPLC-grade 6 N HCl, trifluoroacetic acid, and micro-BCA protein assay kit were from Pierce. Polystyrene Petri dishes (Falcon 1058) were from Becton-

R.N. Achur et al. / Biochimica et Biophysica Acta 1701 (2004) 109–119

Dickinson Labware. Rabbit antiserum LF-94 against the synthetic peptide with sequence corresponding to the IGPEEHFPEVPEC motif at the NH2-terminal end of the bovine decorin core protein and rabbit antiserum LF-96 against the synthetic peptide motif corresponding to the LPDLDSLPPTYSC sequence at the NH2-terminal end of the bovine biglycan core protein were generous gifts from Dr. Larry Fisher, Craniofacial and Skeletal Disease Branch, NIDR, NIH, Bethesda. 2.2. Isolation of proteoglycans from bovine cornea Bovine corneas (6 g) were cut into pieces, suspended in ice-cold 50 mM NaOAc, pH 5.8, containing 10 mM EDTA, 0.1 mM PMSF, 0.1 mM TLCK, 0.25 mM TPCK, 1 mM benzamidine, 0.1 mM NEM, and 6 M GdnHCl, and homogenized using a Polytron homogenizer (Brinkmann, Switzerland). The slurry was stirred at 4 8C overnight, centrifuged at 9000 rpm using Sorvall 5C centrifuge; and the supernatant was dialyzed and lyophilized. The material was dissolved in 50 mM sodium acetate, pH 5.8, containing 6 M GdnHCl, 10 mM EDTA, 0.1 mM PMSF, 0.1 mM TLCK, 0.25 mM TPCK, 1 mM benzamidine, 0.1 mM NEM, and 42% CsCl. The solution was centrifuged at 44,000 rpm in a Beckman 50 TI rotor at 14 8C for 48 h, and the fractions were collected from the bottom of the tubes into 13 equal fractions, and the density determined. Absorption at 260 and 280 nm was measured for all the fractions, and aliquots were assayed for uronic acid content [22]. The uronic-acid-containing fractions were pooled, dialyzed, and lyophilized.

111

2.5. Cesium bromide density gradient centrifugation of CSPGs The proteoglycans, recovered after Sepharose CL-4B chromatography, were dissolved in 25 mM sodium phosphate, pH 7.2, containing 4 M GdnHCl, 10 mM EDTA, and 42% CsBr. The solutions were centrifuged in a Beckman TI 70 rotor at 44,000 rpm for 48 h at 14 8C. The solutions were collected from the bottom into 13 equal fractions, absorption at 260 and 280 nm measured, and aliquots assayed for uronic acid contents. The uronic-acid-containing fractions were pooled, dialyzed, and lyophilized. 2.6. Removal of KSPGs from CSPGs The CSPGs from the CsBr density gradient centrifugation step were incubated with endo-h-galactosidase (2 U) in 70 Al of 50 mM Tris–HCl buffer, 1 AM calcium acetate, pH 7.5, at 37 8C for 10 h. The enzyme digests were chromatographed on Sepharose CL-6B columns (149 cm) to remove the enzyme and the digestion products. 2.7. Isolation of GAGs The purified CSPGs were treated with 0.1 M NaOH, 1 M NaBH4 at 45 8C for 18 h under nitrogen, and the released GAGs were recovered by chromatography on Sepharose CL-6B columns (149 cm) in 0.2 M NaCl as previously reported [17]. 2.8. SDS-PAGE and Western blot analysis

2.3. Fractionation of proteoglycans by ion-exchange chromatography The proteoglycan obtained as above was applied onto a DEAE-Sephacel column (114 cm) in 50 mM sodium acetate, 0.1 M NaCl, pH 5.8, washed with the same buffer, and then eluted with a linear gradient of 0.1–1.0 M NaCl. Fractions (1.3 ml) were collected, absorption at 260 and 280 nm was measured, and the uronic acid content determined [22]. The uronic acid-containing fractions were pooled, dialyzed, and freeze-dried. 2.4. Purification of proteoglycans by gel filtration on Sepharose CL-4B The proteoglycan fractions (2–6 mg), obtained by DEAE-Sephacel chromatography, were treated with heparitinase (1 U) as previously reported [17]. The reaction mixture was chromatographed on Sepharose CL-4B columns (1.583 cm) in 50 mM Tris–HCl, 0.2 M NaCl, pH 7.5, containing 6 M GdnHCl. Fractions (2 ml) were collected, the elution of proteoglycans monitored as above, and the uronic acid-containing materials pooled and recovered.

SDS-PAGE of the CSPGs and core proteins was performed using 4–15% gradient polyacrylamide minigels in the presence of 0.5 M 2-mercaptoethanol [23]. The gels were stained with Coomassie blue followed by Alcian blue and ammoniacal silver [24]. For Western blotting, the proteins on the gels, after SDS-PAGE, were electrotransferred onto PVDF membranes. The membranes were blocked with 1% BSA in 50 mM Tris–HCl, 150 mM NaCl, pH 8.0, containing 0.1% Tween 20, and probed with the antidecorin and antibiglycan antisera. The bound antibodies were visualized using alkaline-phosphatase-conjugated goat antirabbit IgG as the secondary antibody and nitroblue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate as color developing reagent. 2.9. Carbohydrate composition analysis The CSPGs (20 Ag) were hydrolyzed with 4 M HCl at 100 8C for 6 h or with 2.5 M trifluoroacetic acid at 100 8C for 5 h. The hydrolysates were dried in a Speed-Vac and analyzed on a CarboPac PA1 high pH anion-exchange column (4250 mm) using Dionex BioLC HPLC coupled

112

R.N. Achur et al. / Biochimica et Biophysica Acta 1701 (2004) 109–119

to a pulsed amperometric detector [25]. The elution was performed with 18 mM NaOH. 2.10. Disaccharide composition analysis of the GAG chains The purified CSPGs (50 Ag) were digested with chondroitinase ABC (20 mU) in 50 Al of 100 mM Tris– HCl, 30 mM NaOAc, pH 8.0, containing 0.01% BSA at 37 8C for 5 h [26]. The released unsaturated disaccharides were analyzed by HPLC on a 4.6250 mm amine-bonded silica PA03 column as previously reported [17]. 2.11. Analysis of DS content in GAGs The GAGs (~100 Ag) were treated with chondroitinase AC-II (200 mU) in 50 Al of 100 mM NaOAc, pH 6.0, containing 0.01% BSA at 37 8C for 30 min [27] or with chondroitinase B (20 mU) in 50 Al of 50 mM Tris–HCl, pH 8.0, containing 0.05% BSA at 30 8C for 2 h [28]. The enzyme digests were analyzed on Bio-Gel P-6 columns (149 cm) in 0.1 M pyridine/0.1 M acetic acid.

2.13. Parasites The placental isolate used for the adhesion assay was from a blood sample collected from a P. falciparum-infected term placenta of a Cameroonian woman, who delivered the baby at the Central Hospital, Yaounde, Cameroon. After washing the placental blood to remove plasma, the cells were suspended in RPMI 1640 incomplete medium and centrifuged on a cushion of Ficoll-hypaque to remove the mononuclear cells. The cell pellet containing IRBCs and RBCs was cultured in RPMI 1640 medium containing10% O-positive human serum [18]. After one life cycle, the parasites at the early trophozoite stages were cryopreserved in glycerolyte solution at 80 8C, transported, and stored at 80 8C until used. C4S-adherent FCR-3 and 3D7 laboratory strains were obtained by panning on plastic Petri dishes coated with placental CSPG. These parasites were cultured in RPMI 1640 medium as above, and the IRBCs used within 10 days of selection. 2.14. IRBC binding and inhibition assays

2.12. NH2 terminal sequencing and mass spectral analysis of the core proteins The CSPGs (100 Ag) were treated with protease-free chondroitinase ABC (30 mU) and the released core proteins were electrophoresed on 4–15% gradient SDSpolyacrylamide gels (25–30 Ag/well) under reducing conditions [23]. The proteins were electroblotted onto PVDF membranes using 10 mM CAPS, pH 11.0, containing 10% methanol. The membranes were stained with Ponceau S, the protein bands were cut out and used for NH2-terminal sequencing, which was performed by the Research Resource Facility at the Pennsylvania State University College of Medicine, and by Professor Keiichi Takagaki, Department of Biochemistry, Hirosaki University School of Medicine, Hirosaki, Japan. For mass spectral analysis, 50 Ag of the CSPG core protein, released by chondroitinase ABC, was electrophoresed on 4–15% gradient SDS-polyacrylamide gel under reducing conditions. The gel was briefly stained with Coomassie blue, the protein band was excised, destained, and dried by Speed-Vac. The gel was suspended in 100 Al of 10 mM DTT in 25 mM ammonium bicarbonate, pH 8.0, for 15 min at 37 8C, and S-alkylated with 20 mM iodoacetamide, and dried by Speed-Vac. The dried gel was rehydrated in 20 Al of 0.02 Ag/Al sequencing grade trypsin in 50% acetonitrile, 40 mM ammonium bicarbonate, pH 8.0, at room temperature for 1 h, and then added 50 Al of the above enzyme solution, and incubated at 37 8C for 16–18 h. The solution was desalted using ZipTip SCX tips, and analyzed using the Applied Biosystems 4700 Proteomics Analyzer MALDI-TOF mass spectrometer by the Research Resource Facility of the Pennsylvania State University College of Medicine.

The purified CSPGs in PBS, pH 7.2, were coated, at the indicated concentrations, onto plastic Petri dishes as circular spots (0.4-cm diameter), blocked with 2% BSA, overlaid with a 2% suspension of parasite culture (20–30% parasitemia) in PBS, pH 7.2. The unbound cells were washed, and the bound cells fixed with 2% glutaraldehyde, stained with Giemsa, and counted under light microscope [18]. For adhesion–inhibition assays, 4% suspension of parasite culture was preincubated with equal volume of 80 ng/ml to 160 Ag/ml of C4S containing 36% 4-sulfate, obtained by the regioselective 6-O-desulfation of bovine tracheal chondroitin sulfate, in PBS, pH 7.2. The cell suspensions were then layered on CSPG-coated and BSA-blocked spots on Petri dishes as above. The bound cells were fixed with 2% glutaraldehyde, stained with Giemsa, and counted using light microscope.

3. Results 3.1. Isolation and purification of CSPGs of bovine cornea Upon CsCl density gradient centrifugation of the 6 M GdnHCl extracts of the bovine cornea, the uronic acidcontaining proteoglycans were distributed throughout the gradient, except the top 20%, with peak concentration in the middle of the gradient. The material obtained by combining the uronic-acid-containing fractions had relatively high levels of both galactosamine and glucosamine. The material was resistant to heparitinase, suggesting that the uronicacid-containing proteoglycan is either CSPG or DSPG. These results are in agreement with the previous findings

R.N. Achur et al. / Biochimica et Biophysica Acta 1701 (2004) 109–119

that corneas of various animals including human contain predominantly CS/DSPGs and KSPGs. Upon DEAESephacel chromatography using a salt gradient, the uronicacid-containing proteoglycans eluted as a broad unresolved peak without clear separation into distinctive species (Fig. 1). The proteoglycan-containing fractions were pooled into three fractions, I, II, and III as indicated in Fig. 1, and chromatographed on Sepharose CL-4B to remove the contaminating proteins. In each case, a substantial amount of protein was separated, and the uronic acid-containing proteoglycans eluted predominantly as a single peak (Fig. 2). The proteoglycans were then purified by CsBr density gradient centrifugation. All the proteoglycan fractions sedimented at moderate, but distinctly different, density regions with q=1.33, 1.36, and 1.39, respectively (Fig. 3). Fractions I, II, and III had, respectively, 14%, 44%, and 76% glucosamine, suggesting the presence of high levels of KSPGs in addition to uronic acid-containing proteoglycan fractions. To remove the KS chains, the proteoglycan fractions were treated with endo-h-galactosidase and purified on Sepharose CL-6B column. In each case, the uronic-acidcontaining proteoglycan eluted as a single peak, separated from the core proteins and N-acetylglucosamineh(1–3)galactose disaccharide formed from the KS chains by the action of endo-h-galactosidase (not shown). The purified proteoglycan fractions I, II, and III were designated as CSPG-I, CSPG-II, and CSPG-III, respectively. Hexosamine analysis revealed that the CSPG fractions contained predominantly galactosamine (Table 1), and they were resistant to further action of endo-h-galactosidase, suggesting the complete removal of KSPGs. 3.2. Characterization of GAG chains of the corneal CSPGs The GAG chains of CSPG-I, CSPG-II, and CSPG-III were released by h elimination and analyzed for their size and composition. In each case, the GAGs were eluted as two

113

Fig. 2. Purification of the bovine corneal CSPGs by size-exclusion chromatography. The CSPGs isolated by DEAE-Sephacel chromatography were chromatographed on Sepharose CL-4B columns (1.583 cm) in 50 mM Tris–HCl, 0.2 M NaCl, pH 7.5 containing 6 M GdnHCl. The elution of CSPG-I (A), CSPG-II (B), and CSPG-III (C) were monitored for uronic acid ( ) and for protein (o). The CSPG fractions were recovered by pooling the fractions as indicated by the horizontal bars. The elution positions of blue dextran (BD), bovine serum albumin (BSA), and glucose (Glc) are indicated.

.

partially resolved peaks on a calibrated Sepharose CL-6B column, indicating the presence of two distinct populations of CS chains with the molecular mass ~70 and 45 kDa (data not shown). The proportions of these GAGs differed significantly for the three CSPG fractions; ~3:1, ~2.2:1,

Fig. 1. Fractionation of bovine corneal CSPGs by ion-exchange chromatography. Bovine cornea CSPGs were chromatographed on DEAE-Sephacel columns (1.014 cm) in 50 mM sodium acetate buffer, pH 5.8, and the bound CSPGs were eluted with the same buffer with a linear gradient of 0.1–1.0 M NaCl. The elution profiles of the CSPGs were monitored by the uronic acid ( ) assay and by measuring OD at 280 nm for protein content (o). The fractions containing uronic acid were pooled as indicated by the horizontal bars.

.

114

R.N. Achur et al. / Biochimica et Biophysica Acta 1701 (2004) 109–119 Table 2 Yield and composition of bovine corneal CSPGs CSPG fraction

CSPG-I CSPG-II CSPG-III

Core protein type

Dermatan sulfate (%)a

Decorin Decorin Decorin

0 15 30

Disaccharide composition (%mol)b Ddi-0S

Ddi-4S

Ddi-6S

85 69 36

14 28 51

1 3 13

a

Based on percent 4-sulfated di- and tetrasaccharides produced upon digestion of the CSPGs with chondroitinase B. b Determined by HPLC analysis of unsaturated disaccharides released by the digestion of the GAG chains with chondroitinase ABC.

Fig. 3. Purification of bovine corneal CSPGs by CsBr density gradient centrifugation. The CSPGs from Sepharose CL-4B chromatography (Fig. 2) were dissolved in 25 mM sodium phosphate, pH 7.2, containing 4 M GdnHCl, 10 mM EDTA, and 42% CsBr. The solutions were centrifuged in a Beckman TI 70 rotor at 44,000 rpm for 48 h at 14 8C. The gradients were collected into 13 equal fractions from the bottom of the tube, and aliquots were monitored for uronic acid ( ) and protein (o). The density (D) of each fraction was determined by weighing 100 Al aliquot. Fractions containing the CSPG-I (A), CSPG-II (B), and CSPG-III (C) were pooled as indicated by the horizontal bars based on the uronic acid content.

.

and ~1.8:1, for the CS chains of CSPG I, CSPG II, and CSPG-III, respectively. The disaccharide compositions of GAGs of the CSPGs were determined by HPLC after treatment with chondroitinase ABC. The GAG chains of CSPG-I, CSPG-II, and CSPG-III consisted of, respectively, 14%, 28%, and 51% 4sulfated, 1%, 3%, and 13% 6-sulfated, and 85%, 69%, and 36% nonsulfated disaccharides (Table 2). The presence of the CS chains with a wide range (very low to moderate levels) of sulfation is also evident from the observed distribution of decorin during CsCl and CsBr density gradient centrifugations because the major portion of Table 1 Yield and composition of CSPGs purified from bovine cornea CSPGs

CSPG-I CSPG-II CSPG-III a b c d

Yielda (mg/100 g tissue weight) 9 20 16

Composition (wt.%)

HexN (%mol ratio)

Proteinb

Uronic HexNd acidc

GalNd GlcNd

36 30 29

30 29 24

86 94 91

Yield of PGs after final purification. By the micro-BCA protein assay. By the carbazole method. By the high pH anion-exchange HPLC.

32 33 36

14 6 9

CSPGs fractionated to low and to the moderate density regions rather than the high-density region as observed for most other CSPGs. To determine the DS content, the CS chains of the CSPGs were separately treated with chondroitinase AC II and chondroitinase B. In the case of CSPG-I, the GAGs were quantitatively converted into disaccharides by chondroitinase AC II, and were totally resistant to the action of chondroitinase B (not shown). In contrast, approximately 85% of the GAGs of CSPG-II were digested into disaccharides by chondroitinase AC II, and ~15% remained as high molecular weight fragments (not shown). Consistent with these results, ~14% of the GAGs were degraded into di- and tetrasaccharides by chondroitinase B, and the remainder resisted the enzyme action. In the case of CSPG-III, ~70% of the GAG chains were digested by chondroitinase AC II, and ~30% was resistant (not shown). Conversely, when treated with chondroitinase B, ~30% of the GAGs were digested into di- and tetrasaccharides and ~70% remained as high molecular weight material (not shown). These results demonstrate that the GAGs of CSPGI were exclusively CS, whereas those of CSPG-II and CSPG-III have, in addition to CS, ~15% and ~30% of DS, respectively (Table 2). 3.3. Analysis of the core proteins of corneal CSPGs The CSPG fractions and core proteins released by treatment of the CSPGs with chondroitinase ABC were analyzed by SDS-PAGE. In each case, the CSPG showed a broad, diffused band in the molecular mass range of 115– 250 kDa (Fig. 4A and data not shown). In all the three cases, the core proteins had a molecular mass of ~46 kDa, although they differ significantly with respect to the GAG chain composition. The identity of core proteins of the CSPG fractions was determined by Western blot analysis using antibodies raised against peptide motifs of the bovine decorin and biglycan core proteins. The core proteins of all the three CSPG fractions strongly reacted with antidecorin antibodies (Fig. 4B, and data not shown). When probed with antibodies against the biglycan peptide motif, a faint ~45-kDa biglycan band was observed (Fig. 4C, and not shown). Because a

R.N. Achur et al. / Biochimica et Biophysica Acta 1701 (2004) 109–119

115

To conclusively identify the core proteins of corneal CSPGs, the NH2-terminal sequencing and mass spectral analysis were performed. The 10-amino-acid sequences of the NH2-terminal stretches of the corneal CSPG core proteins were found to be DEASGIGPEE. This sequence is identical to the previously reported NH2-terminal amino acid sequence of the bovine decorin [8,29]. Further, the MALDI-TOF mass spectrometry showed that the peptides obtained by treatment of the major corneal CSPG core protein with trypsin were identical to the corresponding peptide motifs in the bovine decorin core protein (Table 3) [30].

Fig. 4. SDS-PAGE and Western blot analysis of bovine corneal CSPG. (A) The purified bovine corneal CSPGs were electrophoresed on 4–15% gradient polyacrylamide gel, and stained with Coomassie blue followed by Alcian Blue. Lane 1, untreated CSPG-III (20 Ag); lane 2, chondroitinase ABC treated CSPG-III (25 Ag). (B) Western blot analysis of CSPG treated with chondroitinase ABC using bovine antidecorin antisera. Lane 1, CSPGII (25 Ag); lane 2, CSPG-III (25 Ag). (C) Western blot analysis of CSPG treated with chondroitinase ABC using bovine antibiglycan antisera. Lane 1, CSPG-II (25 Ag); lane 2, CSPG-III (25 Ag). The position of the molecular mass (kDa) marker proteins is indicated to the right. The CSPG-I was also similarly analyzed (not shown).

standard bovine biglycan is not available, it was not possible to accurately quantify the level of biglycan in the purified corneal CSPGs fractions. However, based on the relative intensities of protein bands on Western blots when antidecorin and antibiglycan antibodies were used at similar dilutions, we estimate that the amount of biglycan in the corneal CSPG fractions is b5%. Thus, the CSPGs of bovine cornea consist of N95% decorin and b5% biglycan.

Table 3 Mass spectrometry of the peptides obtained by the trypsin digestion of the bovine corneal CSPG core protein Peptide

1 2 3 4 5 6 7 8 9 10

Calculated mass

873.3818 1292.7684 1497.8787 1106.6567 1454.8035 2224.6694 1325.6154 1045.6251 981.5112 963.5258

Observed mass

873.3674 1292.7037 1497.8043 1106.6051 1454.7646 1224.6133 1325.6024 1045.5262 981.4663 963.4497

Determined peptide sequence

CQCHLR NLHTLILINNK ISPGAFAPLVKLER LYLSKNQLK NQLKELPEKMPK VHENEITKVR SSGIENGAFQGMK ITKVDAASLK ELHLNNNK AAVQLGNYK

Positions of amino acids in the decorin core protein sequencea 59–64 107–117 121–134 135–143 140–151 158–167 189–201 234–243 274–281 352–360

a The amino acid residues corresponding to the bovine decorin amino acid sequence [30].

Fig. 5. P. falciparum IRBC adherence to bovine corneal CSPGs. The bovine corneal CSPG-I ( ), CSPG-II (n), and CSPG-III (5); the human placental CSPGs (o), and bovine tracheal CSA (E), in PBS, pH 7.2, were coated at the indicated concentrations as 4-mm-diameter circular spots on plastic Petri dishes. The spots were blocked with 2% BSA and overlaid with IRBCs. After washing the unbound cells, the bound IRBCs were fixed with 2% glutaraldehyde, stained with 1% Giemsa, and counted under light microscope. The assays were carried out three times each in duplicate and the results are presented as meanFS.E. (n=6). (A) IRBCs obtained from P. falciparum-infected placenta. (B) C4S-adherent IRBCs from FCR-3 laboratory parasites. C4S-adherent IRBCs from 3D7 strain were also studied, and the IRBC-binding pattern was similar to that of the C4Sadherent FCR-3 parasites. Therefore, the results from 3D7 parasites are not included in the figure. The number of IRBCs bound per square millimeter of CSPG-coated plates was in the range of 3000–8000 depending upon the parasitemia of cell suspension used in the assay.

.

116

R.N. Achur et al. / Biochimica et Biophysica Acta 1701 (2004) 109–119

3.4. Assessment of CSPGs for P. falciparum IRBC binding To test whether the corneal CSPGs are suitable for studying the structural interactions involved in the C4SIRBC adherence, the purified CSPGs were analyzed for their ability to bind IRBCs. The binding assays were performed using IRBCs obtained from a P. falciparuminfected placenta and IRBCs from C4S-adherent 3D7 and FCR-3 laboratory isolates. In the case of laboratory isolates, the IRBCs were used within 10 days of selection on placental CSPGs since the binding affinity of IRBCs gradually decreases over successive life cycles. All the three corneal CSPG fractions, coated on plastic dishes, bound IRBCs from both placental and laboratory parasite isolates in a dose-dependent manner (Fig. 5). The IRBC binding was completely abolished on treatment of the CSPG-coated plates with chondroitinase ABC, and was efficiently inhibited by C4S (not shown). These results demonstrate that IRBC binding by corneal CSPGs is

mediated through their GAG chains, and agree with the known specificity of IRBC binding to C4S. To determine the efficiency of IRBC binding by corneal CSPGs, we studied, in parallel, the IRBC binding abilities of the natural receptor, the low sulfated CSPGs of human placenta, and the widely used bovine tracheal CSA. The number of IRBCs bound per unit area on plates coated with the bovine corneal CSPG-I was comparable to that of the human placental CSPG and substantially higher than that of bovine tracheal CSA at various concentrations tested (Fig. 5). The IRBC binding by corneal CSPG-II and CSPG-III was 15–20% higher than that by placental CSPG (Fig. 5). Further, the IRBC binding density to the respective CSPG was similar irrespective of the source of IRBCs, placental or laboratory selected parasites, used in this study (compare Fig. 5A with B). The number of IRBCs that bound to CSPGs was dependent on the assay conditions used, including the extent of CSPG coated on the plastic surface, parasitemia of

Fig. 6. C4S-IRBC adherence inhibition analysis. The human placental CSPG (black bar, 0.2 Ag/ml) and bovine corneal CSPG-I (white hatched bar, 0.2 Ag/ml), CSPG-II (black hatched bar, 0.2 Ag/ml), and CSPG-III (white bar, 0.2 Ag/ml), and bovine tracheal CSA (gray bar, 10 Ag/ml), in PBS, pH 7.2, were coated and the spots were blocked. The IRBCs were preincubated with the indicated concentrations of C4S, and IRBC-adhesion inhibition analysis was performed as described in Fig. 5. (A) IRBCs obtained from P. falciparum-infected placenta. (B) C4S-adherent IRBCs from FCR-3 laboratory parasites. C4S-adherent IRBCs from 3D7 strains were also studied, and the IRBC-adhesion inhibition pattern was similar to that of the C4S-adherent FCR-3 parasites (not shown in the figure).

R.N. Achur et al. / Biochimica et Biophysica Acta 1701 (2004) 109–119

cell suspension used and washing procedures. Thus, the IRBCs bound per unit area alone do not truly reflect the IRBC binding efficiency. Therefore, the IRBC binding strength was assessed at the saturated CSPG-coating concentrations by the adhesion inhibition analysis using a partially sulfated C4S with 36% 4-sulfate groups. This C4S was prepared by the regioselective 6-O-desulfation of bovine tracheal CSA and has been shown to efficiently inhibit IRBC adhesion to placental CSPG [18]. The corneal CSPGs exhibited either comparable or 2-fold higher binding strengths than the placental CSPG with all the three IRBC isolates used and at all concentrations of the inhibitor tested (Fig. 6). Among the corneal CSPGs, however, CSPG-II and CSPG-III exhibited similar IRBC-binding strengths, and their binding strengths were significantly higher than that of CSPG-I. Thus, these results indicate that the CS chains of corneal CSPGs effectively interact with IRBCs. The binding strength of the commercially available bovine tracheal CSA was substantially lower than those of placental and corneal CSPGs, as measured by the concentration of C4S required for the inhibition of IRBC binding to similar levels (Fig. 6). The observed difference in IRBC-binding strength among the CSPGs studied was not dependent on the parasites used, because both placental and the laboratory-selected parasites showed similar results (compare Fig. 6A with B).

4. Discussion The results of this study provide a comprehensive picture of the structures of the bovine corneal CSPGs. The majority of the CS chains of bovine cornea, including the previously reported unusually low sulfated chondroitin-like chains are linked mainly to the decorin core protein and minor proportions of the CS chains are attached to the biglycan core protein. The bovine corneal CSPGs are highly heterogeneous with regard to the sulfate and DS content of their GAG chains. Based on the results of ion-exchange chromatography, there are at least three distinct populations of CSPGs in bovine cornea, CSPG-I, CSPG-II, and CSPGIII. All the three CSPG species have identical core proteins with predominantly decorin and low proportion of biglycan, but markedly differ in their GAG chain structures. A striking difference in their CS chain structures is the extent of sulfated disaccharides, very low (~15%) in CSPG-I, and moderate, ~31% and ~64%, respectively, in CSPG-II and CSPG-III. The presence of the CS chains with a wide range of sulfate contents is likely to be responsible for the observed heterogeneity in the overall negative charge of CSPGs. Other differences in the CS chain structures of corneal CSPG fractions are the extent of 6-sulfation and DS content. The CS chains of CSPG-I and CSPG-II fractions are predominantly 4-sulfated with only 1% and 3% 6sulfated disaccharides, respectively, whereas CSPG-III contains ~13% 6-sulfate groups. Regarding the DS content, the GAGs of CSPG-I completely lacks DS, while those of

117

the CSPG-II and CSPG-III fractions have significant levels of DS, ~15% and ~30%, respectively. The CSPG-I bearing the unusually low sulfated chondroitin-like CS chains accounts for about 20% of the total CSPGs of bovine cornea (Table 1). Therefore, it appears likely that the CS chains of this CSPG fraction accounts for the previously reported chondroitin-like chains of the bovine cornea with 2.1% and 13.1% sulfate by weight [9]. CSPG-I lacks DS structures as indicated by the complete resistance of the GAG chains to chondroitinase B and the quantitative conversion into disaccharides by chondroitinase AC II. This is in contrast to the current belief that only bone decorin carries chondroitin sulfate, and that those of the soft connective tissues have CS/DS or DS chains. The GAG chains of CSPG-II and CSPG-III, however, contain 15% and 30% DS, respectively, and moderate levels (31% and 64%) of sulfate groups. Previous studies have reported a DS content of 30–40% for bovine, rabbit, and monkey corneal decorin GAG chains [2,3,5]. When these data are compared with the results of this study, it is obvious that the previous studies selectively purified the decorins with relatively high levels of DS. In any event, because our study has considered all of the CSPG types present in the tissue, it is clear that corneal CSPGs vary widely with regard to the level of sulfation and in the extent to which the glucuronic acid residues are epimerized to iduronic acid residues. Although the GAGs of various CS/DSPGs of animal tissues are heterogeneous with regard to the chain length and level of sulfation and CS/DS ratios, the variation within a single tissue type is not as much as that is observed here for the bovine corneal decorin. Therefore, it appears that the observed wide variation in the sulfation and epimerization patterns in CSPG-I, CSPG-II, and CSPG-III are characteristics of cornea and these features may have tissue-specific biological relevance. The DS disaccharides in the bovine corneal decorin GAGs are present mainly as consecutive residues to form fairly large blocks of DS within the CS structure. This conclusion is supported by the results of chondroitinase AC II and chondroitinase B digestion. Approximately one half of the GAG chains of both CSPG-II and CSPG III were converted into disaccharides by chondroitinase B and the other half resisted chondroitinase AC II digestion. Because, in both CSPG-II and CSPG-III, the chondroitinase AC IIresistant portion of the GAG chains were relatively large molecular weight fragments (excluded on Bio-Gel P-6), it is evident that these correspond to the areas of GAG chains consisting of consecutive residues of iduronic acid, forming blocks of DS structures. In this study, the adherence of IRBCs to the purified corneal CSPG fractions was evaluated, with regard to the ability of IRBCs to bind CSPG and IRBC-binding strength, using parasites obtained from P. falciparum-infected placenta as well as parasites selected for C4S-adhesion from two different laboratory strains. All the three CSPG

118

R.N. Achur et al. / Biochimica et Biophysica Acta 1701 (2004) 109–119

fractions of bovine cornea could bind both placental and laboratory-selected IRBCs through their CS chains at a density either similar or higher than the placental CSPGs [17]. Within the three CSPGs fractions, the CSPG-II and CSPG-III showed significantly higher binding compared to CSPG-I. As discussed below, this is probably due to the presence of a higher number of IRBC-binding sites in the former compared to the latter (see below). When the number of IRBCs that bound to the corneal CSPGs were compared to those bound to the bovine tracheal CSA, the difference is much more pronounced—at least 2-fold more binding to the corneal CSPGs compared to the tracheal CSA. The efficient binding of IRBCs by bovine corneal CSPGs was also evident by differences in the IRBC-binding strengths. As shown in Fig. 6, the concentration of C4S required for the inhibition of IRBC binding to comparable levels is either similar or about 2-fold higher for the corneal CSPGs than the placental CSPGs. In agreement with the observed higher number of IRBCs bound to CSPG-II and CSPG-III compared to CSPG-I, the binding strength of the former is significantly higher than the latter. Furthermore, it is clear from the results shown in Fig. 6 that the IRBC-binding strength of tracheal CSA is strikingly lower than the human placental and bovine corneal CSPGs. The results of this study also demonstrate that the binding strength of IRBCs is not parasite strain specific. As shown in Fig. 6, the IRBCs obtained from P. falciparuminfected placentas as well as two different laboratory isolates exhibited comparable IRBC binding strengths. These results are consistent with the previous finding that the antibodies in sera from pregnant women from different endemic regions of the world can inhibit the adhesion of IRBCs from different parasite isolates, implying that the parasite ligand expressed on the IRBC surface is highly conserved [31]. The CS chains of the bovine corneal CSPG appear to contain significantly more IRBC-binding sites than the CS chains of placental intervillous space CSPG, the natural receptor for the P. falciparum IRBC adherence in the placenta. In the case of the placental CSPGs, the CS chains contain 8–10% sulfated disaccharides groups [17], and the IRBC binding is mediated by a dodecasaccharide motif containing two 4-sulfated and four nonsulfated disaccharide units [18]. Furthermore, the sulfate groups in the CS chains of placental CSPGs are clustered in certain regions to a density of ~25%, and these sulfate-rich domains containing approximately two sulfate groups per dodecasaccharide motif provide sites for IRBC binding [32]. The GAGs of the bovine corneal CSPGs contain 13–21% of 4-sulfated disaccharides in the form of CS that are significantly higher compared to the ~8–10% of 4-sulfation in the placental CSPG. Therefore, the corneal GAGs can provide higher number of IRBC-binding dodecasaccharide motifs containing two to four 4-sulfated disaccharides. Additionally, as previously shown [18], DS can also bind IRBCs at a level similar to that by the fully 4-sulfated CS, i.e., ~50% binding

efficiency compared to the partially sulfated CS with 36% 4sulfated disaccharides. Thus, the DS blocks of the corneal decorin GAGs can also bind IRBCs, albeit at a 50% lower efficiency compared to the partially 4-sulfated CS, leading to an overall efficient IRBC binding by the GAG chains of corneal CSPGs. In conclusion, the results of this study indicate that the CSPGs of bovine cornea are heterogeneous mixture containing CS chains with very low to moderate level of sulfate groups. The corneal CSPGs can efficiently bind P. falciparum IRBCs with respect to binding strength as well as the number of IRBCs bound per unit area. A comparative analysis of IRBC binding and inhibition demonstrated that bovine corneal CSPG is a better alternative to the widely used bovine tracheal CSA.

Acknowledgements We thank Dr. Larry Fisher, Craniofacial and Skeletal Disease Branch, NIDR, NIH, Bethesda, for providing antidecorin and antibiglycan antisera; Dr. Bruce Stanley, Research Resource Facility, Pennsylvania State University, Hershey; and Professor Keiichi Takagaki, Department of Biochemistry, Hirosaki University School of Medicine, Hirosaki, Japan, for the mass spectral analysis. This work was supported by Public Health Service grant AI-45086 from the National Institute of Allergy and Infectious Diseases, NIH.

References [1] I. Axelsson, D. Heinegard, Characterization of the keratan sulfate proteoglycans from bovine corneal stroma, Biochem. J. 169 (1978) 517 – 530. [2] I. Axelsson, D. Heinegard, Characterization of chondroitin sulfate-rich proteoglycans from bovine corneal stroma, Exp. Eye Res. 39 (1980) 57 – 66. [3] J.D. Gregory, L. Coster, S.P. Damle, Proteoglycans of rabbit corneal stroma: isolation and partial characterization, J. Biol. Chem. 257 (1982) 6965 – 6970. [4] J.L. Funderburgh, G.W. Conrad, Isoforms of corneal keratan sulfate proteoglycan, J. Biol. Chem. 265 (1990) 8297 – 8303. [5] J.R. Hassell, D.A. Newsome, V.C. Hascall, Characterization and biosynthesis of proteoglycans of corneal stroma from rhesus monkey, J. Biol. Chem. 254 (1979) 12346 – 12354. [6] W. Li, J.P. Vergnes, P.K. Corneut, J.R. Hassell, cDNA clone to chick corneal chondroitin/dermatan sulfate proteoglycan reveals identity to decorin, Arch. Biochem. Biophys. 296 (1992) 190 – 197. [7] R.V. Iozzo, Matrix proteoglycans: from molecular design to cellular function, Annu. Rev. Biochem. 67 (1998) 609 – 652. [8] C.T. Brown, P. Lin, M.T. Walsh, D. Gantz, M.A. Nugent, V. TrinkausRandall, Extraction and purification of decorin from corneal stroma retain structure and biologic activity, Protein Expr. Purif. 25 (2002) 389 – 399. [9] E.A. Davidson, K. Meyer, Chondroitin, a new mucopolysaccharide, J. Biol. Chem. 211 (1954) 605 – 611. [10] K. Meyer, A. Linker, E.A. Davidson, B. Weismann, The mucopolysaccharides of bovine cornea, J. Biol. Chem. 205 (1953) 611 – 616.

R.N. Achur et al. / Biochimica et Biophysica Acta 1701 (2004) 109–119 [11] P.E. Duffy, M. Fried, Deadly parasite, susceptible host, in: P.E. Duffy, M. Fried (Eds.), Malaria in Pregnancy, Taylor and Francis Press, New York, 2001, pp. 1 – 245. [12] R.S. Bray, R.E. Sinden, The sequestration of Plasmodium falciparum infected erythrocytes in the placenta, Trans. R. Soc. Trop. Med. Hyg. 73 (1979) 716 – 719. [13] J.G. Beeson, N. Amin, M. Kanjala, S.J. Rogerson, Selective accumulation of mature asexual stages of Plasmodium falciparuminfected erythrocytes in the placenta, Infect. Immun. 70 (2002) 5412 – 5415. [14] I.A. McGregor, M.E. Wilson, W.Z. Billewicz, Malaria infection of the placenta in The Gambia, West Africa; its incidence and relationship to stillbirth, birth weight and placental weight, Trans. R. Soc. Trop. Med. Hyg. 77 (1983) 232 – 244. [15] C. Menendez, J. Ordi, M.R. Ismail, P.J. Ventura, J.J. Aponte, E. Kahigwa, F. Font, P.L. Alonso, The impact of placental malaria on gestational age and birth weight, J. Infect. Dis. 181 (2000) 1740 – 1745. [16] W. Chai, J.G. Beeson, A.M. Lawson, The structural motif in chondroitin sulfate for adhesion of Plasmodium falciparum-infected erythrocytes comprises disaccharide units of 4-O-sulfated and nonsulfated N-acetylgalactosamine linked to glucuronic acid, J. Biol. Chem. 277 (2002) 22438 – 22446. [17] R.N. Achur, M. Valiyaveettil, A. Alkhalil, C.F. Ockenhouse, D.C. Gowda, Characterization of proteoglycans of human placenta and identification of unique chondroitin sulfate proteoglycans of the intervillous spaces that mediate the adherence of Plasmodium falciparum-infected erythrocytes to the placenta, J. Biol. Chem. 275 (2000) 40344 – 40356. [18] A. Alkhalil, R.N. Achur, M. Valiyaveettil, C.F. Ockenhouse, D.C. Gowda, Structural requirements for the adherence of Plasmodium falciparum-infected erythrocytes to chondroitin sulfate proteoglycans of human placenta, J. Biol. Chem. 275 (2000) 40357 – 40364. [19] J.G. Beeson, G.V. Brown, M.E. Molyneux, C. Mhango, F. Dzinjalamala, S.J. Rogerson, Plasmodium falciparum isolates from infected pregnant women and children are associated with distinct adhesive and antigenic properties, J. Infect. Dis. 180 (1999) 64 – 72. [20] A. Muthusamy, R.N. Achur, M. Valiyaveettil, S.V. Madhunapantula, I. Kakizaki, V.P. Bhavanandan, D.C. Gowda, Structural characterization of the bovine tracheal chondroitin sulfate chains and binding of

[21]

[22] [23] [24]

[25]

[26]

[27]

[28]

[29]

[30]

[31] [32]

119

Plasmodium falciparum-infected erythrocytes, Glycobiology 14 (2004) 635 – 645. M. Fried, R.M. Lauder, P.E. Duffy, Plasmodium falciparum: adhesion of placental isolates modulated by the sulfation characteristics of the glycosaminoglycan receptor, Exp. Parasitol. 95 (2000) 75 – 78. Z. Dische, A new specific color reaction of hexuronic acids, J. Biol. Chem. 167 (1947) 189 – 198. U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227 (1970) 680 – 685. R.C. Krueger Jr., N.B. Schwartz, An improved method of sequential Alcian blue and ammoniacal silver staining of chondroitin sulfate proteoglycan in polyacrylamide gels, Anal. Biochem. 167 (1987) 295 – 300. M.R. Hardy, Monosaccharide analysis of glycoconjugates by highperformance anion-exchange chromatography with pulsed amperometric detection, Methods Enzymol. 179 (1989) 76 – 82. Y. Oike, K. Kimata, T. Shinomura, K. Nakazawa, S. Suzuki, Structural analysis of chick-embryo cartilage proteoglycan by selective degradation with chondroitin lyases (chondroitinases) and endo-beta-dgalactosidase (keratanase), Biochem. J. 191 (1980) 193 – 207. H. Saito, T. Yamagata, S. Suzuki, Enzymatic methods for the determination of small quantities of isomeric chondroitin sulfates, J. Biol. Chem. 243 (1968) 1536 – 1542. N. Ototani, Z. Yosizawa, Purification of chondroitinase B and chondroitinase C using glycosaminoglycan-bound AH-Sepharose 4B, Carbohydr. Res. 70 (1979) 295 – 306. A. Oldberg, P. Antonsson, K. Lindblom, D. Heinegard, A collagenbinding 59-kd protein (fibromodulin) is structurally related to the small interstitial proteoglycans PG-S1 and PG-S2 (decorin), EMBO J. 8 (1989) 2601 – 2604. A.A. Day, C.I. McQuillan, J.D. Termine, M.R. Young, Molecular cloning and sequence analysis of the cDNA for small proteoglycan II of bovine bone, Biochem. J. 248 (1987) 801 – 805. M. Fried, F. Nosten, A. Brockman, B.J. Brabin, P.E. Duffy, Maternal antibodies block malaria, Nature 395 (1998) 851 – 852. R.N. Achur, M. Valiyaveettil, D.C. Gowda, The low sulfated chondroitin sulfate proteoglycans of human placenta have sulfate group-clustered domains that can efficiently bind Plasmodium falciparum-infected erythrocytes, J. Biol. Chem. 278 (2003) 11705 – 11713.