Immunochemical and Biochemical Evidence for Distinct Basement Membrane Heparan Sulfate Proteoglycans”

Two antigenically and structurally related heparan sulfate proteoglycans (HSPG), with masses of 200 and 350 kDa, have been isolated and characterized from bovine renal tubular basement membranes (BTBM) using DEAE-Sephacel, octyl-Sepharose CL-4B, and Propac PA- 1 chromatography. Heparitinase treatment revealed core proteins of 145 and 125 kDa, with cor-responding core proteins after trifluoromethanesulfonic acid treatment of 88 and 82 kDa, from the 200-and 350-kDa HSPGs, respectively. The separated HSPGs produced similar tryptic peptide maps, had similar amino acid compositions, and had similarly sized GAG chains. The 200-kDa HSPG had 2.1 mg of protein/pmol of hexuronic acid compared with 1.1 mg/ pmol for the 350-kDa HSPG. and a

Immunochemical and Biochemical Evidence for Distinct Basement Membrane Heparan Sulfate Proteoglycans" (Received for publication, March 16, 1992) Steven G. Hagen$ §, Alfred F. MichaelSV, and Ralph J. ButkowskiV Two antigenically and structurally related heparan sulfate proteoglycans (HSPG), with masses of 200 and 350 kDa, have been isolated and characterized from bovine renal tubular basement membranes (BTBM) using DEAE-Sephacel, octyl-Sepharose CL-4B, and Propac PA-1 chromatography. Heparitinase treatment revealed core proteins of 145 and 125 kDa, with corresponding core proteins after trifluoromethanesulfonic acid treatment of 88 and 82 kDa, from the 200and 350-kDa HSPGs, respectively. The separated HSPGs produced similar tryptic peptide maps, had similar amino acid compositions, and had similarly sized GAG chains. The 200-kDa HSPG had 2.1 mg of protein/pmol of hexuronic acid compared with 1.1 mg/ pmol for the 350-kDa HSPG.
Anti-BTBM HSPG monoclonal antibody (mAb A12) reacted with core proteins derived from the 200-and 350-kDa HSPGs, whereas anti-perlecan polyclonal and monoclonal antibodies did not bind to the BTBM HSPG core proteins described above but reacted with a 230-kDa core protein, which was nonreactive with mAb A12. Immunohistochemical studies of the kidney demonstrated differences in the distribution of BTBM HSPG and perlecan. Comparison of amino acid sequences from BTBM HSPG tryptic peptides with the sequence of perlecan revealed similarities but not extensive identity. Two tryptic peptides show homology to rat agrin, a basement membrane component of synaptic junctions.
These data suggest that the two BTBM HSPGs are immunologically and structurally related and that differences in these molecules may arise from alternative splicing or posttranslational modifications. In addition, the two BTBM HSPGs are immunologically and structurally distinct from perlecan but may share homology with agrin.
* This research was supported by Grants DK36007 and 10704-32 from the National Institute of Health; the Juvenile Diabetes Foundation, International; the American Diabetes Association, Minnesota Affiliate; the Viking Children's Fund National Kidney Foundation of the Upper Midwest; and the University of Minnesota Graduate School. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. § This work is in partial fulfillment for the Doctor of Philosophy Degree.
11 To whom correspondence should be addressed. Basement membranes (BMs)' are specialized extracellular matrices which are important in cell adhesion, migration, and differentiation; in the structural support of cells and tissues; and as a barrier to cells and macromolecules (1)(2)(3)(4). Compositional and structural differences of various BMs support the existence of heterogeneity and functional specialization (5-9).
BM HSPGs, which are immunochemically and structurally distinct from cell surface HSPGs (10-131, have been isolated from tumor matrices (14)(15)(16)(17), cell cultures (12,(18)(19)(20)(21)(22), and native BMs (23-28). These studies indicate that BM HSPGs are heterogeneous molecules differing in size and biochemical and immunochemical characteristics. Most of our understanding of BM HSPGs has been derived from studies of the murine EHS tumor from which two related HSPGs have been isolated (13,29,30). These HSPGs are thought to arise from the same gene and processed in a way that the small (core proteins of 95-130 kDa) HSPG is proteolytically derived from the large (core protein of 400 kDa) HSPG (30). Immunochemical and structural evidence suggest that the large EHS HSPG (perlecan) is found in many if not all BMs (17,25,31,32). Proteolytic processing is also thought to occur in other tissues, including the renal glomerulus where a large (400 kDa) core protein is seen in intact glomeruli, whereas smaller core proteins are observed in isolated GBM (33). However, there is immunological and structural evidence for the existence of distinct BM HSPGs which are unrelated to perlecan (34).
Two antigenically related HSPGs have been isolated from calf lens capsule (25),human,bovine,and equine GBM (26,35,36), and human TBM (27). The two HSPGs isolated from the latter (with core proteins of 160 and 110 kDa) are structurally and immunochemically related to the two HSPGs isolated from human GBM (27).
Although the two HSPGs have not been separated, it has been suggested that one is derived from the other by proteolytic cleavage and that these are related to perlecan (25).
In the present study methods are described to separate the 200-and 350-kDa HSPGs isolated from bovine TBM. Biochemical and immunochemical analyses indicate that these two HSPGs are structurally and immunochemically related to each other but are distinct from perlecan.
Extraction and Purification of HSPG from BTBM-Proteoglycans were extracted from 5 g of BTBM overnight at room temperature using 4 M guanidine HCl, 50 mM Tris/HCl, pH 7.4, 1 mM Nethylmaleimide, and 0.1 mM phenylmethylsulfonyl fluoride. The extract was centrifuged at 2500 X g for 10 min and the supernatant dialyzed into 6 M urea, 50 mM sodium phosphate, pH 6.3. The extract was chromatographed on a DEAE-Sephacel (1 X 50 cm) anion exchange column equilibrated in 6 M urea, 50 mM sodium phosphate, pH 6.3. The column was washed with 0.1 M NaCl in the equilibration buffer, and the unbound fractions were collected for further study. Proteoglycans were then eluted with a 0.1-0.8 M NaCl linear gradient in the equilibration buffer using a flow rate of 30 ml/h collected in 5ml fractions. Fractions were monitored by absorbance at 280 nm and by the dimethylmethylene blue assay (DMMB). The PG-containing fractions were pooled and dialyzed into 25 mM Tris/HCl, pH 7.4. After dialysis the DEAE PG pool was applied to an octyl-Sepharose CL-4B column (1 X 25 cm) equilibrated in 25 mM Tris/HCl, pH 7.4. The column was washed with the equilibration buffer to remove unbound material, and bound PGs were eluted with 4 M GuHCl and 1% Triton X-100, 4 M GuHCl both in 25 mM Tris/HCl, pH 7.4. The 4 M GuHCl pool and 1% Triton X-100 pool were dialyzed into 6 M urea and fractionated over a Dionex Propac PA-1 (4 X 250 mm) anion exchange column using a nonlinear NaCl gradient (0-0.5 M NaCl from 2 to 12 min, 0.5-1.3 M NaCl from 12 to 52 min, and 1.3-2 M NaCl from 52 to 62 min) in 0.5 M urea, 20 mM Tris/HCl, pH 7.0, at a flow rate of 1.0 ml/min. The PG-containing fractions (detected by absorbance at 280 nm, DMMB, and SDS-PAGE) were pooled for further characterization.
Deglycosylation of Proteoglycans-Heparitinase (ICN) (50 milliunits/mg) and chondroitinase ABC (Sigma) (10 milliunits/pg) digestions were carried out for 16 h at 37 "C in 100 mM Tris, pH 7.2, 10 mM CaC1, 1  The guanidine HC1 extract, originating from 5 g of BTBM, was dialyzed against 6 M urea, 50 mM sodium phosphate, pH 6.3, and chromatographed over 50 ml of DEAE-Sephacel equilibrated in the dialysis buffer. The column was washed with 6 M urea, 50 mM sodium phosphate, 0.1 M NaCl, pH 6.3, and the unbound material was pooled. Bound proteoglycans were eluted with a linear 0.1-0.8 M NaCl gradient (----), in 6 M urea buffer, and 5-ml fractions were collected at a flow rate of 30 ml/h. Elution of proteoglycans was monitored by absorbance at 280 nm (0) and using DMMB (0). Proteoglycan containing fractions (indicated by the stippled bar) were pooled for further study. sulfonic acid (TFMS) digestion was carried out for 6 h at 4 "C, and the products were isolated by ether/pyridine extraction (38). Deglycosylated products were visualized by SDS-PAGE and by Western blots.
Peptide Mapping of BTBM HSPG-After purification, 100 mg of the 200-and 350-kDa HSPGs were digested overnight at 37 "C with trypsin (Sigma) (1% w/w) in 0.1 M Tris, 0.15 M NaC1, pH 8.0. The digestion was stopped by addition of an equal volume of 4 M guanidine HCl, 50 mM Tris/HCl, pH 7.4, and 2-mercaptoethanol (8%) was added and incubated at 65 "C for 30 min. Peptide maps were produced by fractionating the tryptic digests on a Vydac C18 reverse phase HPLC column (0.26 cm X 15 cm) using a linear gradient of 0-40% acetonitrile (0.1% trifluoroacetic acid) over 120 min with a flow rate of 0.25 ml/min. The elution of peptides was monitored by absorbance at 215 nm.
Antisera-Monoclonal antibody A12 was produced by standard procedures described previously (39). Spleen cells from Balb/c mice immunized with BTBM HSPGs isolated by DEAE-Sephacel chromatography (see above) were fused with SP2/OAg14 mouse myeloma cells. Initial screening was carried out by indirect IF and by ELISA, using partially purified BTBM HSPG. Positive clones were subcloned by limiting dilution and screened by ELISA, using purified BTBM HSPG, by indirect IF, and by Western blotting. Rabbit anti-perlecan was kindly provided by John R. Hassell (Pittsburgh Eye and Ear Institute). Anti-perlecan rat monoclonal antibody A7L6 (32, 40) was kindly provided by Alexander V. Ljubimov (La Jolla Cancer Research Center).
Western blot analysis was carried out after separation of proteins by SDS-PAGE and transfer to Immobilon-P (Millipore) with a semidry transfer apparatus (American Bionetics, Inc.) using a continuous transfer buffer: 39 mM glycine, 48 mM Tris, 0.0375% SDS, 10% methanol. Rabbit anti-perlecan was used at a 1:200 dilution and mAb A12 and mAb A7L6 were used without dilution. Alkaline phosphatase-conjugated secondary antibodies were detected using nitro blue  Immunofluorescent Microscopy-Indirect IF was performed using fluorescein and rhodamine-labeled secondary antibodies on frozen kidney sections as described previously (43,44). Dual label indirect IF was carried out with mAb A12 and mAb A7L6 in the following sequence: mAb A12, fluorescein-conjugated goat anti-mouse IgG, mAb A7L6, rhodamine-conjugated goat anti-rat IgG. Appropriate controls for single label and dual label studies were utilized, and secondary antibodies that showed nonspecific reactivity were not used (44). mAbs A12 and A7L6 were used without dilution, and rabbit anti-perlecan was used at a 1:20 dilution.
Tryptic Peptide Purification and Sequence Analysis-After purification, the 200-kDa HSPG (500 pg) was digested with trypsin (l%, w/w) in 0.1 M Tris/HCl, 0.15 M NaCl, pH 8.0, overnight at 37 'C. The reaction was stopped by addition of an equal volume of 4 M guanidine HCl, 50 mM Tris/HCl, pH 7.4, and 2-mercaptoethanol to a final concentration of 4% and heated at 65 'C for 30 min. The digests were first fractionated using Spherogel TSK (23000 SW and TSK 2000 SW HPLC gel filtration columns run in tandem and eluted with 4 M GuHCl, 50 mM Tris/HCl, pH 7.4, at flow rate of 0.5 ml/min and monitored by absorbance at 280 nm. Gel filtration pools containing peptides were then chromatographed over a Vydac C18 reverse phase HPLC column (0.26 X 15 cm) using a 15-45% acetonitrile (0.1% trifluoroacetic acid) gradient over 120 min at a flow rate of 0.25 ml/min. Elution of peptides was monitored at 215 nm, and peptides were collected from the reverse phase columns and sequenced using an Applied Biosystems model 470A gas phase sequenator equipped with an in-line Applied Biosystems model 120A HPLC system for detection of phenylthiohydantoin derivatives. The EHS HSPG perlecan sequence was kindly provided by John R. Hassell (45).
Chemical Analysis-Amino acid analysis was carried out using a Beckman 6300 amino acid analyzer after hydrolysis of nonreduced samples at 110 "C in constant boiling HC1. Duplicate samples were hydrolyzed for 19,48, and 62 h, and serine and threonine values were obtained by extrapolation. Protein concentration was determined using the Pierce Chemical Co. BCA assay (46), hexuronic acids using the carbazole reaction (47), and sulfated GAGS using the DMMB assay (48).
HSPG and GAG Chain Analysis-GAG chains were removed by pelimination using 0.1 M NaOH, for 24 h at 25 "C (28). The reaction was neutralized, and for HPLC, an equal volume of 4 M guanidine HCl, 50 mM Tris/HCl, pH 7.4, was added, and for ELISA assays the products were dialyzed and lyophilized. Nitrous acid digestion was carried out as described by Shively and Conrad (49). After 8-elimination reaction and nitrous acid digestion, the products were chromatographed over TSK 3000-TSK 2000 HPLC gel filtration columns in tandem and eluted with 2 M GuHCl, 50 mM Tris/HCl, pH 7.4, with a flow rate of 0.5 ml/min. The fractions were assayed by the DMMB assay. Thyroglobulin (669 kDa), apoferritin (443 kDa), alcohol dehydrogenase (150 kDa), bovine serum albumin (66 kDa), and cytochrome c (12.4 kDa) were used as standards.  (Fig. 4A). Lane 1 in Fig. 4A shows a mixture of the 200-and 350-kDa HSPGs after purification using only DEAE-Sephacel and Propac PA-1 columns without separation using octyl-Sepharose CL-4B, indicating that the octyl-Sepharose CL-4B column resolved the 200-and 350 kDa HSPGs (Fig. 4A, lanes 2 and 3, respectively). The silverstained gels also show that the separated HSPGs, after the Propac PA-1 column, are relatively free of contaminating protein and the 95-kDa CSPG. Western blot analysis using mAb A12 after heparitinase and TFMS treatment of the 200and 350-kDa HSPGs revealed core proteins of 145 and 88 kDa for the 200-kDa HSPG, respectively (Fig. 4B, lanes 1 and  3 ) and core proteins of 125 and 82 kDa for the 350-kDa HSPG, respectively (Fig. 4B, lanes 2 and 4 ) , indicating that the 200-kDa HSPG has a larger core protein than the larger 350-kDa HSPG.
Comparison of Anit-BTBM HSPG and Anti-perkcan Antibodies by Indirect Immunofluorescence-Comparison of the immunoreactivity of anti-BTBM HSPG (mAb A12) and the anti-perlecan polyclonal and monoclonal (mAb A7L6) antibodies by indirect immunofluorescent microscopy on bovine kidney sections revealed different staining patterns (Fig. 7). Reactivity of anti-BTBM HSPG mAb A12 and polyclonal and monoclonal anti-perlecan antibodies was carried out by Western blotting using various pools of BTBM HSPGs. Panel A compares the reactivity of anti-BM HSPG antibodies with the DEAE unbound pool (see Fig. 1). Lane 1 shows the lack of reactivity of mAb A12, whereas rabbit anti-perlecan (lane 2) and monoclonal anti-perlecan (mAb A7L6) (lane 3 ) react with many bands in this fraction. Panel B compares the reactivity of the anti-HSPG antibodies with the 88-kDa core protein, generated by TFMS digestion, of purified 200-kDa BTBM HSPG (pool A-2, Fig. 3). Lane I shows reactivity of the 88-kDa core protein with mAb A12, whereas no reactivity is observed with rabbit anti-perlecan (lane 2) or mAb A7L6 (lane 3 ) . In panel C, reactivity of the 145-and 125-kDa heparitinase-derived core proteins from the 200-and 350-kDa HSPGs, respectively, with the anti-HSPG antibodies is compared. Reactivity of mAb A12 with the 145-and 125-kDa HSPG core proteins is shown in lanes I and 3, respectively. Rabbit anti-perlecan (lanes 2 and 4 ) and mAb A7L6 (lane 5) do not react with either the 145-kDa ( l a n e 2) or 125-kDa (lanes 4 and 5) core proteins but do react with a heparitinasegenerated 230-kDa band in lanes 4 and 5, respectively. Two mg of protein were run in lanes I and 3, and 10 mg of protein were run in lanes 2 and 4. Lane 5. stained with mAb A7L6. was the second layer of Immobilon-P membrane in a double layer transfer in which 10 mg of protein were run on the gel. capsule, the anti-perlecan antibody also stained the extracellular matrix in the interstitium and mesangium, whereas mAb A12 did not stain in these areas. Rabbit anti-perlecan, shown in panel C, shows the identical staining pattern as mAb A7L6.
Within the glomerulus, differences in GBM staining were also seen where mAb A12 stained the lamina densa of the GBM, whereas the anti-perlecan antibodies stained the subendothelial aspect of the GBM, as seen by phase fluorescence micros-COPY.
Sequence Analysis of BTBM HSPG Tryptic Peptides-After purification by HPLC gel filtration and reverse phase HPLC (Fig. 8), three tryptic peptides obtained from the 200-kDa HSPG were each sequenced twice and the resulting sequences compared with the sequence of perlecan, which was determined from a cDNA library (Fig. 9) (45,50,51). Similarities were observed between the three BTBM HSPG peptides and three sequences found in the amino terminal portion of mouse and human perlecan; however, the low identity between the sequences may indicate a lack of homology. Mouse and human perlecan have a high degree of identity, indicating conservation of sequence between species. The perlecan sequence aligned to peptide 2 is located in the spacer region between the first and second low density lipoprotein receptor-like repeats in domain I1 of the perlecan core protein. The perlecan sequences aligned to peptides 1 and 3 are both found in the laminin A chain-like domain 111, which is similar to the short arm of the laminin A chain, where the peptide 1-aligned sequence is found in the second cysteine repeat and the peptide 2-aligned sequence is found in the globular region of subdomain 1. It should also be noted that peptides 1 and 3 do not show homology with the other cysteine or globular repeats in domain I11 of perlecan or the laminin A chain sequences homologous to the perlecan repeats. Sequence similarities, as determined by searching protein databases at the National Center for Biotechnology Information using the Blast network service, were also observed between the BTBM HSPG tryptic peptides and rat agrin, a BM component associated with acetylcholine receptors in the neuromuscular junction (52). Peptides 2 and 3 revealed 59 and 72% identity, respectively, with rat agrin, whereas peptide 1 did not show any sequence similarity with agrin ( Fig. 9).

DISCUSSION
Recent studies suggest that BM HSPGs comprise a heterogeneous population of molecules formed by proteolytic cleavage of the core proteins which removes peptides and attached GAG chains (30,33,53). The present study investigates further the structural and immunochemical characteristics of HSPGs isolated from BTBM and between the BTBM HSPGs and the EHS HSPG perlecan. Methods were adopted to isolate and characterize 200-and 350-kDa HSPGs from BTBM using DEAE-Sephacel, octyl-Sepharose CL-4B, and Propac PA-1 chromatography. Biochemical and immunochemical studies indicate that the 200-and 350-kDa HSPGs are structurally and immunochemically related. Both the large (350 kDa) and small (200 kDa) HSPGs and their core proteins react with a monoclonal antibody (mAb A12) produced against BTBM HSPG.
Characterization of HSPGs from GBM (26, 35, 36), TBM (27), and lens capsule (25) have revealed two immunochemically related HSPGs, but structural studies could not be completed until separation of the two HSPGs was achieved. Structural analysis of the 200-and 350-kDa HSPGs separated in the present study revealed similar peptide maps and amino acid compositions. The amino FIG. 7. Localization of basement membrane HSPGs by indirect immunofluorescent microscopy on bovine kidney. The same bovine kidney section was reacted sequentially with anti-BTBM HSPG (mAb A12) and goat anti-mouse IgG fluorescein isothiocyanate (panel A ) followed by anti-perlecan (mAb A7L6) and goat anti-rat IgG rhodamine (panel B ) . In panel A , within the glomerulus, mAb A12 stains the phase-dense image of the GBM ( G ) intensely but fails to bind to the mesangial matrix (arrow). No reactivity is seen in the interstitium ( I N ) . In panel B, anti-perlecan mAb A7L6, in contrast, reacts strongly with the mesangial matrix (arrow) and weakly with the glomerular capillary wall (C) in the subendothelial region as viewed by phase-fluorescence microscopy. In addition, weak interstitial reactivity is present. Both antibodies bind to similar regions of the TBM (T) and Bowman's capsule (BC). In panel C, another section of bovine kidney was stained with rabbit anti-perlecan followed by goat anti-rabbit IgG fluorescein isothiocyanate. The distribution of staining is similar to that observed with mAb A7L6 in panel B. Magnification in panels A and B is X 670 and in panel C is X 1100. acid compositions are also similar to that reported for bovine GBM HSPG, indicating structural similarities between the HSPGs isolated from bovine TBM and GBM (28). Deglycosylation studies, using both heparitinase and TFMS, indicate that the 350-kDa HSPG has smaller core proteins (125 and 82 kDa) than the 200-kDa HSPG (145 and 88 kDa). Furthermore, since HS GAG chains from the 200-and 350-kDa HSPGs are of similar size and their protein/hexuronic acid ratios are 2.1 and 1.1 (mg/pmol), the larger HSPG most likely has more GAG chains attached to its core protein. GAG chain analysis of GBM and lens capsule HSPGs revealed four to eleven 14-kDa HS chains attached to the core proteins (25,28).
HSPGs isolated from native BMs has shown heterogeneity in the size of both the intact HSPG and also the core proteins when compared between species and between different tissues (31, 53). This size heterogeneity may indicate that these HSPGs are highly modified molecules in which the tissue and species determines the size of the intact HSPGs and their core proteins and also the number of attached GAG chains. It has been proposed that in the GBM (25), glomeruli (33), can and rat agrin. BTBM HSPG amino acid sequences, obtained from tryptic peptides shown in Fig. 8, are compared with sequences of mouse and human perlecan derived from cDNA (45, 51) and rat agrin. Identities with perlecan are shown in boxes and with agrin are in boldface type. Unknown residues are denoted by an X . Each BTBM HSPG peptide was sequenced twice.

G P C G S R D P C A N V T C S F G S T C V P G~L C G P G A V C P P S V E D P G
and EHS tumor (30), a large HSPG core protein is synthesized and then proteolytically processed to produce smaller HSPGs. However, since the 350-kDa HSPG that we isolated has a smaller core protein than the 200-kDa HSPG, it does not appear that the 200-kDa HSPG is a proteolytically processed form of the larger 350-kDa HSPG. Rather, different core protein sizes may be a consequence of alternative splicing of the HSPG gene or may arise as a posttranslational event in which a portion of the core protein of the larger HSPG is removed. It is also possible that the two HSPGs are distinct gene products. Degradation of HSPG and core protein during isolation or deglycosylation may offer another possible explanation for tissue and species size differences.
Differences in core protein size, antibody reactivity, and amino acid sequences indicate that the BTBM HSPGs are distinct from perlecan. The anti-BTBM HSPG mAb A12 reacted with the 145-and 88-kDa core proteins from the 200-kDa HSPG and the 125-and 82-kDa core proteins from the 350-kDa HSPG, whereas anti-perlecan polyclonal and monoclonal antibodies reacted only with a 230-kDa core protein by Western blotting. HSPGs have been isolated from cell culture (18-20) and various other tissues, including glomeruli (24,31,33), in which core proteins of 230 kDa and larger have been shown to be reactive with anti-perlecan antibodies. This suggests that these large core proteins are related to perlecan. However, no small core proteins reactive with anti-perlecan antibodies have been identified. In this study anti-perlecan antibodies react with many bands in the DEAE unbound pool, whereas mAb A12 was shown to be unreactive with these proteins. This result may suggest a difference in the turnover rate of these HSPGs illustrated by the presence of perlecan degradation products and the absence of BTBM HSPG degradation products. This is supported by studies on the turnover of BM HSPGs which revealed two turnover rates, suggesting distinct HSPG populations (54).
Distributions of BTBM HSPG and perlecan are distinct, as shown by immunofluorescence microscopy. Antibodies to BTBM HSPGs and perlecan stain TBM and GBM. However, mAb A12 does not stain mesangial or interstitial matrices, whereas anti-perlecan antibodies stain these structures. Furthermore, there is a difference in the location of antibody binding within the GBM, where mAb A12 binds to the lamina densa and anti-perlecan antibodies bind to the subendothelial aspect of the GBM. These results suggest that mAb A12-reactive HSPGs have a more limited tissue distribution than anti-perlecan-reactive HSPGs, which have been demonstrated in all BMs (17, 31, 32). Recent evidence has also suggested differences in immunoreactivity of antibodies specific for GBM, liver sinusoid, and PYS-2 HSPGs compared with anti-perlecan antibodies (34,55,56).
Similarities exist between the amino acid sequences derived from BTBM HSPG tryptic peptides and perlecan. However, the lack of extensive identity between the BTBM HSPG sequences and both mouse and human perlecan, and the conservation of sequence of perlecan between species, indicate that the peptide sequences are derived from an HSPG distinct from perlecan (21,51,57). The BTBM HSPG peptides, which are aligned with sequences derived from the amino-terminal portion of perlecan, are found in conserved areas of perlecan as shown by the high degree of identity between the mouse and human sequences in Fig. 9. The similarities between the BTBM HSPG peptide sequences and agrin could indicate homology between these BM components. Agrin, a 200-kDa BM component thought to be associated with acetylcholine receptors, contains domains homologous to protease inhibitors, laminin, and epidermal growth factor (52). Agrin has been localized to the BM of the synapses of the neuromuscular junction; however, its distribution in kidney BMs is unknown (58).
Our results suggest that there are two groups of immunochemically and structurally distinct BM HSPGs. The first is the anti-perlecan reactive HSPGs that, in mature BMs, appear to have core proteins of high molecular weight and may reflect tissue-and species-specific processing or degradation of a precursor core protein of 400 kDa (31).
The second BM HSPG consists of the anti-BTBM HSPGreactive 200-and 350-kDa HSPGs. These HSPGs are structurally and immunologically related to each other but appear to be distinct HSPGs. It is unlikely that they are derived from each other by proteolytic processing. These HSPGs have a more restricted tissue localization than perlecan and may be functionally distinct as well.
It has been suggested that through core protein interactions with other BM components (59) HSPGs are anchored into the BM forming a charged barrier important in filtration of macromolecules in the GBM (4) and also play a role in cell adhesion (60). Furthermore, it is thought that BM HSPG may be important in the pathogenesis of diseases such as diabetes and polycystic kidney disease (61,62). In consensus with other BM components, HSPGs are structurally and immunochemically heterogeneous, providing a bases for functional specialization among different BMs.