ARCHIVES

OF BIOCHEMISTRY

Vol. 292, No. 1, January,

AND

BIOPHYSICS

pp. 54-61,

1992

Identification of Chick Cornea1 Keratan Sulfate Proteoglycan Precursor Protein in Whole Corneas and in Cultured Cornea1 Fibroblasts Pamela

K. Schrecengost,

Thomas

C. Blochberger,

and John

R. Hasselll

Department of Ophthalmology, The Eye and Ear Institute of Pittsburgh, University School of Medicine, 203 Lothrop Street, Pittsburgh, Pennsylvania 15213 Received

June

27, 1991, and in revised

form

August

29, 1991

The precursor protein to the chick cornea1 keratan sulfate proteoglycan was identified by immunoprecipitation with antiserum to its core protein from lysates of [35S]methionine-pulsed corneas and cornea1 fibroblasts in cell culture. Antiserum to the keratan sulfate proteoglycan immunoprecipitated a doublet of M, 52,000 and 50,000 and minor amounts of a M, 40,000 protein from pulsed corneas. Pulse-chase experiments, which permitted the conversion of the precursor proteins to proteoglycans and digestion of the glycosaminoglycans on immunoprecipitated proteoglycans with keratanase or chondroitinase ABC, showed that the M,. 52,000-50,000 doublet was converted to a keratan sulfate proteoglycan and the M,. 40,000 protein was converted to a chondroitin sulfate proteoglycan. Chick cornea1 fibroblasts in cell culture primarily produced the smaller (M, 50,000) precursor protein, and in the presence of tunicamycin the precursor protein size was reduced to M, 35,000, which indicates that the core protein contains approximately five N-linked oligosaccharides. Pulse-chase experiments with cornea1 fibroblasts in culture showed that the precursor protein was processed and secreted into the medium. However, its sensitivity to endo-&galactosidase and resistance to keratanase indicate that the precursor protein was converted to a glycoprotein with large oligosaccharides and not to a proteoglycan. This suggests that, although the precursor protein for the proteoglycan is produced in cultured cornea1 fibroblasts, the sulfation o 1992Academie enzymes for keratan sulfate may be absent. Press,

Inc.

Proteoglycans are present in extracellular matrices and on the surface of cells. They can be considered a special class of glycoproteins that contain one or more glycos1 To whom

correspondence

should

be addressed.

of Pittsburgh

aminoglycan side chains (1). The cornea1 stroma is a fibrillar extracellular matrix containing proteoglycans bearing chondroitin/dermatan sulfate side chains as well as proteoglycans bearing keratan sulfate side chains. These proteoglycans have been isolated and characterized from rabbit (2, 3), bovine (4-6), monkey (7,8), and chick (9-12) corneas. Variations in proteoglycan size and charge suggested that there are at least two proteoglycans with chondroitin/dermatan sulfate side chains and two with keratan sulfate side chains. Removal of the chondroitinl dermatan sulfate side chains with chondroitinase ABC, however, revealed a M, 43,000 core protein for both bovine (13) and chick (12) cornea1 proteoglycans. This chondroitin/dermatan sulfate proteoglycan was recently shown to be decorin (14), a proteoglycan found in many fibrillar extracellular matrices. Messenger RNA for biglycan, a similar size chondroitin/dermatan sulfate proteoglycan but a different gene product, was not detected in bovine cornea (14). This would suggest that the chondroitin/dermatan sulfate proteoglycans in the stroma are derived from one gene product. In contrast, keratanase digestion of bovine keratan sulfate proteoglycan revealed at least two core proteins of M, 40,000 and 55,000 (6). Analysis of these cores by immunologic cross-reactivity and peptide mapping indicated that the proteins had similarities but that they also had differences (15). Thus, it is not clear if the smaller core is a breakdown product of the larger, if they were both derived as breakdown products from a larger core, or if they were derived from separate gene products. Antisera raised against the proteoglycans’ core protein have been successfully used to help resolve questions of apparent proteoglycan heterogeneity by identifying the precursor protein to perlecan, a basement membrane proteoglycan containing heparan sulfate side chains (16,17). This involved incubating cultured cells that produce the proteoglycan with a radiolabeled protein precursor for a short period of time, so that all the incorporated radiolabel

54 All

Copyright 0 1992 rights of reproduction

0003.9861/92 $3.00 by Academic Press, Inc. in any form reserved.

CHICK

CORNEAL

KERATAN

SULFATE

is still in the rough endoplasmic reticulum. The radiolabeled material immunoprecipitated under these conditions will have N-linked oligosaccharides but because it has not yet been transported to the Goli or secreted into the matrix it will not have other post-translational modifications such as glycosaminoglycan addition or proteolytic clipping that produce proteoglycan heterogeneity. Using this approach with a.ntibodies to the chicken cornea1 keratan sulfate proteoglycan and radiolabeled chick corneas or cornea1 fibroblasts., we find that only one precursor protein is converted to a lkeratan sulfate proteoglycan. MATERIALS

AND

METHCODS

Corneas Isolation ofproteoglycan coreprotein and antiseraproduction. were removed from chicken eyes received from a local source, immediately frozen in liquid nitrogen, and stored at -80°C. The proteoglycans were extracted and purified using modifications of a previous procedure (18). In brief, 10 g of frozen corneas was extracted with 10 vol of 4 M guanidine-HCl at 4’C for 16 b. The supernatant was saved and the extraction repeated. The supernatants were combined, dialyzed against 6 M urea containing 0.05 M Tris, 0.15 M NaCl, 0.1% Chaps’ (Tris-buffered urea), and applied to a column (2!.5 X 5.0 cm) of DEAE-Sepharose equilibrated with the same solvent. The pass-through fraction was discarded and the proteoglycans were eluted with a 0.15-1.15 M NaCl gradient. The elution position of the proteoglycan was detected by absorbance at 280 nm for core protein and with dimethylmethylene blue for glycosaminoglycan. The tubes containing the proteoglycan fraction were pooled, dialyzed against distilled water, and lyophilized. The proteoglycan fraction was reconstituted in water, digested with 5 units of chondroitinase ABC (Seikagaku AmericaI, Inc.) for 3 h at 37”C, dialyzed against Tris-buffered urea, and reapplied to a column (1 X 3 cm) of DEAESepharose. The pass-through material was collected as “CSPG core preparation” and the bound material eluted with Tris-buffered urea containing 1.15 M NaCl as the “proteoglycan fraction.” Both the CSPG core preparation and the proteaoglycan fraction were dialyzed against deionized water and lyophilized. The proteoglycan fraction was reconstituted in water, digested with 10 units of keratanase (Seikagaku America, Inc.) for 3 h at 37’C, dialyzed against Tris-buffered urea, and applied again to a column (1 X 3 cm) of DEAE-Sepharose. The passthrough fraction was collected as “KSPG core preparation,” dialyzed against distilled water, and lyophilized. Based on the results of a previous study characterizing core protein size for the chondroitin/dermatan sulfate and the keratan sulfate proteoglycan from chick corneas (In), the M, 42,000 core protein for the chondroitin/dermatan sulfate proteoglycan in the CSPG core preparation and the M, 51,000 core protein in the KSPG core preparation were easily identifiable by SDS-PAGE. These core proteins were purified to homogeneity by FPLC and used to raise antibodies in rabbits using MPL,+TDM+CWS emulsion as an adjuvant (RIB1 Immunochem Research, Inc.). The details of the proteoglycan and core protein purification will be published elsewhere. The specificity of the antisera raised against the core proteins is demonstrated in this report. Isolation and dissection of corneas. The corneas were excised from Day 15 chick embryos and dissected into epithelium and stroma by modification of previously published procedures (19). Corneas were removed from Day 15 chick embryos by grasping the cornea at the limbus with small mouse-toothed forceps and sharply pulling the cornea from

’ Abbreviations used: Chaps, 3-[(3-cholamidopropyl)dimethylammoniolpropanesulfonic acid; SDS., sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; FPLC, fast protein liquid chromatography; HBSS, Hanks’ balanced salt solution; DMEM, Dulbecco’s modified Eagle’s medium; IgG, immunoglobulin G; DTT, dithiothreitol.

PROTEOGLYCAN

PRECURSOR

PROTEIN

55

the eye. Corneas were then incubated at 37°C in 2.74 mM EDTA in calcium and magnesium-free Hanks’ balanced salt solution (HBSS) for 15-20 min. The epithelium was then removed as a sheet of cells with a probe, leaving the stroma free of epithelium but probably containing variable amounts of endothelium. The corneas, epithelium, and stroma were used directly for biosynthetic radiolabeling studies. The epithelium was floated, basal lamina side down, onto a Millipore filter to facilitate its handling during the radiolabeling procedures. Cornea1 fibroblasts were isolated after Corneal fibroblast isolation. collagenase digestion of stroma obtained from the dissection described above. The stroma were first predigested with stirring for 5 min with 0.15% collagenase (Worthington) in HBSS to remove any remaining epithelial and endothelial cells. The collagenase was removed and replaced with fresh collagenase, and the digestion continued with stirring until the corneas were completely digested (45-60 min). The digest was passed through a double layer of Nitex cloth (55-Mm mesh) to remove any remaining clumps and the cells were pelleted by low-speed centrifugation. The cells were washed once in Dulbecco’s modified Eagle’s medium (DMEM) containing 15% calf serum, 0.1% gentamicin (ElkinsSinn, Inc.), and 0.5% penicillin-streptomycin (GIBCO Laboratories), counted with a hemocytometer, and then plated out at lo6 tells/35-mm tissue culture dish. The cells were allowed to attach and spread for 2 h before the medium was replaced. Radiolabeling. The corneas, cornea1 stromas, cornea1 epithelium cell layer, and cornea1 epithelium cell layer, and cornea1 fibroblasts in culture were first incubated in a tissue culture incubator at 37’C for 30 min in serum-free and methionine-free DMEM to deplete the pool of free methionine (16). The medium was then replaced with serum-free and methionine-free DMEM containing [35S]methionine (500 &i/ml medium) and incubated for 30 min (cornea1 fibroblasts in culture) or 1 h (cornea, stroma, and epithelium). The medium was discarded and either the cells/ tissues were then harvested immediately (pulse) or the medium was replaced with serum-free DMEM containing methionine and incubated for an additional 5 h (chase). At the end of the chase period both the medium and the cell layer were harvested from cell cultures. In the cornea organ cultures, however, only the corneas were harvested at the end of the chase since most of the incorporated radiolabel is retained by the corneas and not secreted into the medium (11). The fibroblasts and the epithelial cell layer cultures were harvested directly in extraction buffer. The corneas and stromas were frozen in liquid nitrogen, pulverized while frozen, and then homogenized in extraction buffer. Aliquots of the extracts were precipitated with trichloroacetic acid in the presence of carrier bovine serum albumin. The precipitates were dissolved in 0.1 M NaOH and incorporated radioactivity was determined by liquid scintillation counting. Western blot. The specificity of the antisera for the cornea1 proteoglycans was demonstrated by Western blot. Proteoglycans extracted by 4 M guanidine from Day 18 chick embryo corneas were isolated in a single fraction from chromatography on DEAE in 6 M urea. The isolated proteoglycans were digested, in the presence of proteolytic inhibitors, with chondroitinase ABC and keratanase combined (20, 21). The digests were electrophoresed in 7.5% acrylamide gels containing SDS and transferred to nitrocellulose (17). Separate lanes were then reacted with preimmune serum, antiserum to keratan sulfate proteoglycan, or antiserum to chondroitin/dermatan sulfate proteoglycan at 1:250 dilutions and then horseradish peroxidase-conjugated goat anti-rabbit IgG followed by color development. Immunoprecipitation and electrophoresis. Precursor proteins and proteoglycans were immunoprecipitated and visualized according to a previously published methodology (16). In comparisons of the level of precursor protein synthesis by different tissues or cells, immunoprecipitations were done from equivalent amounts of incorporated radiolabel. In brief, the lysates and media from radiolabeled cells or corneas were mixed with protein A-Sepharose beads (Pharmacia) that had been previously incubated with immune or preimmune serum. The beads were washed and the bound materials were released from the beads by boiling in the presence of SDS. The samples were electrophoresed on 7.5%

56

SCHRECENGOST.

BLOCHBERGER,

polyacrylamide gels containing SDS. The a&amide gel was embedded with Fluoro-Hance (Research Products International Corp.), dried under vacuum, and exposed to X-Omat film at -8O’C for 1 day to several weeks. Once the location of the immunoprecipitated material was determined, the bands were cut out of the acrylamide gel and swelled in TS-1 (Research Products International Corp.), and the radioactivity was measured by liquid scintillation counting. Background was determined by cutting out an equivalent-size piece of acrylamide gel from a lane used for immunoprecipitation with preimmune serum and measuring the radioactivity. The completeness of immunoprecipitation was determined by doing three consecutive immunoprecipitations using the same lysate and totaling the radioactivity obtained in the band by liquid scintillation counting. In general, from 51 to 89% of the total radioactivity obtained was in the first immunoprecipitation and 9 to 0% by the third immunoprecipitation. In cases where quantitation was important, the consecutive values obtained for one lysate were totaled and the data expressed in graph form. Initially, the immunoprecipitated precursor proteins were reduced with dithiothreitol (DTT) before electrophoresis. The unlabeled precursor protein to the keratan sulfate proteoglycan migrated, however, not as a typical protein band but as a crescent-shaped arc preceding the smaller IgG chain. This artifact did not occur when reduction with DTT was omitted. Consequently, all SDS-PAGE were run unreduced and the molecular weight estimates were made under these conditions using high-molecular-weight protein standards (Bio-Rad). Glycosaminoglycan digestions. The glycosaminoglycan side chain on the immunoprecipitated [%S]methionine-radiolabeled proteoglycans was identified by enzymatic removal of the side chains and by looking for a shift in the migration of the radiolabeled core protein on SDS-PAGE. The digestion was done on the beads after the last wash but before boiling in the presence of SDS (20). Keratanase, which requires an adjacent sulfated GlcNAc, was used to degrade keratan sulfate, and chondroitinase ABC was used to degrade chondroitin/dermatan sulfate. Endo/3-galactosidase, which does not require a sulfated adjacent GlcNAc, was used to degrade lactosaminolactan. Protease inhibitors were present during digestion (20,211. The keratanase and chondroitinase ABC were purchased from Seikagaku America Incorporated and endo-P-galactosidase was purified from Escherichia freundii (22). RESULTS

The specificity of the antisera was demonstrated by Western blot of cornea1 proteoglycans digested with chondroitinase ABC and keratanase combined (Fig. 1). The antiserum to the chondroitin sulfate proteoglycan core protein reacted with a broad band that ranged from M, 40,000 to 42,000 (Fig. 1, lane 1). The antiserum to the keratan sulfate proteoglycan core protein reacted with a major broad band that ranged from M, 50,000 to 54,000. This antiserum also reacted to a much lesser extent with the same band (M, 40,000 to 42,000) that reacted with the antiserum to the chondoritin sulfate proteoglycan. Preimmune sera were blank (not shown). Previous studies (23,24) showed that bovine and human cornea1 fibroblasts in culture synthesize only low levels of keratan sulfate proteoglycan. Corneas in organ culture, in contrast, synthesize substantial amounts of keratan sulfate proteoglycan (7, 11, 18), although the amount declines by half after only 20 h of organ culture (11). Consequently, the initial immunoprecipitations were conducted from extracts of five Day 15 chick embryo corneas that were radiolabeled for 1 h in organ culture with

AND

HASSELL

FIG. 1. Western blots of chick cornea1 proteoglycans isolated from Day 18 chick embryonic corneas and digested with chondroitinase ABC and keratanase combined. Two micrograms of proteoglycan was applied to each lane. Lane 1, antiserum to the chondroitin sulfate proteoglycan core protein; lane 2, antiserum to the keratan sulfate proteoglycan core protein. Lanes incubated with preimmune serum were blank (not shown).

[35S]methionine. Acrylamide gel electrophoresis of the extract followed by autoradiography shows that the radiolabel was incorporated into proteins with a wide range of molecular sizes (Fig. 2, lane 1). Immunoprecipitation from the extract with preimmune serum did not precipitate any radiolabeled protein bands (Fig. 2, lane 2). Immunoprecipitation with antiserum to keratan sulfate proteoglycan core protein, however, precipitated a major somewhat broad band at M, 51,000. This broad band could sometimes be resolved into a doublet of M, 52,000 and 50,000 with shorter exposure times (Fig. 2, lane 6). There was also a fainter, sharper band at M, 40,000, and there were several broad bands that appeared as smears at the top one-third (M, 96,000 and greater) of the lane. Immunoprecipitation of all these bands was substantially blocked by 5 pg of keratan sulfate proteoglycan core protein (Fig. 2, lane 4) and completely blocked by 50 pg of core protein (Fig. 2, lane 5). A pulse-chase experiment was done next to establish the identity of the various-sized radiolabeled materials that were immunoprecipitated with the antibodies to the keratan sulfate proteoglycan. Ten corneas were radiolabeled for 1 h and five were harvested as before. The remaining five corneas were incubated for an additional 5 h in medium lacking radiolabel but containing normal levels of unlabeled methionine (chase). Immunoprecipitation from extracts of [35S]methionine-pulsed corneas with antiserum to the keratan sulfate proteoglycan produced a banding pattern (Fig. 3, lane 2) similar to that obtained previously (Fig. 2, lane 3). Antiserum to the chick cornea1 chondroitin sulfate proteoglycan, however, immunoprecipitated primarily the n/i, 40,000 protein and

CHICK

CORNEAL 12345

116---

KERATAN

SULFATE 6

1

97--

FIG. 2. Immunoprecipitation from [%I-methionine-pulsed Day 15 embryonic chick corneas with antibodies to chick cornea1 keratan sulfate proteoglycan core protein. Lam+ 1, cornea1 lysate; lane 2, preimmune serum; lane 3, antiserum to the keratan sulfate proteoglycan; lane 4, antiserum to the keratan su1fat.e proteoglycan in the presence of 5 pg of keratan sulfate proteoglycan core protein; lane 5, antiserum to the keratan sulfate proteoglycan core protein in the presence of 50 pg of keratan sulfate proteoglycan core protein; lane 6, same as lane 3.

substantially less of the M, 51,000 protein (Fig. 3, lane 3). The identity of the Mr 51,000 and 40,000 proteins as precursor proteins to keratan sulfate and chondroitin sulfate proteoglycans, respectively, was confirmed by immunoprecipitations from1 extracts of the pulse-chased corneas, where the radiolabeled precursor proteins had sufficient time for processing to proteoglycans. Antiserum to keratan sulfate proteoglycan now immunoprecipitated the material that resolved as the broad smear in the top one-third of the lane and relatively little of the M, 51,000 protein (Fig. 3, lane 5). This smear is artificially divided by the presence of large amounts of the IgG band that migrates between the M,. 116,000 and 200,000 markers. Digestion of this immunoprecipitated material with keratanase before electrophoresis shifted the migration position of most all of the radiolabel to the position of the M, 51,000 core protein and no radiolabel was detected at the position of the M, 40,000 protein (Fig. 3, lane 6). Digestion with chondroitinase ABC, however, did shift some of the radiolabel in the smear in the top one-third of the lane to a M, 42,000 core protein (Fig. 3, lane 7). Antiserum to the chondroitin sulfate proteoglycan also immunoprecipitated material that resolved as a smear at the top one-third of the lane (Fig. 3, lane 8). Digestion of this immunoprecipitated material with keratanase before electrophoresis shifted some of the radiolabel to a n/i,

PROTEOGLYCAN

PRECURSOR

57

PROTEIN

51,000 band (Fig. 3, lane 9). Digestion with chondroitinase ABC, however, shifted considerably more radiolabel to a M, 42,000 band (Fig. 3, lane 10). These results indicate that the broad smears at the top of the gel lanes are proteoglycans, which are known to migrate in this manner on acrylamide gels. The M, 51,000 protein immunoprecipitated from pulsed corneas is a precursor protein to a keratan sulfate proteoglycan with a M, 51,000 core protein, while the M,. 40,000 protein immunoprecipitated in the pulsed corneas is a precursor protein to a chondroitin sulfate proteoglycan with a M, 42,000 core protein. The cornea1 cell type responsible for producing the precursor protein to the keratan sulfate proteoglycan was also determined by immunoprecipitation from [35S] methionine-pulsed cornea1 epithelium, cornea1 stroma, and cornea1 stromal fibroblasts with antiserum to the keratan sulfate proteoglycan. These immunoprecipitations were done from equal amounts of incorporated radiolabel from each of the different lysates so that the relative rates of precursor protein synthesis could be

Pulse1

2

3

4

5

6

7

8

9

10

FIG. 3. Immunoprecipitation from [%]methionine-pulsed and pulsechased Day 15 embryonic corneas with antisera to both the chick keratan sulfate proteoglycan core protein and the chick chondroitin sulfate core protein. Lane 1, pulsed corneas with preimmune serum; lane 2, pulsed corneas with antiserum to the keratan sulfate proteoglycan; lane 3, pulsed corneas with antiserum to the chondroitin sulfate proteoglycan; lane 4, pulse-chased corneas with preimmune serum; lane 5, pulse-chased corneas with antiserum to the keratan sulfate proteoglycan; lane 6, pulsechased corneas with antiserum to the keratan sulfate proteoglycan and digestion with keratanase; lane 7, pulse-chased corneas with antiserum to the keratan sulfate proteoglycan and digestion with keratanase; lane 7, pulse-chased corneas with antiserum to the keratan sulfate proteoglycan and digestion with chondroitinase ABC; lane 8, pulse-chased corneas with antiserum to the chondroitin sulfate proteoglycan and digestion with keratnase; lane 10, pulse-chased corneas with antiserum to the chondroitin sulfate proteoglycan and digestion with chondroitinase ABC.

58

SCHRECENGOST,

BLOCHBERGER,

compared directly. An autoradiogram from a short exposure to the acrylamide gel for this experiment is shown in Fig. 4. Immunoprecipitation from pulsed corneas was included for comparison (Fig. 4, lane 2). Keratan sulfate proteoglycan precursor protein synthesis was not detected in the epithelium (Fig. 3, lane 4) but it was produced by cornea1 stroma lacking epithelium (Fig. 4, lane 6). The precursor protein was also produced by cornea1 fibroblasts 2 h after plating (Day 0) (Fig. 3, lane 8). The precursor protein synthesis was also detected in fibroblasts maintained in culture for 1 day (Fig. 4, lane 10) and 3 days (Fig. 4, lane 12), although it appeared to be in lesser amounts than at Day 0 of culture (Fig. 4, lane 8). The keratan sulfate proteoglycan precursor protein produced by cornea1 fibroblasts in culture was not as broad a band as that usually produced by intact corneas, even when the autoradiogram was overexposed. It migrated at M, 50,000, the size of the lower band of the doublet often seen in immunoprecipitations from cornea1 lysates. The amount of precursor protein synthesis by these cells and tissues was quantitated by cutting out the region of the acrylamide gel containing the precursor protein and measuring radioactivity by liquid scintillation counting (Fig. 5). Using the level of precursor protein produced by intact corneas as a standard for comparison, precursor protein

Cornea1

Cornea1

Fibroblasts

i2QuEaEoith.Stroma~~Dav-3 1

23

4

5

6

7

8

9

10

11

12

200----

z

116----

f:

97----

x

43----

FIG. 4. Immunoprecipitation from [%]methionine-pulsed corneas, cornea1 stroma, cornea1 epithelium, and cornea1 fibroblasts with antiserum to the keratan sulfate proteoglycan. Lanes 1, 3, 5, 7, 9, and 11 contain preimmune serum. Lanes 2,4, 6, 8, 10, and 12 contain antiserum to the keratan sulfate proteoglycans. Lanes 1 and 2, pulsed corneas; lanes 3 and 4, pulsed cornea1 epithelium; lanes 5 and 6, pulsed cornea1 stroma lacking epithelium; lanes 7 and 8, cornea1 fibroblasts pulsed 2 h after plating; lanes 9 and 10, cornea1 fibroblasts pulsed 1 day after plating; lanes 11 and 12, cornea1 fibroblasts pulsed 3 days after plating.

AND

HASSELL

15 Day Chick Embryo

Corneas and Fibroblasts

5oo’ 400

300

200

.\\\\\ I,,,, \\\\\\ I,,,, L\\\\\ ,111, \\\I.\ ,111, ,I\\\\ ,,,,, 100

Identification of chick corneal keratan sulfate proteoglycan precursor protein in whole corneas and in cultured corneal fibroblasts.

The precursor protein to the chick corneal keratan sulfate proteoglycan was identified by immunoprecipitation with antiserum to its core protein from ...
3MB Sizes 0 Downloads 0 Views