PG-M/Versican-like Proteoglycans Are Components of Large Disulfide-stabilized Complexes in the Axolotl Embryo

doi:10.1074/jbc.272.6.3246

PG-M/Versican-like Proteoglycans Are Components of Large Disulfide-stabilized Complexes in the Axolotl Embryo*

From the Department of Veterinary Medical Chemistry, Swedish University of Agricultural Sciences, The Biomedical Center, S-751 23 Uppsala and the
^§ Department of Environmental and Developmental Biology, Uppsala University, Norbyvägen 18 A, S-752 36 Uppsala, Sweden

^‡ To whom correspondence should be addressed:
Dept. of Veterinary Medical Chemistry, SLU, BMC, Box 575, S-751 23 Uppsala, Sweden.
Tel.: +46-18-174276; Fax: +46-18-550762; E-mail: michael.stigson{at}vmk.slu.se

Abstract

Large disulfide-stabilized proteoglycan complexes were previously shown to be synthesized by the epidermis of axolotl embryos during stages crucial to subepidermal migration of neural crest cells. We now show that the complexes contain PG-M/versican-like monomers in addition to some other component with low buoyant density. Metabolically ³⁵S-labeled proteoglycans were extracted from epidermal explants and separated by size exclusion chromatography and density equilibrium gradient centrifugation. The complexes, which elute in the void volume on Sepharose CL-2B, were recovered at buoyant density 1.42 g/ml in CsCl gradients, whereas the monomer proteoglycans, which could only be liberated from the complexes by reduction, had a higher buoyant density (1.48 g/ml). The native complexes did not aggregate with hyaluronan. The purified complexes reacted with antibodies against a portion of a cloned PG-M/versican-like axolotl proteoglycan. These antibodies were found to stain the subepidermal matrix of axolotl embryos, suggesting that the proteoglycan complexes are encountered by neural crest cells during subepidermal migration. From Western blot analysis, the core protein of the PG-M/versican-like monomers was found to be of similar size (≈500 kDa) as those of PG-M/versican variants of other species. Another chondroitin sulfate proteoglycan that was present in small amounts in the epidermal extracts was found to be distinctly different from the similarly sized PG-M/versican-like monomers.

Footnotes

↵* This work was supported by Grant 6525 from the Swedish Medical Research Council, Konung Gustaf V:s 80-Årsfond, and the Swedish Natural Science Research Council Grant B-AA/BU 03810-311. The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵¹ The abbreviations used are:

PG

proteoglycan

HMPG

high molecular weight PG

IMPG

intermediate molecular weight PG

CS

chondroitin sulfate

DS

dermatan sulfate

AxPG

axolotl proteoglycan

GST

glutathione S-transferase

PAGE

polyacrylamide gel electrophoresis

PBS

phosphate-buffered saline

NEM

N-ethylmaleimide.
↵² Currently, an axolotl PG (AxPG) with homology to PG-M/versican is being cloned (M. Stigson and L. Kjellén, work in progress). Briefly, a cDNA probe corresponding to a portion of the conserved C-terminal domain of chicken PG-M (22) prepared from a partial cDNA clone generously provided by Drs. K. Kimata and T. Shinomura was used for the screening of an oligo(dT)-primed embryonic axolotl cDNA library. A 4.67-kilobase pair AxPG partial cDNA clone (the full transcript was estimated by Northern analysis to be around 12-13 kilobase pairs) was sequenced and found to consist of 3422 of coding and 1245 base pairs of 3′-untranslated sequence. The deduced amino acid sequence comprises the C-terminal 1140 amino acid residues of the core protein including 830 amino acid residues of the CS attachment domain and the complete conserved C-terminal end with two epidermal growth factor-like domains, a lectin-like domain, and a complement regulatory protein-like domain. The C-terminal domain of AxPG shows an extensive amino acid sequence identity to that of the published chicken (22), mouse (24), and human (51) PG-M/versican, i.e. 83% identity to the avian and 78% to the mammalian species, whereas the similarity to other large PGs is significantly lower, e.g. 58% identity to chicken aggrecan (52) and 56% to rat neurocan (53). However, the CS attachment region apparently shows limited sequence conservation between the species and was chosen for expression as a fusion protein and generation of AxPG-specific antibodies (cf 54).
↵³ Introducing ion exchange chromatography on DEAE-Sephacel as a first step before Sepharose CL-2B chromatography or CsCl gradient centrifugation was found to result in a significantly reduced recovery of HMPGs compared with IMPGs (M. Stigson, unpublished data). For unknown reasons, HMPGs appear to bind irreversibly to the ion exchange resin and are only partly eluted by increasing the salt concentration. This property of the HMPG was taken advantage of for purification of IMPGs, which were enriched after ion exchange chromatography of HMPG-contaminated IMPG preparations.
↵⁴ Since the buoyant density of the HMPG monomers is lower, 1.48 g/ml (Fig. 2) versus 1.55 g/ml (Fig. 1) for IMPG, while their glycosaminoglycan side chains are longer, M_r ≈150,000 versus M_r ≈ 75,000 for IMPG (Fig. 3), HMPG is likely to have fewer side chains per core protein molecule than IMPG. Nevertheless, the immunoreactivity of IMPGs to anti-CS-stub antibodies was found to be lower than for the HMPGs when equal amounts of ³⁵S radioactivity of purified PGs were analyzed by slot blotting (Fig. 6). A restricted access of primary antibodies due to a relative epitope clustering could account for the lower immunoreactivity of IMPG.
↵⁵ When epidermal explants were simultaneously labeled with [³H]leucine and [³⁵S]sulfate (M. Stigson, unpublished data) and analyzed by chromatography on Sepharose CL-2B, essentially all ³H-labeled macromolecules of the epidermal explants eluted with K_av ≈0.8, whereas <1% of the ³H radioactivity was recovered in the void volume. The low protein content of the HMPG fraction did not allow for further characterization of its protein component.
- Received June 5, 1996.
- Revision received October 11, 1996.