743 NATURE VOL. 323 23 OCTOBf-'-'-..R-'-'.-198"-'-.6------MATTERS ARISING -_.. Osteoclasts and haematopoietic haematogeneous progenitor cells seeded mammalian body either in its free form or stem cells in in the perichondrium or from bound covalently or electrostatically by haematopoietic multi potent progenitors ionic forces to proteins. Heparin is found developing human bones committed to differentiate into osteoclasts in small amounts in the mammalian sys under the influence of the long bone tem and its biosynthesis, functions and THE stimulating paper by Scheven et al. 1 model. Osteoclasts excavate a marrow degradation involve enzyme systems concluded that murine osteoclast-forming cavity in the calcified hypertrophic car different from those involved in the meta capacity rises with increasing haem tilage area, creating room for stroma for bolism of heparan sulphate. The distinc atopoietic stem cell purity when cocul mation. Osteoclasts and stromal cells are tiveness of the two metabolic systems is tured with preosteoclast-free embryonic together responsible for the establishment expressed in human pathology as, for long bones. However, a growth-factor-Iike of a microenvironment for homing and example, Sanfilippo disease type A, effect of introduced stem cells should be maintenance of haematopoietic stem cells, characterized by the lack of heparan sul considered. which are transported via the blood stream phamidase. Only changes in heparan sul In 1982 we called attention to the fact from the early haematopoietic organ phate metabolism can be detected; that the appearance of osteoclasts in May (liver). Considering this sequence of heparin metabolism is unaffected. This is Griinwald-Giemsa-stained smears and events it is not surprising that the osteo cited to show the biological and bio serial sections from different bones in the clasts appear before the onset of bone chemical differences between those two course of human haematopoietic develop marrow haematopoiesis. compounds. The structural differences ment precedes the onset of haematoroiesis Finally, we are convinced that the multi between the two acid glycosaminoglycans in each bone by about two weeks. This nucleated cells formed during co-culture are well documented. The major difference accords with the development of the with periostless metatarsal bones are true is the degree of sulphation and the length haematopoietic stroma. So if osteoclasts osteoclasts and not macrophage polykary of chains. As the degree of sulphation is are derived from haematopoietic stem ons. The multinucleated cells have responsible for the binding mechanisms cells their presence in the marrow could morphological, enzyme-histochemical of this highly negatively charged poly be regarded as a very early 'footprint' of (tartrate-resistant acid phosphatase) and anion, heparin should have been used these otherwise undetectable stem cells. functional (resorption) characteristics of rather as a control substance in addition Osteoclasts, formerly believed to be of osteoclasts. Furthermore, in several non to the use of heparan sulphate in the pro the monocyte/ macrophage lineage could published experiments we have used quail tein interaction studies. be derived from a marrow stem cell unre bone marrow as exogenous source of The use of heparitinase, which like lated to the stem cell for other blood cells. osteoclast progenitors. The multinu heparan sulphate, is available, could have There could also still be some doubt about cleated cells developed during co-culture helped or could be suggested for confirma the nature of giant cells appearing during with fetal mouse long bones could be rec tion of the data obtained. long-term bone marrow culture, that is, in ognized by their specific chromatin Chondroitin sulphate is not a good con long-term culture of newborn rabbit bone organization as quail cells, and by their trol in our opinion as it is completely marrow giant cells have been character reaction with osteoclast-specific mono different from heparan sulphate and ized as macrophage polykaryons rather clonal antibodies I as osteoclasts. heparin: the sulphate in this compound is 3 than osteoclastic giant cells • BEN A. A. SCHEVEN bound to the polysaccharide skeleton by E. KELEMEN PETER J. NUWEIDE O-sulphation and not, as in heparan sul First Department of Medicine, Laboratory of Cell Biology phate and heparin by N -sulphation. Thus, Semmelweis University, and Histology, even if chain length and degree of sulpha tion of the glycosaminoglycan were com Koranyi u 2/ A, University of Leiden, parable, the unknown charge distribution 1083 Budapest, Hungary Rijnsburgerweg 10, 1 2333 AA Leiden, The Netherlands of 'chondroitin sulphate mixed isomers' could not be compared. The reactivity of J. Scheven, B. A. A., Visser, 1. W. M. & Nijweide, P. J. Natun heparan sulphate (and heparin) with pro 321. 79-81 (1986). I. Nijweide. P. J. tI aL Hisloch.mislry 83, 315-324 (1985). 2. Kelemen, E. &: Calvo, W. in 11tr Human Bon~ Morrow (eds teins is enormous. This has been shown Trobowitz. S. '" Davis,S.) Vol. I. 3-41 (CRC, Boca by several authors. Gelman and coworkers Raton. Florida, 1982). 3. Honan. M. A.• Rimmer, E. F. '" Chambers, T. J. J. Bo•• studied the glycosaminoglycan-protein Mintra/ R.s. 1,5-14 (1986). Interactions of N-CAM interaction using circular dichroism2 and with heparin-like molecules Doyle and associates published the inter SCHEVEN AND NIJWEIDE REPLY action between collagen and acid gly Because the haematopoietic cell popula COLE ET AL.1 tried to answer the question cosaminoglycans indicating that sul tions used in our study were not 100% of whether the binding of heparan sul phated glycosaminoglycans were binding pure, we cannot completely exclude the phate to the neural cell adhesion molecule, (nonspecifically) to clustered, basic, posi 3 possibility that other stem cells, not iden N-CAM, is also required for cell adhesion. tively charged amino-acid residues • tical to the haematopoietic stem cell, were They showed that the binding of retinal Laminin, an intrinsic glomerular base co-purified and gave rise to osteoclast for probe cells to retinal cell monolayers was ment membrane protein which is involved mation. However, the fact that interleukin- inhibited by heparin, a substance similar in cell-basement membrane interactions, 3 pretreatment of the stem cell prepar to heparan sulphate; but not by chon and fibronectin interact with heparan sul 4 ations stimulated osteoclast development, droitin sulphate. Monoclonal antibodies phate and heparin ,5. So does collagen as well as formation of granulocytes and that recognized two different domains on type IV and the native glomerular base macrophages, strengthen the possibility N-CAM, the homophilic and heparin ment membrane, containing covalently 6 that the haematopoietic stem cell is also binding domains, inhibited cell adhesion. linked heparan sulphate-proteoglycan . the progenitor for osteoclasts. The heparin-binding domain isolated Thus, heparan sulphate seems to be a The first appearance of osteoclasts in from N-CAM by selective proteolysis also highly reactive substance resembling other developing embryonic bones should not inhibited cell-cell adhesion when bound compounds (such as laminin, actin and be regarded as a footprint of the presence to the probe cells. fibronectin) with 'glue-activity'. of haematopoietic stem cells. Osteoclasts Heparan sulphate is one of the major In our hands this acid-glycosamino differentiate in the cartilagenous models constituents of the extracellular matrix, is glycan reacts with commercially available of long bones either from committed found on cell surfaces, and throughout the polylysine as a marker substance for
© 1986 Nature Publishing Group