Proc. Nadl. Acad. Sci. USA Vol. 91, pp. 2888-2889, April 1994 Commentary B lymphopoiesis: Global factors, local control Paul W Kincade

Okahom Medical Research Foundation, 825 Northeast 13th Sheet, Oklahoma City, OK 73104

The bone marrow is the only site in adult unique strategy. They prepared cDNA sites found in a number of cytokines. mammals where B lymphocytes differen- libraries enriched for genes with consen- These were used to screen a cDNA li- tiate from stem cell precursors. Much sus signal sequences of transmembrane brary. One of several clones encoded a effort has therefore been devoted to and secreted proteins and used a vector 52-kDa protein that stimulates growth of learning what is unique about this highly that places an epitope tag on expressed pre-B cells in the presence of IL-7 and specialized tissue. While a number of proteins. Selection of transfected COS-7 stem cell factor. Transcripts for this pro- factors influence the replication and dif- cells with an antibody to that epitope tein, pre-B-cell colony-enhancing factor ferentiation of lymphohemopoietic pre- increased the probability of isolating (PBEF), were found in many tissues. cursors (1-3), none is restricted in its open reading frames. Although PBSF/ IL-7, PBSF, BST-1, IGF-I, and PBEF distribution to bone marrow. This situa- SDF-la was cloned from a stromal cell all appear to be positive regulators of tion can now be generalized to five more library, its expression is widespread. B-cell formation. While negative regula- molecules that have recently been found The route to discovery is often as in- tors have been less extensively studied, to influence the survival and proliferation teresting as the discovery itself. This is they may also enjoy a wide tissue distri- of B-cell precursors. Indeed, some of certainly the case for another pre-B-cell bution. A potentially important regula- them are hormones previously known for costimulatory factor, BST-1 (12). Hirano tory mechanism for B lymphopoiesis was their effects on other biologic processes. and colleagues prepared numerous trans- unexpectedly discovered in this labora- It appears likely that B-lineage differen- formed bone marrow stromal cell lines tory (18). B-cell precursors, including tiation in the bone marrow microenviron- from normal individuals and patients IL-7-responsive cells, markedly and se- ment is determined by a unique combi- with rheumatoid arthritis. Some of these lectively declined in bone marrow of nation and concentration of molecules human cell lines, and especially those pregnant mice. Recent immirants in the rather than by tissue-restricted factors. from arthritis patients, supported sub- spleen also declined, but mature B-cell Interleukin 7 (IL-7), which is made by stantial proliferation ofthe murine DW34 numbers were unaffected. Suspecting keratinocytes and stromal cells in bone pre-B-cell line. A monoclonal antibody that these changes related to hormonal marrow, spleen, thymus, and kidney, is that preferentially recognized stromal fluctuations during pregnancy, we found the best established replication factor for cells with high support capability was that estrogen was particularly effective pre-B cells (4). Especially compelling are prepared and used to isolate the gene for depletion of B-lineage precursors in two recent studies showing that pre-B encoding this antigen. The BST-1 gene nonpregnant mice (19). Progesterone cells are virtually absent from mice in- encoded a glycosyl phosphatidylinositol- alone had no effect, but it dramatically jected with antibodies to either IL-7 or its anchored protein that was, as expected, reduced the effective dose of estrogen receptor (5, 6). However, while IL-7 is highly expressed on rheumatoid stromal when the two hormones were given to- necessary for B-lymphocyte formation, it cells. Moreover, when introduced by gether. Glucocorticoids also deplete is clearly not sufficient. Additional sig- transfection, the BST-1 gene conferred pre-B cells (20), but initial culture exper- nals, presumably present in the bone the DW34 support capability upon other- iments suggest an important difference in marrow microenvironment, are required wise nonsupporting fibroblasts. Finally, mechanisms of action. While glucocorti- for efficient recruitment to, and differen- as was true for PBSF, BST-1 was ex- coids may act directly to elicit apoptotic tiation of cells within, this lineage (3, 7, pressed in many normal tissues (T. pre-B-cell death, estrogen is suppressive 8). It is these signals that recent studies Hirano, personal communication). only in the presence of stromal cells (21, have sought to define in molecular terms. A third factor was isolated from stro- 22). Kishimoto and colleagues (9) used the mal cells because ofits ability to enhance Additional studies with hypogonadal stromal cell-dependent pre-B-cell clone the emergence of B-lineage cells in cul- animals suggest that sex steroids contrib- DW34 as a biological indicator to clone a ture and was identified as insulin-like ute to normal steady state regulation (G. soluble mediator, which they designated growth factor I (IGF-I) (13). There is no Smithson, W. G. Beamer, K. L. Shultz, pre-B-cell growth stimulatory factor indication that this ubiquitous hormone S. W. Christianson, L. D. Schultz & (PBSF) (9). The 89-amino acid protein alone stimulates growth of B-cell precur- P.W.K., unpublished data). The mice are stimulated proliferation of DW34 cells sors. However, it does promote prolifer- genetically incapable of synthesizing and enhanced their response to IL-7. ation of IL-7-dependent lymphocyte these hormones and have greatly ele- Moreover, while PBSF markedly aug- clones and normal marrow cells, which vated B-lineage precursors in bone mar- mented the proliferative effect of IL-7 on have an early phenotype (14, 15). Fur- row. Sex steroids must therefore be freshly isolated bone marrow cells, it had thermore, lymphocytes, including pre-B added to arelatively small list ofpotential little effect when used alone. cells in bone marrow, increased substan- negative regulators of B lymphopoiesis PBSF has structural homology to tially in mice infused with recombinant (22). There are interesting parallels to members of the intercrine a family, IGF-I (ref. 16; K. Dorshkind, personal studies of the thymus, where the same which includes proinflammatory cyto- communication). hormones suppress proliferation of im- kines already known to influence repli- RNA from activated lymphocytes was mature T lymphocytes, and age-related cation of stem cells and myeloid progen- the starting point for isolation of another involution of the thymus is partially re- itors (10). A PBSF clone, termed SDF- cytokine gene (17). Samal and colleagues stored by castration (reviewed in ref. 22). la, had previously been isolated by used degenerate oligonucleotides corre- All of these results argue against the Honjo and colleagues (11), who devised a sponding to signal peptidase processing hypothesis that B-cell differentiation in 2888 Downloaded by guest on October 2, 2021 Commentary: Kincade Proc. Natl. Acad. Sci. USA 91 (1994) 2889

marrow is determined by tissue-specific bone marrow microenvironment have kawa, S. & Nishikawa, S. (1990) J. Exp. cytokines. What are the alternatives? been identified, and discovery of these Med. 171, 1683-1695. The effects of hormones such as eryth- cytokines/interaction molecules raises a 9. Nagasawa, T., Kikutani, H. & Kishi- ropoietin are primarily restricted to mar- With the moto, T. (1994) Proc. NatI. Acad. Sci. number of questions. excep- USA 91, 2305-2309. row because erythroid progenitors there tions of IL-7 and sex hormones, we have 10. Oppenheim, J. J., Zachariae, C. 0. C., are among the very few cell types with little idea of their importance in vivo and Mukaida, N. & Matsushima, K. (1991) receptors for it. In contrast, estrogen and it is probable that a degree ofredundancy Annu. Rev. Immunol. 9, 617-648. IGF-I receptors are found on many cell will be demonstrated. Indeed, overlap- 11. Tashiro, K., Tada, H., Heilker, R., Shi- types in many tissues. Bone marrow ping functions at the level ofcytokines or rozu, M., Nakano, T. & Honjo, T. (1993) could be the unique site of B-cell emer- their receptors provide hemopoietic cells Science 261, 600-603. gence because the combination and con- with the potential forhighly sophisticated 12. Kaisho, T., Ishikawa, J., Oritani, K., centrations of essential cytokines are responses (31). All five of the recently Inazawa, J., Tomizawa, H., Muraoka, unique, even though none of those fac- described molecules influence the sur- O., Ochi, T. & Hirano, T. (1994) Proc. tors is individually restricted to that lo- vival and/or proliferation of IL-7- Nati. Acad. Sci. USA, in press. In this respect, marrow might 13. Landreth, K. S.,Narayanan,R. &Dorsh- cation. responsive cells. There is a pressing need kind, K. (1992) Blood 80, 1207-1212. resemble the gonads, which also provide to learn about other cellular targets and 14. Gibson, L. F., Piktel, D. & Landreth, unique microenvironments for differenti- whether these newly discovered media- K. S. (1993) Blood 82, 3005-3011. ation of highly specialized cells. Local tors also influence differentiation of 15. Funk, P. E., Kincade, P. W. & Witte, production of sex steroids in the gonads B-lymphocyte precursors. While B cells P. L. (1994) Blood 83, 361-369. depends on sequential conversion of are well established as models for selec- 16. Clark, R., Strasser, J., McCabe, S., Rob- blood-borne precursors by two closely tive gene rearrangement and expression, bins, K. & Jardieu, P. (1993) J. Clin. coordinated cell types (Sertoli/Leydig Invest. 92, 540-548. we know comparatively little about the Y., G., Xie, C., and theca/granulosa cells) (23). Simi- environmental cues that trigger these 17. Samal, B., Sun, Stearns, aromatases made by fat cells in Suggs, S. & McNiece, I. (1994) Mol. larly, events. Cell. Biol. 14, 1431-1437. marrow might permit local production of 18. Medina, K. L., Smithson, G. & Kin- estrogen in excess ofsystemic levels (24). Dr. Philip Silverman provided particularly cade, P. W. (1993) J. Exp. Med. 178, Immobilization of cytokines on extracel- helpful comments on the manuscript. Studies 1507-1515. lular matrices or via transmembrane an- done in the author's laboratory are supported 19. Medina, K. L. & Kincade, P. W. (1994) choring domains could also achieve high by Grants AI-20069 and AI-33085 from the Proc. Natd. Acad. Sci. USA, in press. local concentrations of essential signals. National Institutes of Health. 20. Garvy, B. A., King, L. E., Telford, Locally produced enzymes might con- W. G., Morford, L. A. & Fraker, P. J. tribute to a unique bone marrow microen- 1. Kincade, P. W. (1991) Semin. Immunol. (1993) Immunology 80, 587-592. vironment in other ways. Stromal cells 3, 379-390. 21. Smithson, G. M., Medina, K. L. & Kin- 2. Hirayama, F., Katayama, N., Neben, S., cade, P. W. (1993) J. Immunol. 150, 20A synthesize both IGF-I and several of the Donaldson, D., Nickbarg, E. B., Clark, (abstr.). six known IGF binding proteins (IGF- S. C. & Ogawa, M. (1994) Blood 83, 22. Kincade, P. W., Medina, K. L. & Smith- BPs) (25). IGF-BPs can either enhance or 92-98. son, G. (1994) Immunol. Rev. 137, in inhibit the cellular effects of IGF-I, with 3. Kee, B. L., Cumano, A., Iscove, N. N. press. the outcome determined by posttransla- & Paige, C. J. (1994) Int. Immunol. 6, in 23. Dorrington, J. H., Bendell, J. J. & tional modifications catalyzed by kinases press. Khan, S. A. (1993) J. Steroid Biochem. and proteases (26). Proteolysis of IGF- 4. Namen, A. E., Lupton, S., Hjerrild, K., Mol. Biol. 44, 441-447. BPs can be triggered by IGF-I binding Wignall, J., Mochizuki, D. Y., 24. Frisch, R. E., Canick, J. A. & Tulchin- (27). A number of membrane-bound ec- Schmierer, A., Mosley, B., March, sky, D. (1980) J. Clin. Endocrinol. C. J., Urdal, D., Gillis, S., Cosman, D. Metab. 51, 394-3%. toproteases are known to be expressed & Goodwin, R. G. (1988) Nature (Lon- 25. Abboud, S. L., Bethel, C. R. & Aron, on stromal cells and/or B-cell precursors don) 333, 571-573. D. C. (1991) J. Clin. Invest. 88,470-475. (reviewed in ref. 28). Secreted proteases 5. Grabstein, K. H., Waldschmidt, T. J., 26. Jones, J. I., D'Ercole, J. D., Comaco- might also contribute to the bone marrow Finkelman, F. D., Hess, B. W., Alpert, Hubner, C. & Clemmons, D. R. (1991) microenvironment through effects on cy- A. R., Boiani, N. E., Namen, A. E. & Proc. Natd. Acad. Sci. USA 88, 7481- tokines. For example, cathepsins re- Morrissey, P. J. (1993) J. Exp. Med. 178, 7485. leased by breast stromal cells in response 257-264. 27. Conover, C. A., Kiefer, M. C. & Zapf, J. to estrogen are capable of activating la- 6. Sudo, T., Nishikawa, S., Ohno, N., Ak- (1993) J. Clin. Invest. 91, 1129-1137. tent type transforming growth factor iyama, N., Tamakoshi, M., Yoshida, H. 28. Shipp, M. A. & Look, A. T. (1993) & Nishikawa, S. (1993) Proc. Nati. Blood 82, 1052-1070. (TGF-f8) (29). Other studies suggest a Acad. Sci. USA 90, 9125-9129. 29. Westley, B. R. & May, F. E. B. (1987) connection between sex hormones and 7. Lee, G., Namen, A. E., Gillis, S., Ell- Nucleic Acids Res. 15, 3773-3786. TGF-(3 activity in the thymus and ovaries ingsworth, L. R. & Kincade, P. W. 30. Olsen, N. J., Zhou, P., Ong, H. & Ko- (23, 30). (1989) J. Immunol. 142, 3875-3883. vacs, W. J. (1993) J. Steroid Biochem. It is unlikely that all of the local and 8. Hayashi, S., Kunisada, T., Ogawa, M., Mol. Biol. 45, 327-332. widely disseminated components of the Sudo, T., Kodama, H., Suda, T., Nishi- 31. Metcalf, D. (1993) Blood 82, 3515-3523. Downloaded by guest on October 2, 2021