Evidence is provided in a companion paper for an IL-7-associated molecular complex that selectively stimulates the proliferation and presumptive differentiation of pre-pro-B cells in our long-term bone marrow culture system and “primes” them to proliferate in response to monomeric IL-7. Here, Western immunoblot analysis reveals that this pre-pro-B cell growth-stimulating factor (PPBSF) is a self-assembling heterodimer of IL-7 and a cofactor with a Mr of 30,000. Thus, when developed with anti-IL-7 mAb, PPBSF migrates electrophoretically as a covalently bound ∼55-kDa molecule under nonreducing conditions but dissociates under reducing conditions. Furthermore, the addition of rIL-7 or native IL-7 to medium conditioned by stromal cells from IL-7 gene-deleted (−/−) mice results in the formation of active 45-kDa and 55-kDa molecular complexes, respectively. Antiserum prepared in IL-7(−/−) mice against affinity-purified PPBSF contained separable reactivities for IL-7 and the non-IL-7 component of PPBSF. The PPBSF cofactor detected by this antiserum migrates as an ∼30-kDa molecule and is able to maintain the viability, but not the proliferation, of pre-pro-B cells. Furthermore, the cofactor is produced constitutively by IL-7(−/−) and IL-7(+/+) bone marrow stromal cells under pro-B- but not pre-B-type culture conditions. Consequently, IL-7 appears to exist almost entirely as a heterodimer (i.e., PPBSF) in pro-B-type cultures, whereas it exists almost entirely as a monomer in pre-B-type cultures. Although the identity of the PPBSF cofactor remains to be determined, it does not appear to be stem cell factor, insulin-like growth factor-1, thymic stromal-derived lymphopoietin, flt3, stromal cell-derived factor-1, or IL-7R.
Despite the demonstration over the past decade of the key role of IL-7 in the development and differentiation of murine pre-B cells and pro-B cells, the nature of its involvement, if any, at the pre-pro-B cell stage of lymphopoiesis remains controversial (1, 2, 3). Inasmuch as our long-term xenogeneic (rat/mouse/human) bone marrow (BM)5 lymphoid culture system selectively generates pre-pro-B cells and pro-B cells (4, 5, 6, 7), we have used it as a model with which to identify the stromal cell-derived growth factors that are responsible for regulating the proliferation and/or differentiation of these primitive lymphoid precursors.
Results of earlier studies have demonstrated that serum-free BM stromal cell conditioned medium (CM) from our culture system selectively stimulates the proliferation of pre-pro-B cells from freshly harvested rat BM and supports the accumulation, but not the proliferation, of pro-B cells in vitro (8). Adsorption of CM with anti-IL-7 mAb removes this activity, whereas rIL-7 restores this activity to CM by BM stromal cells from IL-7 gene-deleted mice (9). Furthermore, IL-7 with a nominal molecular mass of 25 kDa coisolates with pre-pro-B cell growth-stimulating activity in the apparent 50- to 100-kDa molecular mass fraction as determined by ultrafiltration. Yet IL-7 itself does not induce proliferation of pre-pro-B cells, even in the presence of stem cell factor (SCF) or insulin-like growth factor (IGF)-1, and anti-IL-7 mAb is unable to neutralize the growth-stimulating activity in CM. These results, in aggregate, suggested that the unique lymphopoietic properties of our BM lymphoid culture system were due to the presence of a molecular complex of IL-7 and a second stromal cell-derived factor. Although this molecular complex can also induce thymocyte proliferation (9), we have not yet detected its presence in thymic stromal cell CM (4, 8). Therefore, we have designated it pre-pro-B cell growth-stimulating factor (PPBSF).
The present study directly confirms, by Western blot analysis, that PPBSF is an ∼55-kDa heterodimer that consists of one molecule of IL-7 and one molecule of an as yet unidentified 30-kDa cofactor. The results also demonstrate that the cofactor is constitutively produced by IL-7(−/−) stromal cells under pro-B cell- but not pre-B cell-type culture conditions and that it associates covalently with IL-7. Finally, the results confirm our previous observation (9) that PPBSF “primes” pre-pro-B cells and their immediate descendants to proliferate in response to IL-7 alone. The possible role of PPBSF in early B-lineage development and the exclusion of several candidates for the PPBSF cofactor are discussed below.
Materials and Methods
Male 4- to 6-wk-old IL-7 gene-deleted (IL-7(−/−)) and nondeleted (IL-7(+/+)) mice (10) were bred from (129 × B6)F2 stock that was generously provided by Drs. Richard Murray and Ursula von Freeden-Jeffry (DNAX Research Institute of Cellular and Molecular Biology, Palo Alto, CA). The mice were used as donors of BM-adherent cells and stromal cell lines. Male 4- to 6-wk-old Lewis strain rats bred from stock originally obtained from the National Cancer Institute, National Institutes of Health (Bethesda, MD) were used as donors of BM lymphoid precursor cells.
rIL-7 was purchased from Genzyme Corporation (Cambridge, MA). Thymic stromal-derived lymphopoietin (TSLP) (11, 12) was generously provided by Dr. Philip J. Morrissey (Immunex Research and Development Corporation, Seattle, WA); rflt-3 ligand (13) was generously provided by Dr. Satish Menon (DNAX); and rSDF-1β (14) was purchased from R&D Systems (Minneapolis, MN).
Neutralizing mAbs cross-reactive with both human and mouse IL-7 or SDF-1β were purchased from Genzyme Corporation or R&D Systems, respectively. Rabbit polyclonal Abs against TSLP or mAbs against flt-3 ligand were kindly provided by Immunex or DNAX, respectively. Mouse IgG2b isotype control was obtained from Sigma (St. Louis, MO). Murine mAbs to the HIS40 (anti-IgM) (15), HIS24 (anti-CD45RC-B220) (16, 17), and HIS50 (anti-heat stable Ag (HSA)) (18) rat B-lineage-associated Ags were generously provided by Dr. Davine Opstelten (Department of Pathology, University of Hong Kong, Hong Kong, China). Mouse anti-bromodeoxyuridine (anti-BrdU) mAb (with nuclease) was purchased from Amersham International (Little Chalfont, U.K.). Affinity-purified FITC-conjugated goat IgG F(ab′)2 anti-mouse IgM (heavy chain-specific) Ab was obtained from Kirkegaard and Perry Laboratories (Gaithersburg, MD). Affinity-purified rabbit Ab to calf thymus terminal deoxynucleotidyl transferase (TdT) as well as FITC- and tetramethylrhodamine isothiocyanate (TRITC)-conjugated goat anti-rabbit IgG were purchased from Supertechs (Bethesda, MD). Phycoerythrin-conjugated goat anti-mouse IgG was obtained from Caltag Laboratories (San Francisco, CA). Horseradish peroxidase (HRP)-linked sheep anti-mouse IgG or anti-rabbit IgG were purchased from Amersham Life Sciences (Arlington Heights, IL).
Indirect immunofluorescence of cell surface Ags was performed by incubating 1 × 106 freshly harvested or culture-generated BM cells with mouse or rat primary Abs (10 μl) and developing with appropriate FITC- or phycoerythrin-conjugated goat anti-IgG or anti-IgM Abs. To detect intranuclear TdT, cytocentrifuge-prepared cell smears were fixed in 4°C absolute methanol, stained with rabbit Abs to TdT, and developed with FITC- or TRITC-conjugated Abs to rabbit IgG (19). Double immunofluorescence for Cμ or Sμ Ig heavy chains and TdT was performed on cell smears that were fixed in cold absolute ethanol with 5% glacial acetic acid for 20 min at 4°C, sequentially stained for TdT and HIS40, and developed with FITC goat anti-mouse IgG and TRITC goat anti-rabbit IgG (17).
To detect the incorporation of BrdU, cultured cells were pulsed overnight with BrdU cell-proliferation labeling reagent (Amersham International) in a final concentration of 1:1000. Cytosmears prepared from these cells were fixed in cold absolute ethanol with 5% glacial acetic acid, stained with the anti-BrdU/nuclease reaction mixture for 60 min, and developed with FITC goat anti-mouse IgG. Double immunofluorescence for BrdU and TdT was accomplished by staining for TdT at this step. Double immunofluorescence for BrdU and cell-surface Ags was performed by staining viable cells in suspension with the appropriate Abs and then staining cytocentrifuge smears of the same cells for BrdU.
Lymphoid culture systems
Rat BM pre-pro-B cells and pro-B cells were generated in our culture system as previously described (4). Briefly, single cell suspensions of mouse BM (8 × 106 cells) were added to 2 ml RPMI 1640 containing 20% lot-selected, defined FBS (HyClone, Logan, UT) in 35-mm diameter culture plate wells and incubated at 37°C in 5% CO2. After 10 days, the confluent adherent cell layers were washed and seeded with 5 × 105 freshly harvested rat BM cells/ml. In some experiments, the rat BM cells were seeded into microporous membrane culture inserts (0.4-μm pore size; Transwell-3408, Costar, Cambridge, MA) that were placed over, but not in contact with, the mouse BM-adherent cell layers. Total cells from the culture inserts and nonadherent lymphoid cells from the standard cultures were recovered in serum-free medium on day 10 for cytologic and phenotypic analysis (8).
Rat BM pre-B cells were generated in long-term culture by a modification of the method of Whitlock and Witte (20). Briefly, adherent cell layers were established by incubating 8 × 106 mouse BM cells/well at 37°C (5% CO2) in 2 ml RPMI 1640 medium containing 5% lot-selected, defined FBS (HyClone), 5 × 10−5 M 2-ME, and 40 mg/L gentamicin. The cultures were refed with 50% fresh medium twice weekly. After 20 days, the confluent adherent cell layers, containing only an occasional mouse lymphoid cell colony, were washed with RPMI 1640 and seeded with 5 × 105 freshly harvested rat BM cells/ml as above. Culture-generated lymphoid cells (>98% rat origin) were recovered on day 15 for cytologic and phenotypic analysis.
Washed confluent mouse BM-adherent cell layers or stromal cell lines therefrom were used to condition medium for 10 days (8). The CM for cell stimulation was filtered to remove any cells, concentrated twofold by ultrafiltration in Centriprep-10 concentrator units (Amicon, Danvers, MA), dialyzed for 16 h in serum-free normal medium at 4°C, and stored at −70°C. For cell stimulation, CM was diluted to twofold its original concentration with medium containing 20% FBS; for immunoadsorption or Western immunoblotting, 10× concentrated CM in serum-free normal medium was used.
Immunoadsorption of CM with anti-IL-7 mAb
Anti-IL-7 mAb (mouse IgG2b) was conjugated to protein A-Sepharose by incubating 15 μl of Ab with 80 μl of packed beads for 4 h. The beads were extensively washed with PBS to remove unbound Ab. Immunoadsorption was accomplished by incubating 10× concentrated CM with Ab-conjugated protein A-Sepharose beads (1 ml CM/80 μl packed beads) in a rotating mixer for 2 h at 4°C. The beads were pelleted in a microfuge (8000 rpm), and the supernatant was removed. This process was repeated three times. Nonspecific binding was controlled by incubating CM with unconjugated protein A-Sepharose beads and beads conjugated with a mouse IgG2b isotype control. The bound Ag was recovered from the beads by elution with 0.1 M NaHCO3 buffer (pH 9.3) containing 0.5 M NaCl, and the eluate was dialyzed for 16 h in PBS (pH 7.2) at 4°C.
To evaluate cell proliferation induced by CM, 1 × 105 freshly harvested rat thymocytes or day 10 culture-generated rat BM lymphoid cells were pulsed with 1 μCi/well of [3H]TdR (New England Nuclear, Boston, MA) 12 h before harvesting. Incorporation of [3H]TdR was determined by liquid scintillation spectroscopy.
Preparation of antisera to PPBSF
An anti-IL-7 immunoaffinity column was prepared by mixing 3 mg IL-7-specific mAb in coupling buffer with 0.3 g cyanogen bromide-activated Sepharose 4B according to the manufacturer’s instructions. One hundred milliliters of 10× concentrated serum-free CM was loaded onto the column, which was then washed with 50 ml PBS. The bound Ag was eluted with 10 ml of 0.1 M NaHCO3 buffer (pH 9.3) containing 0.5 M NaCl and dialyzed for 16 h in PBS (pH 7.2) at 4°C. The activity of the eluate was then tested by thymocyte proliferation analysis.
Equal volumes of eluate and Freund’s adjuvant were mixed, and IL-7(−/−) mice were injected with a total of 2 ml eluate from IL-7(+/+) pro-B CM, IL-7(+/+) pre-B CM, or 20 μg rIL-7 at multiple s.c. sites on a biweekly basis. The first two biweekly immunizations were performed using Ag emulsified with CFA. All subsequent immunizations were carried out with the same Ag emulsified with IFA. After four successive biweekly immunizations, sera were collected and pooled. To remove anti-IL-7 Abs, some aliquots of antisera were adsorbed with rIL-7 coupled to cyanogen bromide-activated Sepharose 4B (5 μmol rIL-7/ml gel).
Western immunoblotting of CM for PPBSF
For Western immunoblotting, 25 μl of eluate from anti-IL-7, anti-PPBSF, or anti-IL-7 mAb-adsorbed anti-PPBSF Ab immunoaffinity columns was mixed with 25 μl of 2× SDS sample buffer, with or without 0.1 M DTT, and boiled for 5 min. The samples were loaded onto different slots of a 12% SDS-PAGE gel and run overnight at 45 V. The proteins were then transferred onto Immobilon-P membrane (Millipore, Bedford, MA) using a trans-Blot SD Semidry Transfer Cell (model 200/2.0, Bio-Rad, Hercules, CA) at 300 mA for 1 h. After blocking with 5% blocking reagent in PBS-Tween 20, the membrane was incubated with appropriate dilutions of anti-IL-7 mAb, anti-TSLP polyclonal Ab, anti-flt-3 ligand mAb, anti-SDF-1β mAb, or antiserum to PPBSF; washed; incubated with 1:2000 HRP-labeled anti-mouse IgG or HRP-labeled anti-rabbit IgG; washed again; and developed with enhanced chemiluminescence Western blotting analysis system (Amersham Life Sciences).
PPBSF activity is present in pro-B- but not pre-B-type cultures
As previously documented (6) and illustrated in Figure 1,A, the pre-pro-B cell and pro-B cell compartments in our culture system progressively expand with time after inoculation with freshly harvested rat BM cells, whereas the pre-B cell compartment progressively contracts. In contrast (Fig. 1 B), the pre-B cell compartment progressively expands with time, after a brief lag, under Whitlock-Witte (W-W)-type culture conditions (20), whereas the pre-pro-B cell and pro-B cell compartments progressively contract, after a brief period of expansion. Therefore, for convenience, these culture systems will be referred to as pro-B-type and pre-B-type cultures, respectively.
The differences in the generative potentials of the two culture systems were further documented using inoculum of >95% pure mixtures of pre-pro-B cells and pro-B cells obtained from day 10 pro-B-type cultures (6). Under pro-B-type culture conditions (Fig. 1,C), the pre-pro-B cell/pro-B cell compartments progressively expanded and, as reported (8), there was a disproportional increase of pre-pro-B cells with time. However, under pre-B-type culture conditions (Fig. 1 D), pre-pro-B cells decreased and pro-B cells increased during the first 2 wk of culture, after which pro-B cells decreased and pre-B cells increased.
The results in Figure 2 show that CM from pro-B- and pre-B-type cultures supported the same patterns of lymphopoiesis, albeit at lower efficiency, as cultures containing BM-adherent cells (cf Figs. 1,A and 1B; day 11). Furthermore, as shown in Figure 3, the growth-stimulating activity for thymocytes in pre-B CM, unlike that in pro-B CM, was both neutralized and adsorbed by anti-IL-7 mAb. These results suggested that the IL-7 in pre-B-type cultures, unlike that in pro-B-type cultures, is not complexed with the cofactor previously detected in IL-7(−/−) pro-B CM (9). To verify this, we evaluated the ability of rIL-7 to restore PPBSF activity to IL-7(−/−) CM generated under pre-B-type culture conditions. No activity for pre-pro-B cells was detected (data not shown). The results below show that this was due to the absence of the PPBSF cofactor from pre-B CM.
Electrophoretic mobility and molecular mass of PPBSF
As shown in Figure 4, the difference in form of IL-7 in pro-B and pre-B CM was confirmed by electrophoresis and Western immunoblotting. Under nonreducing conditions, the IL-7 in pro-B CM migrated with a molecular mass of 55 kDa (lane 2, arrow), whereas that in pre-B CM migrated at 25 kDa (lane 1). However, under reducing conditions, the IL-7 in both pro-B and pre-B CM migrated as 25-kDa molecules. Furthermore, the IL-7-associated molecule in pro-B CM did not dissociate after treatment with 8 M urea, 1 M acetic acid, or 0.1 M NaOH (Fig. 5), suggesting that it exists as a covalently bound (presumably disulfide-linked) molecular complex.
PPBSF is a heterodimer of IL-7 and an Mr 30,000 cofactor
To confirm that PPBSF is a self-associating heterodimer, rIL-7 was added to IL-7(+/+) and IL-7(−/−) CM generated under pro-B- and pre-B-type culture conditions. One hour later, the total amount of IL-7 in these CM was affinity purified, electrophoresed under nonreducing conditions, and subjected to Western blot analysis.
The results in Figure 6 show that all detectable rIL-7 added to IL-7(−/−) pro-B CM (lane 5, arrow) migrated as part of a 45-kDa molecule, whereas the rIL-7 added to IL-7(+/+) pro-B CM (lane 3) migrated at 14.5 kDa. This suggested that the rIL-7 formed a heterodimer with an ∼30-kDa molecule in IL-7(−/−) CM. Again, the endogenous IL-7 in pro-B CM migrated as part of a 55-kDa molecule (lanes 2and 3). Conversely, all detectable rIL-7 added to either IL-7(−/−) pre-B CM (lane 5, arrow) or IL-7(+/+) pre-B CM (lane 3) migrated at 14.5 kDa, and the endogenous IL-7 migrated at 25 kDa (lanes 2and 3).
The 55-kDa molecular mass of PPBSF suggests that it consists of one molecule of IL-7 and one molecule of cofactor. Nonetheless, to exclude the possibility that IL-7 and the cofactor can associate in multiple proportions, graded amounts of rIL-7 were added to constant volumes of IL-7(−/−) pro-B CM. Only a single species of PPBSF (∼45 kDa) was detected by Western blot analysis, even when rIL-7 was added in amounts that ranged between 10-fold above and below that required for maximum complex formation (data not shown).
Similarly, results in Figure 7 show that native IL-7 in pre-B CM complexes with an ∼30-kDa molecule when added to IL-7(−/−) pro-B CM (lane 3, arrow). However, when mixed with IL-7(+/+) pro-B CM (lane 1), native IL-7 continues to migrate as a 25-kDa molecule. The failure of IL-7 to exist as a heterodimer in pre-B CM does not appear to be due to an inhibitory effect of 2-ME on complex formation. This was shown by the addition of 2-ME to pro-B CM. Under these conditions, 2-ME neither caused the PPBSF in IL-7(+/+) CM to dissociate nor prevented rIL-7 from forming PPBSF when added to IL-7(−/−) CM (data not shown). Therefore, as confirmed below, the PPBSF cofactor appears to be absent from pre-B CM.
Antiserum to PPBSF has separable specificities for IL-7 and the PPBSF cofactor
Antisera were raised in IL-7(−/−) mice against Ags isolated from IL-7(+/+) pro-B and pre-B CM by affinity purification with anti-IL-7 mAb. The ability of these antisera to neutralize PPBSF activity in pro-B CM was then tested against freshly harvested rat BM cells. The results in Figure 8 show that antiserum to the IL-7-associated Ag(s) in pro-B CM completely neutralized PPBSF activity. Therefore, we have termed this anti-PPBSF antiserum. However, the antisera to the IL-7-associated Ag in pre-B CM and to rIL-7 itself did not neutralize PPBSF activity, although both neutralized rIL-7 activity (data not shown).
Western blot analysis with the anti-PPBSF antiserum (Fig. 9 A) detected a single band of ∼55 kDa in IL-7(+/+) pro-B CM under nonreducing conditions (lane 2), and two bands of ∼25 kDa (asterisk) and 30 kDa (arrow) under reducing conditions. Only the 30-kDa band was observed in IL-7(−/−) pro-B CM (lane 4, arrows), and only the 25-kDa band was observed in IL-7(+/+) pre-B CM (lane 1, asterisks). Neither band was detected in IL-7(−/−) pre-B CM (lane 3). Hence, the 30-kDa molecule was selectively produced under pro-B cell culture conditions.
Adsorption of the anti-PPBSF antiserum with rIL-7 (Fig. 9 B) eliminated reaction with the 25-kDa band but not with the 30-kDa band (lanes 2 and 4, arrows). However, adsorption with rIL-7 did not alter the ability of this antiserum to neutralize PPBSF activity in IL-7(+/+) pro-B CM (data not shown). These results suggested that the 30-kDa molecule was the PPBSF cofactor. Evidence confirming this possibility was obtained by adsorbing IL-7(+/+) pro-B CM with anti-IL-7 mAb before Western blot analysis and demonstrating the absence of both the 30-kDa (PPBSF cofactor) and 55-kDa (PPBSF) bands as well as the 25-kDa (IL-7) band (data not shown).
The PPBSF cofactor is not TSLP, flt3 ligand, or SDF-1
We have previously demonstrated that neither rSCF nor rIGF-1 can substitute for the PPBSF cofactor in enabling IL-7 to induce proliferation of pre-pro-B cells in vitro (9). Here, we determined whether any of three other cloned cytokines/chemokines that synergize with IL-7 in regulating early B-lineage development were identical with the PPBSF cofactor. The results in Figure 10 show that neither TSLP (11, 12), flt3 ligand (13), nor pre-B cell stimulation factor/SDF-1β (14) is detected by Western blot analysis using antiserum to the PPBSF cofactor, and that the PPBSF cofactor is not detected by mAbs to TSLP, flt3 ligand, or SDF-1β. Furthermore, none of these cytokines/chemokines formed a heterodimer when mixed with rIL-7 (data not shown).
When combined with the observations in a companion paper (9), the present results appear to permit the following conclusions regarding PPBSF. Structurally, 1) PPBSF is a covalently linked heterodimer consisting of IL-7 and a cofactor with an Mr of ∼30,000; 2) PPBSF can form in solution by the spontaneous association of its two components; 3) the production of the two components of PPBSF by BM stromal cells is independently regulated; but 4) the formation and/or release of PPBSF under pro-B culture conditions is coordinated such that little if any of either component normally appears in monomeric form in the supernatant; and 5) the PPBSF cofactor, and hence, PPBSF, is neither formed nor released under pre-B cell culture conditions, thereby leaving only monomeric IL-7 in the supernatant. Although the nature of the covalent binding reaction is not known, dissociation of the PPBSF complex by DTT suggests that it occurs through a disulfide/sulfhydryl exchange mechanism similar to that observed in solution between platelet-derived growth factor and α2-macroglobulin (21). Functionally, 1) PPBSF stimulates the proliferation of pre-pro-B cells and some thymocytes but not pro-B cells or pre-B cells; 2) PPBSF primes pre-pro-B cells and their immediate descendants to proliferate in the presence of IL-7 alone; and 3) both IL-7 and the PPBSF cofactor maintain the viability of “unprimed” pre-pro-B cells, but neither induces their proliferation.
Having previously demonstrated that CM from pro-B-type cultures selectively stimulates the proliferation of pre-pro-B cells in vitro (8), the combined results of the present studies suggest that PPBSF, rather than IL-7 alone or in combination with SCF and/or IGF-1 (9), is the responsible agent. We also postulate that PPBSF normally induces the observed differentiation of pre-pro-B cells to pro-B cells (5, 6, 8, 22, 23). However, other cytokines, including monomeric IL-7 itself (9) and some as yet undefined SDFs, appear to amplify this process by inducing proliferation and/or differentiation of pro-B cells (2, 3, 11, 12, 13, 14, 24, 25, 26, 27, 28). Although none of these defined factors efficiently induces the development of pre-B cells in our culture system (our unpublished observations), such differentiation is rapidly induced in vitro by incubation of pro-B cells in IL-7(+/+) pre-B CM (29). Furthermore, surface IgM (sIgM)+ B cells appear in vivo within 2 wk of adoptive transfer of culture-generated pre-pro-B/pro-B cells to irradiated recipients (8; and our unpublished observations). Hence, additional factors not effectively represented in our culture system appear to be required for the differentiation of pro-B cells to pre-B cells (3, 14, 30, 31).
The observation that pro-B cells develop in IL-7 gene-deleted mice (10) would appear to challenge the postulated role of PPBSF in early B-lineage development. A more cautious interpretation, which we favor, is that PPBSF is the preferred ligand under physiologic conditions but that compensatory mechanisms for stimulating the proliferation and, at the very least, the phenotypic differentiation of pre-pro-B cells exist under nonphysiologic circumstances. Furthermore, such compensatory mechanisms may be incomplete, given that phenotypic differentiation to pro-B cells may occur in the absence of IgH gene rearrangements (32), and that pro-B cell proliferation is markedly reduced in IL-7(−/−) mice (C. Wei and I. Goldschneider, unpublished observations). It is important to emphasize, therefore, that the in vivo administration of anti-IL-7 mAb prevents the development of pro-B cells in normal mice (33), and that pro-B cells apparently fail to develop in IL-7R α-chain (−/−) mice (34). Inasmuch as the absence of IL-7 itself does not prevent pro-B cell formation (10), the former results suggest that anti-IL-7 mAb causes the coordinate elimination of IL-7 and an associated cofactor, thereby providing indirect evidence for the existence of PPBSF in vivo (also see 35 . In addition, the latter results suggest that the compensatory factor in IL-7(−/−) mice transduces a signal via the IL-7R. An intriguing candidate is a TSLP-like molecule (11, 12), whose function might be negatively affected in IL-7R α-chain (−/−) mice. Whatever its identity, the present results suggest that this compensatory factor is absent from IL-7(−/−) stromal cell CM.
Regarding the nature of the PPBSF cofactor, the most obvious candidate is the soluble form of the IL-7R (36), especially because other soluble ligand-receptor complexes in the hemopoietin family have been found to have enhanced functional activity over the ligand alone (37). Although not formally excluded, this possibility seems unlikely for several reasons. First, the form of the soluble IL-7R with the lowest molecular mass thus far described (38) is still significantly greater than that of the PPBSF cofactor. Second, the PPBSF cofactor appears to bind IL-7 covalently. Third, adsorption of CM with anti-IL-7R mAb does not remove the PPBSF cofactor (our unpublished observations). Nevertheless, it will be important to demonstrate that the PPBSF cofactor is produced by BM stromal cells from IL-7R α-chain (−/−) mice.
Inasmuch as IL-7 is avidly bound by heparin (39), it is possible that the PPBSF cofactor is a component of the stromal cell-associated extracellular matrix in our culture system (5, 6). Despite our inability to detect PPBSF activity in extracellular matrices (ECMs) extracted from BM-adherent cell layers with hypertonic saline (8), continued efforts are warranted based upon reports of the regulation of growth-factor signaling by ECM proteins (40, 41) and especially the description by Oritani and Kincade (42) of a series of ECM glycoproteins that selectively increase the IL-7-dependent proliferation of pre-B cells.
A number of stromal cell-derived cytokines/chemokines that act synergistically with IL-7 have been described, and, based on molecular mass and certain functional attributes, at least some might theoretically be candidates for the PPBSF cofactor. Therefore, Western blot analysis was conducted to determine whether any of the candidates bind to IL-7 and/or cross-react serologically with the PPBSF cofactor. The negative results obtained with TSLP (11, 12), flt3 ligand (13), and pre-B cell stimulation factor/SDF-1 (14) appear to exclude these factors, and previous mixing experiments with rIL-7 have similarly excluded SCF and IGF-1 (9). Other IL-7-synergizing factors are similarly being analyzed. However, studies of the primary amino acid sequence of affinity-purified PPBSF cofactor are most likely to reveal its identity.
We thank Drs. Richard Murray and Ursula von Freeden-Jeffry (DNAX Research Institute of Cellular and Molecular Biology) for providing the breeding stock of the IL-7 gene-deleted mice and Dr. Paul Kincade (Oklahoma Medical Research Foundation) for his insightful review of the manuscript. We also thank Mrs. Leigh Maher for expert technical assistance and Ms. Ruth Faasen and Ms. Cathy Mitchell for excellent secretarial assistance.
This work was supported in part by Grant No. AI-32752 from the National Institutes of Health.
Abbreviations used in this paper: BM, bone marrow; CM, conditioned medium; PPBSF, pre-pro-B cell growth-stimulating factor; TSLP, thymic stromal-derived lymphopoietin; SDF, stromal cell-derived factor; HSA, heat-stable Ag; BrdU, bromodeoxyuridine; TRITC, tetramethylrhodamine isothiocyanate; SCF, stem cell factor; IGF, insulin-like growth factor; TdT, terminal deoxynucleotidyl transferase; ECM, extracellular matrix; sIgM, surface IgM; HRP, horseradish peroxidase.