C a l c i f T i s s u e I n t (1990) 4 6 : 3 2 7 - 3 3 2
Calcified Tissue International 9 1990 Springer-Verlag New York Inc.
Effect of Cortisone on Cells at the Bone-Marrow Interface David J. Simmons, 1 Louis Kidder, 1 and Mary Thomas 2 University of Texas Medical Branch, ~Departmentof Surgery, DivisionOrthopedicSurgery; and 2Departmentof Pharmacologyand Toxicology,Galveston, Texas, USA
Summary. A study of the association between the rate of proliferation of marrow fibroblast-like stromal cells (in vitro) and the rate of endosteal bone mineralization (EsMR) (in vivo) was undertaken in an osteopenic rat model. We report that 200 g male rats treated with cortisone acetate (5 mg/day for 7 days) exhibit decreases in marrow fibroblast colony-forming units (FCFU) and tetracycline-based measurements of EsMR at the level of the femoral midshaft. In cortisone-treated rats recovering for I-3 weeks, the FCFU census and EsMR normalized during the first posttreatment week, remained at control levels after 2-3 weeks, and exhibited a relapse in the third week which signified only partial recovery. These changes were unrelated to patterns of body weight gain. The data indicate that the FCFU census can serve to index endosteal osteoblast vigor. Key words: Glucocorticoid - - Stromal cell - - Marrow - - Bone formation - - Osteopenia.
The ability of extraosseous marrow grafts to form bone is evidence that marrow contains a complement of mesenchymal osteogenic stem cells. Many recent clinical and laboratory studies have shown that the addition of viable autologous marrow cells enhances the osteogenic performance of bone bank grafts [1], accelerates the rate at which large segmental osseous defects heal [2-8], and promotes the repair of experimentally produced nonunions [6]. Such grafts have also been useful in treating bone defects in irradiated tissue sites that do not otherSend reprint requests to David J. Simmons, Ph.D., University of Texas Medical Branch, Department of Surgery, Division of Orthopedic Surgery, Galveston, TX 77550, USA.
wise heal [9-11]. The fibroblast-like stromal cells have been identified as the medullary osteogenic component. In situ, these cells are more concentrated near the endosteum than in the axial marrow [12-15]. They appear surficially as a condensation of epithelioid-like cells which cover the bone marrow, and extend as perivascular cells into the vascular channels of the bone cortex [16, 17]. We have explored the question of whether marrow stromal cells function in part to maintain populations of committed osteogenic cells at the marrow bone interface. The question presumes that stromal cells are somehow involved in osteoprogenitor-osteoblast interactions, and it suggests by extension that the failure of the mechanism might be responsible for age-related phenomena such as cortical involution, osteopenia, and/or osteoporosis. This thesis finds support in the association between involution and increased marrow fat. A deficit in bone formation and bone mass occurs coordinately with the conversion of stromal cells to adipocytes [18, 19] and a similar phenomenon can be precipitated by ovariectomy [20, 21], hypokinesia [22], and treatment with glucocorticoids [23-25]. Moreover, within a circadian context, the numbers of proliferative stromal cells in marrow are strongly associated with their ability to support osteogenesis in composite bone grafts [26]. There is some evidence that a stromal cell-derived cytokine drives this association [27-29], but the putative growth factor has been incompletely characterized [28]. Here we report that marrow stromal cell proliferation and the rate of endosteal bone formation in rats are inhibited coordinately by cortisone acetate, and that these processes r e c o v e r coordinately within a few weeks posttreatment.
Materials and Methods Young adult male rats (170-180 g body wt) were housed in
328 groups of 3-4. To establish baselines for the rate of endosteal bone mineralization (EsMR)d and marrow stromal cell proliferation, all animals were labeled with tetracycline (Achromycin, 10 mg/kg i.m.) on day 1 and 7 of the study, and marrow was taken for culture on day 8 (see below). All marrow biopsies were performed under general anesthesia (30 mg/kg sodium pentobarbital), and the animals were subsequently sacrificed by opening the chest cavity. The remaining animals were then randomly divided into control and steroid-treatment groups, and pair-fed Purina Lab chow, with tap water freely available. Body weights were recorded daily. During the second week of the study (day 8-15), the steroidtreatment group was injected with 5 mg/day cortisone acetate (Merck). The controls were injected i.m. with an equal volume of 0.9% saline (0.2 ml). All of these animals received a third tetracycline injection on day 13, and 5-6 rats from each group was sacrificed on day 16 to obtain bone and marrow samples. No steroid or saline injections were delivered during the succeeding 3 weeks, but all animals were injected with tetracycline on the 5th, 12th, and 19th days to obtain a fairly continuous record of the EsMR during this time period. Five animals from each group were sacrificed 48 hours after each fluorochrome injection to obtain bone and marrow samples. Plasma samples were collected by heart puncture and frozen for calcium and phosphorus determinations.
At Autopsy Using sterile conditions, the femurs and tibias were recovered. The right femur was fixed in 100% ethyl alcohol for bone growth studies. The other long bones were stripped of soft tissues and their epiphyses were cut away and used as a source of marrow.
D . J . Simmons et al.: Cortisone and Bone-Marrow
3s0
f
9 Control
o CortisoneAcetate5mg/di.m.
In 300
y
j
o
'
/~" "
"
-
250 ('~
s-
150
Pre-Rx 7 d
I
Rx 7 d
I
1 Wk
I
2 Wks
I
3 Wks
I
Post-Rx Recovery Fig. 1. Effect of cortisone acetate (5 mg/day i.m.) on body weight gain during the treatment and posttreatment recovery periods in male Sprague Dawley rats.
of the growing/mineralizing endosteal bone when each of the six tetracycline injections were administered. Pairs of fluorescent markers, then, identified the bone deposited/mineralized during the baseline, injection, and postinjection recovery periods. To obtain the EsMR for each timepoint, we measured the distances between band pairs on the anterior, posterior, lateral, and medial endosteal surfaces, and the averages were divided by the number of days between injection(s). The data were expressed in terms of micrometer growth/day.
Plasma Determinations. Plasma calcium was determined using atomic absorption spectrophotometry following dilution of samples in lanthanum-HC1 to minimize interference by serum phosphates. Plasma phosphorus was measured colorimetrically using reagents obtained from Stanbio (San Antonio, TX).
Marrow Cultures. At each sacrifice period, femoral (left) and tibial marrow from each group was expressed into a-MEM medium and pooled. The cells were disaggregated by passing them through a fine steel mesh, resuspended in a-MEM supplemented with 15% fetal calf serum (FCS; Hyclone Laboratories, Logan UT), and a 1% penicillin-streptomycin-fungizone mixture (Whittaker M.A. Bioproducts, Walkersville, MD). Erythrocytes were lysed in Tris-ammonium chloride. The cells were washed in fresh medium and counted in a hemocytometer using the Trypan blue exclusion technique to identify damaged elements. Viable cells were plated-out in 25 cm a T-flasks at a concentration of 10 7 cells/ flask and grown in a humidified tissue culture incubator at 37~ under an atmosphere of 5% CO2. The medium was first changed after 24 hours and thereafter at intervals of 2-3 days. There were no intergroup differences in the recovery of nonadherent cells (data not shown). After 10 days, the cultures were fixed in absolute methanol, stained with Giemsa, and visualized under a dissecting microscope equipped with an ocular micrometer. Two indices of stromal cell proliferation were used: (1) the number of fibroblast colonies (FCFU) with 50 or more cells; (2) the average size in mm of a random sample of 100 FCFUs.
EsMR. The alcohol-fixed femurs were defatted in acetone, embedded undecalcified in methylmethacrylate, and sectioned transversely (90-100 p.m) at the level of the midshaft with an Isomet low-speed rotary saw, using a diamond blade. The sections were visualized by UV microscopy to identify the positions
Statistics The significance of intergroup differences was calculated with the unpaired Student's t test when the variances were equal. When the variances were unequal, we used a nonparametric Ftest. P values < 0.05 were considered statistically significant.
Results
Body Weight Cortisone treatment inhibited body weight gain. The greatest disparity between the control and steroid-treated groups occurred during the injection p e r i o d ( F i g . 1). D u r i n g t h e 3 w e e k r e c o v e r y p e r i o d , the growth curve for the treated animals paralleled that of their controls, but they never normalized in terms of absolute weight. Similar changes were recorded in a confirmatory study in rats, which were approximately 220 g at the time of steroid treatment; their weights normalized within 2-3 weeks posttreatment.
329
D. J. S i m m o n s et al.: Cortisone and Bone-Marrow
Table 1. Effect of cortisone acetate on plasma calcium and phosphorus in rats
T a b l e 2. E f f e c t o f c o r t i s o n e a c e t a t e o n f e m u r l e n g t h in rats (SEM)
Experimental period
Experimental period
Group
N
F e m u r length (cm)
Treatment"
Control Cortisone Control Cortisone Control Cortisone Control Cortisone
5 5 5 4 3 3 3 3
3.216 3.132 3.266 3.183 3.370 3.290 3.390 3.550
Treatment" 1 week recovery 2 week recovery 3 week recovery
Group
N
Calcium (mg/dl)
Phosphorus (mg/dl)
Control Cortisone Control Cortisone Control Cortisone Control Cortisone
5 5 5 5 5 5 3 3
9.96 9.42 9.22 9.92 9.97 9.94 10.19 9.52
8.81 10.41 8.87 8.64 9.54 10.09 9.28 8.33
--- 0.14 -+ 0.22 -+ 0.38 -+ 0.19 -+ 0.41 - 0.59 -+ 0.20 -+ 0.23
-+ 0.25 b - 0.41 -+ 0.38 - 0.44 - 0.27 -+ 0.64 -+ 0.24 b +- 0.07
1 w e e k recovery 2 week recovery 3 w e e k recovery
-+ 0.018 - 0.021 b -+ 0.037 -+ 0.028 +- 0.078 + 0.020 -+ 0.062 -+ 0.012
5 mg/day i.m. for 8 days b p < 0.02 a
" 5 mg/day for 8 days b p < 0.02 vs. control
NS 100
Serum Calcium and Phosphorus Cortisone-treated rats were normocalcemic at all experimental intervals. They displayed significant hyperphosphatemia at the end of the treatment period, and were mildly hypophosphatemic at the end of the third recovery week (Table 1). The stability of calcemia suggests that the shifts in phosphatemia were more reflective of tissue breakdown patterns than changes in renal function or parathyroid-bone physiology [40].
Femur Length Cortisone treatment caused a slight but statistically significant decrease in linear bone growth (P < 0.02), but this difference was not evident a week thereafter or at any subsequent recovery period (Table 2).
Marrow FCFU Census The FCFU census in marrow cultures from salineinjected control rats increased over baseline (see comparable data for endosteal bone mineralization below). This failed to occur in the cortisone-treated rats, and their FCFU census was 50% lower than control values at the end of the treatment period. During the posttreatment recovery period, the colony counts from the steroid-treated groups normalized within 7 days, remained at control levels during the second week, but declined at the end of the third week (Figs. 2, 4). The confirmatory experiment in somewhat older rats (220 g rats, Table 3) produced normal colony counts at the end of the third and sixth posttreatment recovery weeks, indicating that the steroid dosage schedule could result in either partial or total recovery of FCFUs. There were no
/~'--..
90
NS
80 (n
70
u.
60
E
50
u. (J u.
40 30 20
/_/' /.---"I
if...'-
10 0 I i Baseline
J.
Rx 7 d
",,,; 9 conro,
0 Cortisone Acetate 5mg/d I.m.
1 Wk
2 Wks
3 Wks
Post-Rx Recovery Fig. 2. Effectof cortisone acetate (5 mg/dayi.m.) on the clonal growth of cultured fibroblast-like marrow cells (FCFU) during the treatmentand posttreatmentrecoveryperiods.
intergroup differences in the size distribution of FCFUs (data not shown). Endosteal Bone Mineralization Rate Daily saline injections produced a slight increase in EsMR to accompany the change in the FCFU census. Eight days of cortisone treatment decreased the EsMR by 25-30%, but recovery was rapid and the rates normalized within the first week. During the 3 week recovery period, the EsMR values for the control and cortisone-Rx rats were not statistically different (1-2 ~rn/day), but the mean values in the controls during the second and third week tended to be higher. This difference were also observed in the FCFU census in the third week (Figs. 3, 4; Table 3). Overall, the in vivo EsMR data were reflective of the in vitro stromal cell proliferation assay (see confirmatory 3 week recovery study above). Discussion
The steroid injection protocol was designed to
330
D . J . Simmons et al.: Cortisone and Bone-Marrow
Table 3. F e m u r length, the endosteal mineralization rate, and marrow stromal cell clonal growth (FCFU) in rats 21 and 52 days following an 8-day course of cortisone injection a Experimental
F C F U m a r r o w census
period
Group
N
F e m u r length 21 days (cm)
EsMR (21 days) (t~m/day)
21 days
52 days
Recovery
Control Cortisone
4 4
3.550 -4- 0.039 3.386 • 0.046
0.625 • 0.625 0.625 • 0.625
47.5 --- 23.0 33.0 • 10.6
25.33 • 5.70 22.75 • 3.47
Average body wt 212 g a Injection--5 mg/day i.m.
"o E
8.0
<.02
7.0 6.0
9 Control 0 Cortisone Acetate 5 mg/d I.m.
_~______..~,,~
2 s.0 9
4.0
~
3.0
m 2.0 om "w
1.0 0
i
i
L
i
i
Baseline
Rx 7 d
1 Wk
2 Wks
3 Wks
Post-Rx Recovery Fig. 3. Effect of cortisone acetate (5 mg/day i.m.) on the midfemoral endosteal mineralization rate during the treatment and posttreatment recovery periods.
.~" . . . . .
5.Or
4.0t
,o._ __~;~,.-_o- . . . . .
~O) 3.0 "I2.~0
+| |
,_0~
sss
jJ
s'ss~ ~''~
-~
::=6,~..
s7
"-"~" - ~a I Marrow
~z~s/
~
N
90
~ \
Control
9 ~ co.,..-..
0.6 0.4
_
9/ 0/ /
FCFU's
~.g,,,.
~.....,..,.,..,.,.~ El~lo~l~l G. . . . .
~
Baseline
, Rx 7 d
, 1 Wk
, 2 Wks
3 Wks
POst-Rx R e c o v e r y
Fig. 4. Relative changes in the marrow F C F U census and the endosteal bone mineralization rate in rats treated with 5 mg cortisone acetate i.m. during the treatment and posttreatment
recovery periods9
achieve impairment of linear and radial bone growth in rats [24, 25, 30], and to ensure a posttreatment recovery time that was long enough for animals to sustain a period of catch-up growth [30]. Those aims were accomplished. The model served as a vehicle to test the validity of a proposition that the in vitro proliferation of marrow fibroblast-like stromal cells (FCFU) could be used to indirectly assess the level of endosteal bone formation (EsMR). As noted above, we developed this hypothesis on the basis of experiments in rats that developed osteopenia following ovariectomy [20] and short-term tail
suspension (hypokinesia [31]). Because in both situations the progressive thinning of the femur is associated with a decrease in stromal cell proliferation and because some of those cells are osteoprogenitor elements, the thesis provided an insight about the mechanism(s) responsible for the cortical involution and osteopenia/osteoporosis that accompanies aging and some metabolic bone diseases. The association also persisted in trials to reverse the osteopenic trend. In the ovariectomized animal, the deficit in the FCFU census and EsMR could be corrected by treating the animals with an analogue of 1,25 dihydroxyvitamin D (dihydrotachysterol) alone or in combination with beta-estradiol. Identical responses to dihydrotachysterol were observed in ovariectomized rats implanted intramuscularly with osteoinductive composite grafts of allogeneic bone and autologous marrow. The hypokinetic rat presents a less coherent model because the EsMr values alone spontaneously normalize within the second week of tail suspension [31]. Yet, the present study in cortisone-treated rats, and Owen's recent study in rabbits [32], points to the validity of the association between the EsMr and marrow stromal cell proliferation. It should be noted that in the present study, the relationship was less than perfect. There was a delay in the recovery of baseline FCFU scores relative to the EsMR, but this appeared to be due to the relatively larger amplitude of the changes in the stromal cell population. The differences in lag-to-recovery does not, however, contraindicate a relationship between stromal cell proliferation and osteoblast performance. The question remains as to whether this association has any real functional meaning for the normal maintenance of active osteogenic cell populations at the bone-marrow interface. This is a consideration apart f r o m B u r w e l l ' s [1] original and oftenconfirmed report that adjunctive autologous marrow improves the survival and osteogenic performance of banked bone grafts. In all aging vertebrate species, the marrow cavities do not fill with bone; rather, there is a pattern of increased tissue fat due to the conversion of the fibroblast-like stromal cells to adipocytes, and a corresponding decline in bone growth [34]. In situ, some "pathologic process"
D. J. Simmons et al.: Cortisone and Bone-Marrow
such as fracture or radiation appears requisite for endosteal marrow stromal cells to express their osteogenic potential [33]. In terms of our findings, the real issue is whether there is any evidence to support a concept that there are stromal cell-osteoprogenitor-osteoblast interactions that do not involve the direct modulation of stromal cells to osteoblasts. Gazit et al.'s [27] and our own study [29] show that stromal cells produce one or more as yet unidentified growth factors that stimulate the proliferation of neonatal calvarialderived rat osteoblasts and chondrocytes, but not transformed ROS 17/2 or 17/2.8 cells. In the absence of stimuli such as endotoxin and irradiation [35-38], stromal cell conditioned media appears to contain little fibroblast mitostatic IL1/IL-3, and a 35KDa fraction isolated by filtration on Sephadex G75 was apparently without PDGF-Iike activity [28]. The activities of two other fractions [27] with molecular weights of <10,000 and 18,000 Da are unknown, perhaps representing mitogenic insulinlike growth factor (7500 Da) and monokines [39]. Thus, the agewise decline in the level of bone formation and the onset of osteoporosis probably reflects differing degress of loss/impairment of marrow stromal cell cytokine production. The alternate proposition may be equally feasible--that osteoprogenitor cells and osteoblasts produce a stromal cell mitogen, but this question has never been put to the test. The precise relationship between the FCFU component of the marrow stroma and endosteal osteogenesis is unknown, but the communality of experience with the clonal stromal cell assay in osteopenic animal models indicates that the system should be equally applicable to situations in which osteogenesis is accelerated.
Acknowledgment. This study was supported by a UTMB Intramural grant.
References 1. Burwell RG (1966) Studies on the transplantation of bone. VII. The fresh composite homograft-autograft of cancellous bone. An analysis of factors leading to osteogenesis in marrow transplants and in marrow-containing bone grafts. J Bone Jt Surg 46-B:110-140 2. Boyne PJ (1973) Methods of osseous reconstruction of the mandible following surgical resection. J Biomed Mat Res Symp No 4:195-204 3. Salaama R, Burwell RG, Dickson IR (1973) Recombined grafts of bone and marrow. J Bone Jt Surg 55-B:402-417 4. Salama R, Weissman SL (1978) The clinical use of combined xenografts of bone and autologous red marrow. J Bone Jt Surg 60-B: 111-115
331 5. Wittbjer J, Palmer B, Rohlin M, Thorngren K-G (1983) Osteogenic activity in composite grafts of demineralized compact bone and marrow. Clin Orthop 173:229-238 6. Takagi K, Urist MR (1981) The role of bone marrow in bone morphogenetic protein-induced repair of femoral massive diaphyseal defects. Clin Orthop 171:224-231 7. Plenk H, Hollmann K, Wilfert K-H (1972) Experimental bridging of osseous defects in rats by the implantation of Kiel bone containing fresh autologous marrow. J Bone Jt Surg 54-B:735-743 8. Leake D, Habal M, Maniscalco J, Freeman S (1978) Neurocranial reconstruction using an elastomer-coated cloth mesh and bone grafting. J Biomed Mat Res 12:83-87 9. Marciani RD, Gonty AA, Synhorst JB, Page LR (1979) Cancellous bone marrow grafts in irradiated dog and monkey mandibles. Oral Surg 47:17-24 10. Marciani RD, Gonty AA, Giansanti JS, Avila J (1977) Autogenous cancellous-marrow bone grafts in irradiated dog mandibles. Oral Surg 43:365-367 11. Marciani RD, Trodahl JN (1976) Cancellous-marrow bonegrafts in irradiated tissue. Oral Surg 42:431-441 12. Ashton B A, Allen TD, Howlett CR, Eaglesom CC, Hattori A, Owen M (1980) Formation of bone and cartilage by marrow stromal cells in diffusion chambers in vivo. Clin Orthop 151:294-307 13. Bab I, Ashton BA, Syftestad GT, Owen ME (1984) Assessment of an in vivo diffusion chamber method as a quantitative assay for osteogenesis. Calcif Tissue Int 36:77-82 14. Yang EO, Rooney P, Wright EG (1982) The radial distribution of fibroblastic colony-forming cells in mouse femoral marrow. IRCS Med Sci 122:1141-1142 15. Xu CX, Hendry JH (1981) The radial distribution of fibroblastic colony-forming units in mouse femoral marrow. Biomedicine 35:119-122 16. Menton DN, Simmons DJ, Orr BV, Plurad SB (1982) A cellular investment of bone marrow. Anat Rec 203:157-164 17. Gupta V, Rajaraman S, Costanzi JK (1987) Effect of oxygen on the clonal growth of adherent cells (CFU-F) from different compartments of mouse bone marrow. Exp Hematol 15:1153-1157 18. Castro-Malespina R, Gay RE, Resnick G, Napoor N, Meyers P, Chiarieri D, McKenzie S, Broxmeyer HE, Moore MAS (1980) Characteristics of human bone marrow colonyforming cells (CFU-F) and their progeny. Blood 56:289-301 19. Kaneko S, Motomura S, Ibayashi H (1982) Differentiation of human bone marrow-derived fibroblastoid colony-forming cells (CFU-F) and their roles in haemopoiesis. Br J Haematol 51:217-225 20. Tabuchi C, Simmons DJ, Fausto A, Avioli LV, Binderman I (1986) The bone deficit in ovariectomized rats: the functional contribution of the marrow stromal cell population. J Clin Invest 78:637-642 21. Wronski TJ, Morey ER (1982) Skeletal abnormalities in rats induced by simulated weightlessness. Metab Bone Dis Rel Res 4:69-75 22. Jee WSS, Wronski TJ, Morey ER, Kimmel DB (1983) Effects of spaceflight on trabecular bone in rats. Am J Physiol 244:R310-R314 23. Jee WSS, Park HZ, Roberts WE, Blackwood EL, Kenner GH (1970) Corticosteroids and bone. Am J Anat 129:477-480 24. Storey E (1963) The influence of adrenal cortical hormones on bone formation and resorption. Clin Orthop 30:197-216 25. Simmons DJ, Kunin AS (1967) Autoradiographic and bio-
332
26.
27. 28.
29.
30.
31.
32.
D . J . Simmons et al.: Cortisone and Bone-Marrow chemical investigations on the effect of cortisone on the bones of the rat. Clin Orthop 55:201-215 Simmons DJ, Loeffelman K, Frier C, McCoy R, Friedman B, Melville S, Kahn AJ (1984) Circadian changes in osteogenic competence of marrow stromal cells. In: Haus E, Kabat HF (eds) Chronobiology 1982/1983. Karger, Basel, pp 37-42 Gazit D, Shteyer A, Bab 1 (1986) Marrow-derived factors regulating bone cell proliferation and activity. Bone 7:389 Gazit D, Shteyer A, Muhlrad A, Bab I (1988) Partial purification and biological activity of an osteogenic growth factor derived from healing bone marrow. Calcif Tissue Int (suppl) 42:A30 Simmons DJ, Seitz P, McPherson MB, Tabuchi C, Rifas L, Klein GL, Gundberg CM, Kidder LS, Waeltz M, Yang CY, Lozano R (1989) Partial characterization of rat marrow stromal cells (abstract). XXI Eur Symp on Calcified Tissues, Jerusalem, Israel, Calcif Tissue Int (suppl) 44:S-49 (Abstr #H9) Simmons DJ (1976) Comparative physiology of bone. In: Bourne GH (ed) The biochemistry and physiology of bone. Academic Press, New York, pp 445-516 Kidder LS, Simmons DJ (in press) Endosteal bone formation in hypokinetic osteopenia. Am J Phys Anthrop 78:252 (Abstr) Owen ME, Cave J, Joyner CJ (1987) Clonal analysis in vitro of osteogenic differentiation of marrow CFU-F. J Cell Sci 87:731-738
33. Simmons DJ (1985) Fracture healing perspectives. Clin Orthop 200:100-113 34. Simmons DJ, Grynpas MD (in press) Mechanisms of bone formation in vivo: the osteoblast. In: Hall BK (ed) Bone, vol. 1. The Telford Press, Caldwell, NJ 35. Felix R, Elford PR, Stoerckle C, Cecchini M, Wetterwald A, Trechsel U, Fleisch H, Stadler BA (1988) Production of hemopoietic growth factors by bone tissue and bone cells in culture. J Bone Min Res 3:27-36 36. Bagby GC, Rigas VD, Bennett RM, Vanderbark AA, Garewol HS (1981) Interaction of lactoferrin monocytes and lymphocyte T-subsets in the regulation of steady state granulopoiesis in vitro. J Clin Invest 68:56--63 37. Cline MJ, Golde DW (1974) Production of colonystimulating activity by human lymphocytes. Nature 248:703704 38. Ruscetti FW, Chersenick PA (1975) Regulation of the release of colony-stimulating activity from mitogen-stimulated lymphocytes. J Immunol 114:1513-1517 39. Centrella M, Canalis E (1985) Local regulators of skeletal growth: a perspective. Endocrine Rev 6:544-551 40. Sorensen BM (1966) Influence of glucocorticoids and anabolic steroid hormones on the wound-healing process. In: Asboe-Hansen G (ed) Hormones and connective tissue. Munksgaard, pp 106-123 Received November 28, 1988, and in revised form February 2, 1989.