THE ANATOMICAL RECORD 216:528-533 (1986) Strategies of Hemopoietic Stress Adaptation Within the Medullary Cavity DENNIS C. GIULIANI, JAMES C. HALL, AND BERNARD S. MORSE Department of Medicine-Hematology Division, Monte-Scaglione Memorial Research Laboratory, New Jersey Medical School, University of Medicine and Dentistry at New Jersey, Newark, N J 07103 (D.C.G.,B.S. M.); Department of Zoology and Physiology, Rutgers University, Newark, N J 07102 (J.C.H.) ABSTRACT Gravimetric determination of total bone water space was used as a n index of available bone marrow space in mice following various specific stressors, i.e., splenectomy, hypoxia, bone fracture, and estrone-induced osteosclerosis. Data was corrected for bone weight and was reported as specific bone marrow volume (total bone water spacetmg dry bone x 100). A direct relationship was observed between specific bone marrow volume and medullary hemopoietic activity induced by stress. Absolute andior relative marrow space increased following splenectomy, hypoxia, and fracture. Osteosclerotic animals shift most hemopoietic activity from marrow to spleen, and splenectomized osteosclerotic animals become anemic. Both intact and splenectomized hypoxic animals develop increased specific bone marrow volume and successfully compensate for hypoxia with enhanced erythropoiesis. Animals sustaining a fracture callus increase both specific bone marrow volume and hemopoietic activity a t the callus without an increase in hemopoietic demand. Increased specific bone marrow volume extends the marrow bone interface, where primitive stem cells accumulate, while expanding marrow stromal space, where stem cells lodge, proliferate and differentiate. Therefore, it would appear that availability of competent marrow space may play a n integral part in passively permitting hematopoiesis and in determining hemopoietic reserve capacity. Stem cell migration increases during intensified hemopoietic demand, which also may be related to available marrow space. Mice have a low medullary hemopoietic reserve capacity; subsequently, when available medullary hematopoietic stroma becomes occupied, stem cells are more likely to migrate from the marrow to extramedullary sites where they mature before entering the circulating pool. Adaptation of hemopoietic tissue to stress has received considerable attention in the literature. Yoffey (1966) suggested strategies by which the hematopoietic tissue may adapt in order to support the erythroid hyperplasia following hypoxic hypoxia. He proposed: 1) accelerated stem cell maturation, 2) increased numbers of mitoses during maturation (amplification), 3) increased numbers of stem cells (proliferation), and 4)expansion of hematopoietic reserve space. Current thought on these adaptive strategies Worse et al., 1969) indicates that progenitor cells may shorten their cell cycle, primarily by reducing the time in G1 phase. Questions regarding the number of mitoses preceding final maturation remain unresolved, and although stem cells reportedly proliferate in order to support increased demand, the mechanism for maintaining the differentiated pool (Bessis and Brecher, 1975)remains uncertain. Finally, hematopoietic reserve space may be gained at the expense of scattered fat cells within the predominantly active (red) marrow in those species with greater hemopoietic reserve, and available marrow space may be expanded through bone resorption, as reported by Yoffey (1966). 0 1986 ALAN R. LISS, INC. Independent studies in dogs by Drinker et al. (19221, Johnstone (1922), and Gong (1965) present data that support a direct relationship between hemopoietic demand and available hematopoietic tissue space within the medullary cavity. In contrast, Hudson (1958a,b) reported no change in either total or red marrow space of guinea pigs exposed to hypoxia. Increased spleen weight in the hypoxic guinea pigs indicates extramedullary hemopoietic adaptation in this species, which has a low hemopoietic marrow reserve capacity according to Yoffey (1966). In this study, hemopoietic adaptation was studied in mice, a species with low medullary hemopoietic reserve, as they responded to various specific stresses. Both normal and splenectomized animals were exposed to the following: hypoxic hypoxia, estrone-induced osteosclerosis, and fracture repair. Hematologic parameters as well as dry bone weight, bone mineral volume, and total Received February 3, 1986; accepted July 15, 1986. HEMOPOIETIC STRESS AND BONE MARROW VOLUME 529 bone water space were measured. Strategies of adapta- bone water space were made by projecting and tracing tion to these stressors are discussed. cross- and longitudinal sections of fully calcified bone (Carl Ziess projection microscope). For this study, bones MATERIALS AND METHODS previously digested in pronase were embedded in clear Animals casting resin and allowed to harden under vacuum. MiFemale 8-week-old, 20-g CF1 outbred albino mice croscopic sections were made with the aid of Isomet saw (Charles River Farms) were used exclusively in these and Ecomet I1 grinder-polisher. Each cross-sectional cut experiments. was 0.10 inch thick. Area and length measurement were Animals were randomly divided into four groups: con- made from the tracings by means of a computer aided trol, experimental, splenectomized control, and splenec- planimeter. Absolute medullary cavity volume and bone tomized experimental. Spleens were surgically removed volume were calculated from the tracings by projecting from the animals so designated, and individual experi- a stage micrometer along with the bone sections as ments were started 2 weeks after splenectomy. modified from reports by Green et al. (1981).Differences between medullary cavity volume and total water space Peripheral Blood Parameters in the whole bone were taken to represent bone vascular Mice were exsanguinated via cardiac puncture under and lacunae space as well as water of hydration. light ether anesthesia. Whole blood, anticoagulated with Bone Mineral Density disodium ethylenediamine tetraacetic acid (EDTA), was used to measure hematocrit and hemoglobin concentraBone mineral density was obtained by employing Artion; peripheral blood smears (Wright’s-Giemsa stain) chimedes’ Principle. Mineral volume was measured in were prepared according to standard methodology re- pronase-digested, defatted bone. Bones were first satuported by Wintrobe (1972). rated with water in vacuo then weighed in a 1-ml pycnometer in order to determine the weight of water Bone Marrow Volume Measurement displaced by the bone mineral. Weight of displaced water Animals were weighed to the nearest gram, exsan- was converted to volume (dividing by the density of guinated, then killed by cervical dislocation while under water at laboratory temperature). Density (dry bone ether anesthesia. Intact bones (tibias and femurs) were weightlunit volume of bone) reflects contributions of both dissected free from the soft tissue. Bones were then the bone mineral and organic matrix. incubated in vacuo at 37°C in 0.01 M Tris buffer (pH Fracture and Fracture Repair 7.2) containing 5 mglml pronase in order to digest the marrow. Incubation was continued for 2 weeks, during Midshaft fracture of the left tibia was performed by which the buffered pronase solution was changed every digital manipulation under light ether anesthesia. Aniother day. Following pronase digestion, fat was ex- mals were sacrificed 23 days later. Both fractured and tracted from the bone by refluxing in a mixture of 4:l contralateral unfractured tibias were processed a s preethyl etherabsolute alcohol for 1 hour at room temper- viously described. ature. The remaining soluble substance was removed by three changes of deionized water in vacuo, with one Estrone change per day. Experimental osteosclerosis was induced by weekly Microscopic examination of representative bones after subcutaneous injections of a n aqueous estrone suspendigestion showed them to be completely devoid of all marrow tissue. These bones were then dried under vac- sion (0.5 mg) for 6 weeks (Simmons, 1966). uum a t 80°C for 6 hours, allowed to reach room temperHypoxic Hypoxia ature in a dessicated chamber, and weighed to the Hypoxia was induced by housing mice in cages lined nearest 0.0001 g. Intact bones were then placed in deionized, distilled water under vacuum for 24 hours. Upon with a silicone membrane enclosure, which selectively removal from the water, bone surfaces were wiped dry, reduces oxygen permeability. Maintaining seven mice and the bone was immediately weighed to the nearest per cage has been found (Lang et al., 1966) to lower the 0.0001 g. Both dry and water-saturated weights were oxygen concentration to 9% which is equivalent to a n altitude of 21,000 feet. Cages were changed twice weekly repeated until constant values were achieved. Marrow volume (total bone water space) was deter- during the 23-day course of the experiment, and animals mined by subtracting the dried bone weight from the unable to sustain hematocrits of 60% or greater were water-saturated weight. Differences in these weights eliminated from the experiment. represented the total weight of water within the bone. Data Analysis This value was converted to total volume within the bone by dividing by the density of water at laboratory Data was analyzed by means of standard statistical temperature. These volume measurements include bone methods. Sample size with mean values are reported lacunae space, vascular space, water of hydration, and k standard error. One-way analysis of variance was used available marrow space. Bone marrow volume measure- to analyze for differences between means. If means were ments were repeated using selected bones that had been significantly different by analysis of variance, (P < .05), cleaved longitudinally in order to remove medullary Duncan’s multiple range test was used to determine cavity water. This made it possible to determine the significant differences between all possible pairs of group contribution of nonmedullary cavity water space to the means. Animals that sustained a fractured tibia were total bone water space. statistically analyzed using the two-tailed paired t-test. Morphometric studies designed to confirm the contri- Comparison was made between the fractured tibia and bution of nonmedullary cavity water space to the total the contralateral unfractured tibia. D.C. GIULIANI, J.C. HALL, AND B.S. MORSE 530 TABLE 1. Cumulative data' Body weight (gm) Mean No. f SE Hematocrit 9'0 Mean No. f SE Hemoglobin conc. (gram %I Mean No. f SE No. Control 22 22 Hypoxic 14 Hypoxic splenectomy 15 13.81 f 0.41 14.4 f 0.73 14.3 f 0.38 27.11" f 0.88 25.35* f 0.67 22 13 45.7 .-t 0.54 . 45.5 -t . - 0.76 47.1 -t 0.67 78.0" .- 1.32 -t 76.5* -t . - 0.69 22 Control splenectomy Shamoperated 28.6' f 0.71 27.8 f 1.34 28.7 1.21 16.8" f 0.38 15.9* f 0.32 29.3 f 0.44 28.6 f 0.55 9 42.8" .- 1.09 -t 33.1" .- 1.44 -t 8 12.0" f 0.07 9.77" f 0.61 8 Estrone Estrone splenectomy 9 9 14 * Body weight Mean No. f SE 11 9 9 15 14 Hematocrit _____ % No. Mean k SE 11 9 9 15 14 Hemoglobin conc. (gram %) Mean No. f SE 12 7 13 13 14 No. Dry bone weight (mg) Tibia No. Femur 30.3 f 0.93 27.8 f 1.36 30.2 f 1.07 22.3" f 0.52 21.0" f 0.67 22 43.8" f 1.37 43.2" f 0.86 8 13 9 14 14 14 Bone marrow volume (total bone water space) (mm3) No. Tibia No. Femur 37.0 f 1.10 33.3 2.00 33.9 f 1.71 22.8* k 0.51 21.9" It 0.60 21 54.7" f 1.53 51.1" 1.23 8 * Dry bone weight (mg) Tibia No. Tibia2 11 7 11 12 8 11.8 f 0.27 11.8 f 0.38 11.5 k 0.53 11.3 f 0.35 10.3" f 0.51 22 8.5" f 0.2 8.1" k 0.17 8 13 9 14 13 14 16.6 f 0.46 16.7 f 0.51 16.1 f 0.64 14.6" f 0.39 14.7" f 0.62 11.8* k 0.16 12.2* f 0.25 Bone marrow volume (total bone water space) (mm3) No. Tibia No. Tibia2 Fracture (tibia only) 10 28.4' k 0.72 10 4-6.1 ._ -t 0.80 9 13.42 f 0.64 9 45.8" f 1.17 9 30.7 f 0.45 9 26.6" f 1.51 9 12.0 + 0.39 - Fracture (tibia only) splenectomy 10 27.7 f 0.58 10 4.4.8 ._ -t 0.55 10 12.83 f 0.23 10 42.7" f 1.76 10 30.1 f 0.84 8 25.6" k 1.37 8 12.25 f 0.36 (continued) RESULTS Hematologic and bone and marrow volume parameters are listed in Table 1. Splenectomized animals respond with a small decrease in bone mineral volume (tibia and femur) and increased specific bone marrow volume (femur only). Other parameters did not vary significantly from those of control animals. Sham-operated splenectomized animals showed no variation in any parameter when compared to controls. Mice exposed to atmospheric hypoxia showed significant variation from control animal values. Body weight, dry bone weight, bone mineral volume, and bone marrow volume (total bone water space) decreased. Hematocrit, hemoglobin concentration, and specific bone marrow volume (total bone water space/mg bone x 100) all increased versus control values. Only tibia1 bone marrow volume remained unaltered in nonsplenectomized animals exposed to hypoxia. Animals subjected to bone fracture and fracture repair were studied in order to assess their influence on the bone and marrow. Peripheral blood parameters remained unchanged. Bone and bone marrow parameters in the contralateral unfractured tibia also did not vary from that of control. All bone and marrow parameters in the fractured tibia, both measured and derived, increased-with the exception of bone density, which remained unchanged. Splenectomized animals responded similarly to fracture. Estrone-treated animals show a slight but significant reduction in hematocrit and hemoglobin concentration. Bone and marrow parameters, on the other hand, all change in response to estrone. Dry bone weight increases, total bone water space decreases, and bone mineral volume increases. Derived parameters were also altered; bone density increases, and specific marrow volume decreases. Splenectomized animals treated with estrone respond with a pronounced reduction in hemoglobin concentration and hematocrit. Both normal and splenectomized estrone-treated animals had similar changes in their bone and marrow parameters. Conformational studies of the displacement technique are shown in Table 2. Both morphometric measurement (projection microscopy)and displacement technique (longitudinally cleaved bones) are in close agreement and confirm the contribution of medullary cavity water space to total bone water space. Increased specific marrow volume in hypoxic animals indicates a redistribution of bone water space, with a larger portion of this water distributing in the marrow cavity. DISCUSSION Enigmas associated with the stem cell concept of hematopoiesis reported by Cronkite (1964) remain unresolved. For example, whether individual stem cells multiply sequentially as they mature into a specific functional pool or have continuous access among the various stem cell compartments is not clear (Mackey 531 HEMOPOIETIC STRESS AND BONE MARROW VOLUME TABLE 1. (Continued) Bone mineral volume (mm3) No. Tibia No. Femur No. Tibia No. Femur Control 21 21 22 Control splenectomy Shamoperated 12 Hypoxic 13 Hypoxic splenectomy Estrone 12 Estrone splenectomy 14 2.30 f 0.026 2.33 f 0.031 2.28 0.063 2.43 0.069 2.43 k 0.069 2.51* & 0.028 2.55" f 0.032 2.32 f 0.027 2.38 f 0.044 2.28 f 0.054 2.43 k 0.068 2.47 k 0.063 2.53* f 0.023 2.48* f 0.017 7 8 13.4 f 0.29 12.0" f 0.45 13.4 f 0.51 9.2" f 0.40 8.9" f 0.45 17.5" f 0.62 17.1* ir 0.36 22 13 9 14 14 8 14 16.0 f 0.49 14.0* & 0.83 14.5 0.91 9.4" f 0.43 9.0* f 0.30 21.5" & 0.71 20.6* k 0.46 * Bone density 12 7 13 12 8 14 Bone mineral volume (mm3) Fracture (tibia only) Fracture (tibia only) splenectomy Tibia No. Tibia2 No. Tibia 9 19.2 f 0.54 19.6* f 0.76 9 13.1 f 0.39 13.0 f 0.45 8 2.34 0.017 2.30 f 0.024 8 9 13 14 8 14 Bone density No. 8 13 9 No. 8 9 Tibia' 2.33 f 0.05 2.40 f 0.039 Specific bone marrow volume No. Tibia No. Femur 21 11 7 11 12 8 14 38.9 f 1.48 42.5 & 1.58 37.5 f 1.01 49.1* f 1.58 48.8* f 1.96 19.5" f 0.39 18.7* f 0.33 22 13 9 14 13 8 14 45.1 f 1.59 51.6* f 1.35 47.6 3.13 64.1" f 2.06 65.8* f 2.01 21.7" f 0.58 24.1* & 0.49 Specific bone marrow volume No. Tibia No. Tibia2 8 10 56.0* + 1.87 55.6* f 2.51 9 10 39.1 f 1.15 39.8 f 0.81 'Values are mean k standard error. Data analyzed by means of Duncan multiple comparison procedure. No., No. of animals. 'Contralateral unfractured tibia-two-tailed paired t-test with fractured tibia. All other data analyzed with Duncan multiple comparison procedure. *P < .05. and Dormer, 1982; Dormer and Ucci, 1984). Furthermore, the mechanism of commitment to a functional pool, studied by Patt and Maloney (1972) and Till (1982), also remains in question. Stem cells may enter the committed stage of development randomly, as suggested by the stocastic theory, or may be induced into final commitment by the hemopoietic microenvironment, according to the deterministic theory. Further insight into the mechanism of stem-cell-mediated hematopoiesis in mice might be gained by studying hemopoietic stress adaptation. Hemopoietic adaptation to specific stresses includes changes in available tissue space. Present studies evaluate such changes that occur in the medullary cavity and relate these changes to hemopoietic activity. Strategies linking medullary cavity space and hemopoietic activity remain uncertain. However, expanded specific marrow volume, as seen in mice following hypoxia, suggests a direct relationship between increased marrow volume and the adaptive strategy, which compensates for increased erythroid demand. Similarly, the medullary cavity expands at the site of a fracture callus. Both 59Fedistribution and histological inspection of the callus during repair was studied by Van Dyke and Harris (1969) and Giuliani et al. (1977). These reports indicate localized increased hemopoietic activity. Data support localized expansion of active he- matopoietic tissue into space made available through resorption of the callus, as reported by Raisz (1981). Although the influence of trauma during fracture cannot be fully assessed, it would appear that increased marrow space alone may serve to support active hematopoietic stroma in the callus. Splenectomized control animals show significant loss of bone mineral volume. This, however, contributes litTABLE 2. Conformation-morphometric measurement and cleaved bone studies' Medullary cavity space as percentage of bone marrow volume (total bone water spacetibias only) Morphometric measurement projection microscopy Cleaved bone method Control Hypoxic 48.6 f 4.0 control animals 45.6 f 3.2 57.8 f 0.8* 'Data indicate that medullary space is less than half the total bone water space. Hypoxic animals with increased specific marrow volume contribute a greater portion of their bone water space to the marrow. Values are mean SE. * P = < ,051vs. control. 532 D.C. GIULIANI, J.C. HALL, AND B.S. MORSE tle to available marrow space, since specific marrow volume increases only in the femur (12.2%). Present data tend to support reported increases in marrow space (Drinker et al., 1922; Johnstone, 192%)associated with the loss of normal splenic hemopoietic activity following splenectomy. Animals treated with estrone show a slight decrease in hematocrit and hemoglobin concentration, which reflects a combination of mild hemolysis and decreased production, as previously reported by Morse et al. (1974). Bone and marrow parameters are consistent with endosteal deposition of a denser mineral complex. Histological examination further confirms the loss of marrow space to bone. Splenectomized, estrone-treated animals respond similarly, but they show a more marked decrease in hematocrit and hemoglobin concentration. These data suggest a greater dependence upon splenic hemopoietic capacity following osteosclerosis. There appears to be a direct relationship between the availability of hematopoietic tissue space and hemopoietic activity in mice. During hypoxia, relative marrow space expands at the expense of bone, as the animals respond to increased erythroid demand. Furthermore, active hematopoietic tissue infiltrates space made available through resorption of a fracture callus in the absence of altered hemopoietic demand, according to earlier studies by Van Dyke and Harris (1969). Increased splenic hematopoiesis accompanies the loss of medullary space to estrone-induced endosteal bone deposition, while splenectomized estrone-treated animals demonstrate frank anemia. Quantitative study has shown that active hematopoietic tissue expands in response to increased demands made upon the functional blood cell pool. This increased space may be achieved 1)at the expense of medullary fat in species with high medullary hemopoietic reserve capacity, 2) through extramedullary hematopoiesis (primarily in the spleen) (Kozlov et al., 1980a, 1980b; Aggio et al., 1973),and 3) through bone resorption. Though the data suggest that increased medullary space may serve as a direct incentive for increased hematopoietic activity, the mechanism for this response remains obscure. Mass action phenomenon, however, might serve to explain foci of intensified hemopoietic activity in this space even in the absence of increased hemopoietic demand. In the mouse, intensified hemopoietic demand is compensated, in part, by selective migration of stem cells from the marrow to the spleen (Kozlov et al., 1980a,b; Aggio et al., 1973). Selective migration might be explained by direct competition between stem cells, i.e., preferential differentiation into one stem cell compartment a t the expense of another. Rather, competition for a limited medullary niche remains indistinguishable from direct competition. If marrow constitutes a preferred site for stem cell selfperpetuation, while the spleen provides a preferred site for their maturation, stem cell migration becomes more understandable. Competition for a limited medullary niche preferred by the self-perpetuating pool of stem cells might initiate the expulsion of daughter cells from the marrow. Multiplication of mature circulating cells from the differentiated pool might proceed in the marrow, if space were available, andor a t extramedullary (migratory) sites, such as the spleen. Evidence in support of this proposal is indirect, though compelling. Clinically, marrow volume increases during conditions of chronic persistant hemopoietic demand, as seen in patients with sickle cell disease and thalassemia major (Raisz 1981). In these diseases, a s in the present studies, active marrow fills the space made available by bone resorption. Proliferating primitive stem cells, (CFUS),tend to distribute in greatest numbers at the marrowbone interface Cord et al., 1975); pronounced increases in specific marrow volume will extend this site while providing increased potential stromal space required for lodging, differentiation, and maturation. Therefore, expanded (endosteal) surface area a t the site of the fracture callus may permit the localized increase in hemopoietic activity, which also potentially increases hemopoietic reserve capacity. Results from estrone-treated animals provide further support for this proposal. Although a lead-shielded limb from a lethally irradiated estrone-treated mouse may provide increased endogenous (migrating) stem cells, splenectomized estrone-treated animals become anemic much more rapidly than normal estrone-treated animals (Morse et al., 1974). Therefore, without adequate sites for proliferating marrow stem cells to lodge and multiply into the functional pool, normal erythroid homeostasis cannot be maintained. In contrast, normal splenectomized animals, exposed to hypoxia, but with increased specific marrow volume, may effectively compensate for the increased erythroid demand. These data strongly support a direct relationship between hemopoietic activity and the availability of hematopoietic tissue space. Furthermore they support a functional distinction between splenic and medullary hematopoiesis, proposing the marrow a s a preferential site for stem cell proliferation. Maturation of stem cells into the function pool may also proceed in the marrow; however, this depends, a t least in part, upon availability of hematopoietic tissue space. During conditions of hemopoietic stress or when available medullary space is otherwise occupied, mouse stem cells preferentially lodge, proliferate, and differentiate at extramedullary (migratory) sites, primarily in the spleen. ACKNOWLEDGMENTS This report was taken in part from a thesis submitted by Dennis Giuliani to the faculty of the Graduate School of Zoology-Physiology in partial fulfillment of the requirements for the Degree of Doctor of Philosophy a t Rutgers University, Newark, New Jersey. LITERATURE CITED Aggio, M.S., M.T. Bruzzo, L.M. Fernandez, and M.J. Montana (1973) Influence of the occlusion of splenic vessels on bone marrow radioiron uptake in the mouse. Acta Physiol. Lat. Am., 23:165-170. Bessis, M., and G. Brecher (1975) A second look at stress erythropoiesis-unanswered questions. Blood Cells, I:409-414. Cronkite, E.P. (1964) Enigmas underlying the study of hemopoietic cell proliferation. Fed. Proc., 23:649-661. Dormer, P., and D. Ucci (1984)Pluripotent stem cells do not completely maintain normal human steady state haemopoiesis. Cell Tissue Kinet., 17:367-374. Drinker, C.K., K.R. Drinker, and C.C. Lund (1922) The circulation in the mammalian bone marrow. Am. J. Physiol., 62:l-92. Giuliani, D., M. Nussbaum, and B.S. Morse (1977)Trauma as a stimulus to marrow regeneration in the osteosclerotic mouse. J. Lab. Clin. Med., 89t804-808. HEMOPOIETIC STRESS AND BONE MARROW VOLUME Gong, J.K. (1965) Effects of altitude acclimatization and deacclimatization on bone and marrow volume in dogs. Am. J. Physiol., 209:347-352. Green, D., G.R. Howells, and M.C. Thorne (1981)Morphometric studies on mouse bone using a computer based image analysis sytems. J. Microsc., 12249-58. Hudson, G. (1958a) Effect of hypoxia on bone marrow volume. Br. J. Haematol., 4:239-248. Hudson, G. (1958b) Bone marrow volume in guinea pigs. J. Anat., 92:150-161. Johnstone, E.M. (1922) Changes in the morphology and function of the bone marrow after splenectomy. Arch. Surg., Fir159-187. Kozlov, V.A., I.N. Zhuravkin, R.M. Coleman, and N.J. Rencricca (1980a) Splenic plaque-forming cells (PFC) and stem cells (CFU-S) during acute phenylhydrazine induced enhanced erythropoiesis. J. Exp. Zool., 213:199-203. Kozlov, V.A., I.N. Zhuravkin, R.M. Coleman, and N.J. Rencricca (1980b) Alerations in the levels of stem cells (CFU-S)and plaque-forming cells (PFC) in mice during chronic phenylhydrazine induced hemolytic anemia. J. Exp. Zool., 211:357-360. Lange, R.D., M.L. Simmons, N.R. Dibelius (1966) Polycythemic mice produced by hypoxia in silicone rubber membrane enclosure: A new technique. Proc. SOC. Exp. Biol. Med., 122,761-764. Lord, B.I., N.G. Testa, and J.H. Hendry (1975) The relative spatial distribution of CFU-S and CFU-C in the normal mouse femur. Blood, 46:65-72. 533 Mackey, M.C., and P. Dormer (1982) Continious maturation of proliferating erythroid precursors. Cell Tissue Kinet., 15381-392. Morse, B.S., N.J. Rencircca, and P. Stohlman (1969) The mechanism of action of erythropoietin in relationship to cell cycle kinetics. In: Hemopoietic Cellular Proliferation. F. Stohlman, ed. Grune and Stratton, New York, pp. 160-170. Morse, B.S., D.C. Giuliani, M. Soremkun, J. DiFino, and E.R. Giuliani (1974) Adaptation of hemopoietic tissue resulting from estrone Induced osteosclerosis in mice. Cell Tissue Kinet., 7:113-123. Patt, H.M., and M.A. Maloney (1972) Bone formation and resorption Exp. Biol. as a requirement for marrow development. Proc. SOC. Med., 140:205-207. Raisz, L.G. (1981) What marrow does to bone. N. Engl. J. Med., 304:1485-1486. Simmons, D.J. (1966) Collagen formation and endochondral ossification in estrogen treated mice. Proc. SOC.Exp. Biol. Med., 121,11651168. Till, J.E. (1982) Stem cells in differentiation and neoplasia. J. Cell. Physiol. (Suppl.], 1:3-11. Van Dyke, D., and N. Harris (1969) Bone marrow reactions to trauma. Stimulation of erythropoietic marrow by mechanical disruption, fracture and endosteal curettage. Blood, 34,257-275. Wintrobe, M.M. (1972) Clinical Hematology. Lea and Febiger, Philadelphia. Yoffey, J.M. (1966)Bone Marrow Reactions. The Williams and Wilkins Co., Baltimore.