Characterization of rabbit stromal fibroblasts derived from red and yellow bone marrow.

Rabbit stromal fibroblasts subcultured from red and yellow bone marrow and implanted beneath the renal capsule form ossicles the hemic cellularity of which mirrors the cellularity of the marrow used for culture. Although the cultured red and yellow marrow cells are similar in fine-structural appearance, they differ strikingly in enzymatic content of alpha-naphthylbutyrate esterase, which is abundant only in the cells derived from yellow marrow. Other observers (20, 21) have proposed that stromal fibroblasts are preadipocytes, and this data suggests that those derived from yellow marrow have the phenotype of more differentiated adipocytes. On the other hand, fibroblasts derived from red and yellow bone marrow show no differences in their profiles of procollagen synthesis. Both types of fibroblasts secrete type III procollagen as the major species, with a I/III ratio of 1:3; in contrast, rabbit dermal fibroblasts have a prominent peak of type I procollagen. The similarity of stromal cells derived from red and yellow bone marrow in procollagen synthesis suggests that the collagen part of the extracellular matrix is not the only basis for their intrinsic difference in capacity for hematopoiesis.

The question of bone marrow stromal fibroblast traffic.

Bone marrow stromal fibroblasts (CFU-F) normally do not exchange bone marrow sites in vivo. Restitution of the CFU-F after radiation damage is primarily recovery by the local fibroblasts from potentially lethal damage. Migration of stromal fibroblasts from shielded sites to an irradiated site makes a minimal contribution, if any, to CFU-F recovery. Determination of the relative contribution of donor stromal cells in bone marrow transplants by karyotyping the proliferating bone marrow stromal cells in vitro may not reflect the relative distribution of fibroblasts in the marrow. If there is residual damage to the host stromal fibroblasts from treatment before transplantation, these cells may not be able to proliferate in vitro. Therefore, an occasional transplanted fibroblast may contribute most of the metaphase figures scored for karyotype.

Decrease in hematopoietic stem cell domains as a delayed effect of x-irradiation.

Although the hematopoietic integrity of locally X-irradiated sites can be restored for a time even after fairly large doses, a secondary aplasia often occurs some months later. To gain further insight into this delayed effect within the framework of the stem cell regulatory domain hypothesis, we characterized the growth kinetics of spleen colony forming units (CFU-S) in WBB6FI-+/+ bone marrow transplanted into WBB6FI-W/WV mice in which one leg had been exposed to 10-30 Gy of X rays 4-5 months previously. Compared to unirradiated contralateral marrow, fewer CFU-S either reached the previously irradiated marrow or were seeded into sites that could support growth. The initial exponential growth of effectively seeded CFU-S was unchanged, but growth deceleration (inflection point) occurred at a lower level of CFU-S in marrow previously irradiated with 20-30 Gy. This change in the inflection point indicates a radiation dose-dependent decrease consistent with the decrease in bone marrow cellularity. The decrease in effective stem cell domains after 20 Gy was calculated to be about 35%. We interpret these results to reflect the highly localized nature of delayed radiation damage to the marrow microenvironment.

Hematopoietic microenvironment transfer by stromal fibroblasts derived from bone marrow varying in cellularity.

Autologous fibroblast derivatives of red and yellow marrow of rabbits were shown to differ in their capability to transfer a hematopoietic microenvironment upon implantation under the renal capsule. Although a heterotopic ossicle formed in each instance, the quality of the associated medullary tissue mirrored the quality of the bone marrow used to generate the stromal fibroblasts. Thus, fibroblasts cultured from a cellular marrow produced a stroma with numerous hematopoietic foci whereas those cultured from a severely hypocellular marrow produced a stroma with mainly fat cells. The results with 21 implants point to a transmittable regulatory role of a class of stromal fibroblasts.

Partitioning of bone marrow into stem cell regulatory domains.

To examine the hypothesis that bone marrow consists of discrete stem cell regulatory volumes or domains, we studied spleen colony-forming unit (CFU-S) population growth kinetics in unirradiated WBB6F1-W/Wv mice receiving various doses of +/+ bone marrow cells. Assay of femoral marrow CFU-S content in the eight recipient dose groups revealed a family of growth curves having an initial dose-independent exponential phase and a subsequent dose-dependent deceleration phase. CFU-S content at the growth transition (inflection point) was not a simple linear function of inoculum dose but was shown rather to reflect a random distribution of initially seeded donor CFU-S in discrete volumes of recipient bone marrow. The inoculum dose resulting in a mean of 1 CFU-S per bone marrow sampling unit was estimated to be 17 x 10(6) bone marrow cells, corresponding to a total marrow uptake of approximately 5100 CFU-S (based on a seeding efficiency factor of 10%). If we assume single-hit kinetics, it follows that the recipient W/Wv bone marrow may contain approximately 5100 domains in which stem cell proliferation is geared to the density of the stem cell population. When the various inocula were corrected for multiple seeding in a given domain, the mean inflection point per domain was similar and indicative of five or so divisions before departure from exponential growth at approximately 20% of final CFU-S content 8 days after bone marrow injection. The partitioning of bone marrow into highly localized functional units is consistent with the putative regulatory role of short-range interactions between stem cells and essential stromal elements.

Hematopoietic stem cell proliferative behavior as revealed by bromodeoxyuridine labeling.

Hematopoietic stem cells (CFUS) are thought to represent a heterogeneous population with different probabilities of self-renewal and different rates of proliferation. As an approach to further characterization of this population, we determined the disappearance of bromodeoxyuridine (BrdUrd)-induced CFUS sensitization to ultraviolet light in normal mice infused with BrdUrd for three weeks and in hydroxyurea-pretreated mice regenerating their CFUS pool during a one-week BrdUrd infusion. the same exponential disappearance with a T1/2 of six days was found in each case. From this and from the apparent absence of a secondary slope on the sensitization decay curve for the recovered hydroxyurea-treated bone marrow, we conclude that fewer than 10% of CFUS may be mitotically quiescent for prolonged periods and that the age structure of the CFUS population reflects a relatively short proliferative history.

Influence of a marrow stromal factor on survival of hemopoietic stem cells in vitro.

When mouse bone marrow or spleen cells were incubated for 24 hours in medium conditioned by marrow adherent cells, the survival of pluripotent hemopoietic stem cells (CFUs) was considerably greater than that of stem cells incubated in fresh or spleen-conditioned medium. Medium conditioned by endosteal cells also increased marrow CFUs survival. Survival of CFUs more than 24 hours was greater when fresh marrow was incubated directly with marrow adherent cells than with medium conditioned by these cells. Although marrow-conditioned medium greatly increased CFUs susceptibility to 3H-thymidine suicide, DNA synthesis inhibitors did not prevent the increased survival of CUFs in marrow-conditioned medium. These results indicate that marrow stromal cells produce a factor that increases CFUs survival in vitro independently of proliferation.

Turnover of circulating hematopoietic stem cells.

Short-term parabiosis of male and female CBA/CaJ mice was used to investigate the turnover of circulating hematopoietic stem cells. The exchange and subsequent disappearance of donor stem cells were monitored by spleen colony assay and chromosome analysis of individual colonies. The results revealed an exponential disappearance of pluripotent stem cells from blood with a characteristic half time of 1.7 h. Blood-borne stem cells were shown to be equilibrated with a subpopulation of marrow stem cells exhibiting a disappearance half time of 9.5 h. Splenectomy did not change the apparent rate of stem cell removal from the blood.

Increased survival of CBA pluripotent haemopoietic stem cells in vitro induced by a marrow stromal factor in Sl/Sld mice.

Media conditioned by marrow adherent cells from anaemic Sl/Sld and W/Wv mice increased the 24-h survival of CBA CFUs in vitro compared to fresh medium to about the same extent as marrow-conditioned medium from normal Sl+/Sl+, W+/W+, and CBA mice. Sl/Sld marrow-conditioned medium also increased the percentage of CFUs in DNA synthesis to the same extent as CBA marrow-conditioned medium. These results demonstrate that Sl/Sld mice produce a marrow stromal factor that increases both survival of CFUs and the percentage of CFUs in DNA synthesis in vitro. Therefore, the defective haemopoietic microenvironment of Sl/Sld mice is not due to a deficiency in the production of this factor.

Differential response of early erythropoietic and granulopoietic progenitors to dexamethasone and cortisone.

The sensitivity of erythropoietic (BFU-E) and granulopoietic (CFU-C) progenitor cells to dexamethasone and cortisone was studied in cultures of mouse bone marrow. Although the log dose-response relationships had a similar form, the BFU-E were much more sensitive than the CFU-C to either glucocorticoid. The dexamethasone concentration for 50% inhibition was 3 X 10)-9) M for BFU-E and 60 X 10(-9) M for CFU-C. The differential sensitivity to cortisone was even greater, with 60% inhibition of BFU-E and 18% inhibition of CFU-C at 0.1 microgram/ml. These findings suggest a specific rather than a general response to glucocorticoids and indicate that granulocyte-macrophage progenitors are less affected than early erythroid progenitors by physiologic concentrations of these hormones.

Hematopoietic stem cell regulatory volumes as revealed in studies of the bgj/bgj:W/WV chimera.

The kinetics of bone marrow replacement was studied in W/WV mice implanted with gbj/bgj (beige) stem cells, with the characteristic beige neutrophil marker as a criterion of the takeover of host marrow by donor marrow. A hyperbolic pattern of W/WV marrow replacement conforming to a log dose-response was observed in experiments encompassing a 50-fold range of bgj/bgj inoculum doses and a 2-yr period of observation. The dose-response relationships were consistent with random seeding of stem cells in the host marrow coupled with a decreasing efficiency of secondary colonization by local migration. Application of single-hit Poisson sampling statistics to the dose-response data led to the hypothesis that mouse bone marrow is compartmentalized into essentially self-contained stem cell regulatory volumes or domains. We estimate that W/WV marrow contains about 2,600 stem cells regulatory units with an average volume of about 10(8) micron3, a dimension consistent with the presumptive role of short-range cell-cell interactions in the regulation of pluripotent stem cells. Our analysis of the dose-response data is also indicative of the discontinuous and limited nature of local stem cell migration in a cellular marrow, a consideration that may be of practical as well as theoretical interest.

Concentration gradient of blood stem cells in mouse bone marrow--an open question.

The distribution of pluripotential blood stem cells (CFU-S) in bone marrow was studied in four strains of mice. Analyses of paired samples containing various fractions of axial and marginal bone marrow cells, as well as longitudinal 200-micrometer sections, revealed a rather uniform spatial distribution of CFU-S across the diameter of the femoral medullary cavity, in contrast to the finding by others of a CFU-S concentration gradient extending from the endosteum to the central longitudinal axis of bone marrow. The existence of a stem cell gradient is therefore open to question.

Marrow stem cell release in the autorepopulation assay.

The early migration of stem cells from a shielded marrow to an irradiated spleen has been re-evaluated, and the findings have been compared with the results of earlier studies. The composite data reveal a constant rate during the first 24 h after irradiation, with a slope of 1.6 cells per h and an intercept of 2.4. The positive intercept is interpreted to signify an immediate brief perturbation of CFU's release. The low concentration of CFUs in the bloodstream, despite their continuous migration from the shielded marrow, is indicative of a rapid, and probably greatly increased, blood turnover. Despite the constancy of stem cell seeding, it is not yet possible to determine whether the rate of stem cell release is different in shielded marrow than in normal marrow. The resolution of this question requires more precise information about spleen seeding efficiency in the autorepopulation assay and about the normal turnover rate of stem cells in the bloodstream.

The bgJ/bgJ:W/WV bone marrow chimera. A model for studying stem cell regulation.

In studies with bgJ/bgJ:W/WV chimeric mice, we used the biege neutrophil marker as a criterion of of W/WV marrow replacement by implanted bgJ/bg/J stem cells. Data from a 50-fold range of inoculum doses and a 2-year period of observation indicate a hyperbolic pattern of replacement expressed as a log dose-response relationship. The saturating effect with increasing inoculum dose was interpreted as reflecting random initial stem cell seeding in bone marrow coupled with a decreasing efficiency of colonization by migration. From the statistics of random sampling and the exponential decrease of the 63% replacement dose with time, we estimate that W/WV marrow contains about 2600 stem cell regulatory volumes of about 10(8) mu3 (50 cell diameters) each, a dimension consistent with concepts of short-range cell-cell interactions. Our observations suggest that each regulatory volume is essentially self-contained and that stem cell migration is generally restricted to contiguous volumes.

Effect of hypertransfusion on bone marrow restoration after localized depletion.

The reconstitution of marrow in a mechanically depleted medullary cavity was compared in normal and hypertransfused mice. Polycythemia (hematocrit of approximately 70%) prevented the emergence of recognizable erythroid cells but not of erythropoietin-responsive cells as revealed by autoradiographic studies with 55Fe after erythropoietin treatment. The similar regeneration of erythropoietin-responsive cells in normal and hypertransfused mice was shown by comparison of 55Fe labeling in the repopulating relative to the contralateral marrow; in each case, labeling in the repopulating marrow was about 20% of that in the intact contralateral marrow. Polycythemia did not influence the recovery of granulocytic elements in the regenerating marrow even though the myeloblast-metamyelocyte population of the intact contralateral marrow was increased. The findings are intrepreted to support the concept that hematopoietic stem cell commitment to one or another pathway of differentiation is basically a local phenomenon.

Increased survival of haemopoietic pluripotent stem cells in vitro induced by a marrow fibroblast factor.

When mouse bone marrow was incubated in medium conditioned by marrow fibroblasts, the survival of pluripotent stem cells (CFUs) was considerably greater than when marrow was incubated in fresh medium. This increase in CFUs survival depended on the age of the marrow fibroblast culture, the initial number of cells in the culture, and the concentration of the conditioned medium. Medium conditioned by fibroblasts from other adult tissues--spleen, bone, and subcutaneous tissue--did not increase CFUs survival, but medium conditioned by embryo bone did. The increase in CFUs survival by marrow-fibroblast-conditioned medium was not accompanied by any change in the total number of nucleated cells of the incubated marrow nor by any comparable increase in the survival of granulopoietic stem cells (CFUc) or erythropoietic stem cells (BFUE). These results indicate that marrow fibroblasts produce a factor that increases the survival of CFUs, which may be involved in the role of marrow stroma in the control of haemopoiesis.

On the origin of hematopoietic stem cells after local marrow extirpation.

The early hematopoietic regeneration in a depopulated segment of femur shaft is compared in +/+ and W/Wv mice and in W/Wv mice previously treated with +/+ marrow. Since the W/Wv mouse has an intrinsic CFU deficiency on spleen colony assay and since immigrant cells play a negligible role in the onset of regeneration after marrow extirpation, the W/Wv(+/+) chimera provides a model for evaluation of the contribution of residual cells to the regenerative program. There was little difference in the relative recovery of CFU in +/+, W/Wv, and W/Wv-(+/+), Moreover, +/+ derived CFU were responsible for nearly all of the CFU repopulation in chimeric mice. Thus, recovery of hemic cellularity must be due to residual stem cells rather than to stem cells derived by transformation of more primitive mesenchymal elements. The residual CFU are probably intimately associated with bone, most likely within the endosteum and haversian system.

Bone marrow regeneration after local injury: a review.

This paper is focused on a mechanically depleted medullary cavity as an experimental model for analysis of marrow regenerative programs. The reconstitution of marrow in an evacuated cavity is basically a local phenomenon in respect to the stimulus for regeneration and the origin of the responsible cells. The nature of the triggering stimulus is unknown, but it is probably related to disruption of the continuity of the marrow stroma and endosteum. The initiating cells appear to be independent lines of mesenchymal and hematopoietic stem cells bound to bone, most likely within the endosteum and haversian system. The mesenchymal cells form the characteristic marrow stroma. Hemic cell regeneration can occur without immigrant hematopoietic stem cells, although such cells are known to contribute to later stages of repopulation. The formation and resorption of trabecular bone appears to be intimately related to the development of a sinusoidal matrix, perhaps by serving as a callus or supporting lattice and perhaps by providing a mechanism for distribution of stromal progenitors. Hematopoiesis is initiated in sites of active bone resorptive. The interplay of events consequent to marrow removal is strikingly similar to that seen with heterotopic marrow implants. Because stromal stem cells, unlike hematopoietic stem cells, do not migrate from distant sites, marrow stroma is the limiting factor in recovery from localized injury. Stromal stem cells are fairly radiosensitive but are not as sensitive as hematopoietic stem cells. The apparent radioresistance of stromal elements in an intact marrow seems to be due to their very low turnover rate. Latent radiation damage can be readily unmasked by conditions that promote their proliferation. This no doubt accounts for the radiosensitivity of stroma in an evacuated femur or heterotopic implant in contrast to its continued functional integrity with similar irradiation of in situ marrow. Even in an intact marrow, however, exposures in the 1000 rad range can lead to slowly evolving hypocellularity associated with diminished blood flow. With higher doses, aplasia of the irradiated site becomes progressively more generalized. It remains to be seen whether this limiting condition is due to the loss of specific regulatory functions or stromal components or merely reflects sinusoidal damage.