A comparative morphometric study on the ultrastructure of adherent cells in long-term bone marrow culture from normal and congenitally anemic mice

The relationship between structure and function of bone marrow stromal tissue in adherent layers of long-term bone marrow cultures (LTBMCs) from normal and congenital anemic mice (C57BL, Sl/Sld, Sl+/Sl+, W/Wv, and W+/W+) was investigated. Many previously reported features were confirmed. However, in LTBMC from all strains of mice examined, isolated cilia with the axonemal structure of a 9 + O pattern with obvious dynein arms were observed in the blanket cells. The frequency of cilia was approximately 2%-5% of total number of profiles of blanket cells examined. Crystalloid inclusions (CI) were observed in cultured macrophages similar to those reported in vivo in all strains of murine LTBMC. The CI could be classified into four types according to their structure in the same way as in vivo (type A to type D), with a predominance of type A in the cultures. Viral particles were also apparent in adherent cells of all strains (except W/Wv and W+/W+), which were compatible with a type C retrovirus. Gap junctions occurred regularly between the adherent cells of LTBMC, particularly between blanket cells and preadipocytes. The most frequent appearance of gap junctions was found in Sl/Sld cultures. The phenomena of normal and abnormal hematopoiesis appear to be accurately reproduced in culture, thus retaining the same relationship between function and structure as occurs in vivo. The surface of isolated cilia of blanket cells, CI of macrophages, viral particles among adherent cells, and gap junctions between blanket cells and preadipocytes is discussed.     

Sl/Sld mice have an increased number of gap junctions in their bone marrow stromal cells

To study the defect of the hematopoietic inductive microenvironment (HIM) in Sl/Sld mice, femoral bone marrow tissue of 10 of each mutant, (Sl/Sld and W/Wv) their normal littermates (Sl+/Sl+ and W+/W+), and 20 normal C57BL mice were examined by electron microscopy using morphometric and statistical methods. Gap junctions were observed in all strains of mice, in the following stromal cell types: 1) reticular cells, 2) between reticular cells and periarterial adventitial cells, and 3) between periarterial adventitial cells. The frequency of gap junctions in bone marrow stromal cells of Sl/Sld mice (mean = 2.2/9.4 X 10(-3) mm2) was significantly higher than in control mice. It is suggested that there is a relationship between the increased numbers of gap junctions in bone marrow stromal cells of Sl/Sld mice and the defect in HIM function in these genetically anemic animals.     

Ultrastructural morphometric study of efferent nerve terminals on murine bone marrow stromal cells, and the recognition of a novel anatomical unit: the "neuro-reticular complex"

In order to extend our understanding of the role of nerve fibers in the structure and function of bone marrow stroma, we have examined nerve terminals, arterioles, and capillaries in femoral bone marrow tissues of 50 C57BL strain mice, using electron microscopy and morphometric methods. Within the adventitia of arterioles, a particular type of cell, termed periarterial adventitial (PAA) cell, is characterized by a thin veil-like cytoplasm which concentrically surrounds both nerves and arterioles. Nerve fibers containing both unmyelinated and myelinated axons are distributed mainly between the layers of PAA cells, but are found rarely on the sinus walls or within the hematopoietic parenchyma. Quantitatively, the efferent nerve terminals with many synaptic vesicles are distributed mainly beside arterial smooth muscle cells (Type I: 58.8%) or between the layers of PAA cells (Type III: 33.2%), and rarely in hematopoietic parenchyma (Type II: 5.3%) or on sinus walls (Type IV: 2.7%). In the case of Type II-IV nerve terminals, efferent (autonomic) nerves and bone marrow stromal cells which are connected by gap junctions (sinus adventitial reticular cells, intersinusoidal reticular cells, and PAA cells) appear to constitute a potential functional unit for signal conduction. We would like to propose a new term for this anatomical unit in marrow, the "neuro-reticular complex."     

Macrophages and stromal cells phagocytose apoptotic bone marrow-derived B lineage cells

It has been hypothesized that B cell precursors that undergo programmed cell death due to nonproductive Ig gene rearrangements are cleared from the bone marrow by macrophages. However, a role for macrophages in this process is supported only by micrographs showing their association with apoptotic-appearing, B lineage cells. Functional data demonstrating phagocytosis of apoptotic, bone marrow lymphocytes by macrophages have not been presented, nor have receptors potentially involved in that process been identified. The data in this report demonstrate that macrophages isolated from murine bone marrow efficiently phagocytose apoptotic murine B lineage cells using multiple receptors that include CD14, integrins, class A scavenger receptor, and CD31 (PECAM-1). In addition, the results further reveal a new role for the hemopoietic microenvironment in B cell development in view of data demonstrating that murine bone marrow stromal cells are also capable of clearing apoptotic cells via an integrin-dependent mechanism.     

Stromal cells of the bone marrow in growing bone

By means of electron microscopy, cytochemistry and radioautography with 3H-thymidine, the bone marrow stromal cells have been studied in the zones of endochondral osteogenesis in the rabbit and rat femoral bones. In the stromal cells demonstrating a high alkaline phosphatase activity are distinguished: perivascular, reticular fibroblastic, osteogenic cells. Populations of the perivascular phosphatase-positive cells include poorly differentiated DNA-synthesizing forms, as well as cells with signs of differentiation into stromal fibroblasts. Cleft-like spaces in cytoplasm of the fibroblastic reticular cells are, probably, formed as a result of lymphocyte-like mononuclears passing through. Phagocyting stromal elements are presented by macrophages, having perivascular localization and including into composition of erythroblastic islets. Mononuclear macrophages are revealed also on the surface of osseous trabecules, where they participate in destruction of hemopoetic and osteogenic cells.     

Ultrastructural and morphometric alterations in bone marrow stromal tissue after 7 Gy irradiation.

To evaluate the response of marrow stroma to 7 Gy irradiation, femoral bone marrow was fixed by vascular perfusion (so as to avoid the artificial destruction of sinus endothelia), and was examined using light and electron microscopy with morphometric methods. The radiation caused a marked decrease in hematopoietic cell number (NHC) within 3 days post-irradiation, followed by total recovery of hematopoiesis, which occurred gradually over 28 days. An increased number of fat cells was seen by 7 days. During the whole course of hypoplasia and recovery, the continuity of sinus wall, three-dimensional reticular mesh work in hematopoietic parenchyma, gap junctions (GJ) between stromal cells, the adventitial cell cover of sinus wall (ACC), and the stromal cell numbers of reticular cells (RC), sinus endothelia (SE), and macrophages (MP) were maintained. The cellularity of stromal components of RC, SE, and MP seemed passively increased in contrast to a reduction in numbers of NHC. A similar tendency was observed (1) between NHC and ACC and (2) between GJ and the cellularity of fat cells, which had a statistical significant correlation (p less than 0.05; t-test). The mechanism of radio resistance in bone marrow stroma and the possible functional adaptation and cellular coordination after irradiation are discussed.     

Osteogenic properties of adhesive cells in Dexter culture of the mouse bone marrow

Disaggregated cell suspensions obtained by mouse bone marrow fermentative digestion as well as stromal tissue obtained by marrow mild mechanical destruction were explanted. Both methods yield the cultures in which the hematopoiesis duration is comparable with dexter cultures. Adhesive cells from all of these three culture types were resuspended and in the porous gelatin sponges heterotopically transplanted under the kidney capsule of syngenic recipients. In the transplantation site there develops the hemopoietic organ containing reticular stroma, hemopoietic cells, and in most cases the well developed bone tissue. Thus, the adherent layers of mouse bone marrow dexter and similar cultures contain for a long period (not less than 2-3.5 months) the stromal fibroblast population which maintains its osteogenic and hemopoietic microenvironment transfer capacities.     

Adipocytes and reticular cells of the bone marrow stroma in the human iliac bone

The results of the histological and electron microscopic investigation of adipose and reticular cells and their interconnections with blood cells are presented in the material of trephine biopsies of the iliac bone. A possibility for development of adipocytes from the adventitial reticular cells is demonstrated. Close contacts are revealed between pre-adipocytes and young hemopoietic cells. Two types of the reticular cells are characterized, they differ in their position, structural organization and interconnection with the young hemopoietic elements. The peculiarities revealed in the morphofunctional state of the microenvironmental structures demonstrate functional variegation of the stromal elements, and also attest an essential importance of intercellular contacts of the hemopoietic predecessors and the stromal cells in maintaining the hemopoietic function of the bone marrow.     

Pre-CFU-f: young-type stromal stem cells in murine bone marrow following administration of DNA inhibitors

Occurrence of young-type stromal stem cells (defined here as "pre-CFU-f") in murine bone marrow is reported in this study. Two consecutive intraperitoneal (i.p.) cytosine arabinoside (ara-C) injections were administered to C57B1 mice (2 X 200 mg/kg at 6-h intervals). Two days later the bone marrow was collected and assayed for colony-forming units-fibroblastoid (defined here as "CFU-f"). In additional experiments, ara-C-treated marrow was exposed in vitro to hydroxyurea (HU; "hydroxyurea killing test"), prior to plating, to establish the cycling state of stromal stem cells. In separate cultures of ara-C-treated marrow, replating of adherent cells was carried out up to quaternary sub-cultures. The results indicate ara-C-treated marrow produces approximately 20% "huge" fibroblastoid colonies (approximately 5 mm diameter versus 0.5-2 mm normal size); most stromal stem cells producing huge colonies are cycling cells; and adherent cells from primary ara-C-treated marrow cultures replated to secondary cultures produce adherent layers with double the number of cells than in the control secondary cultures. We conclude that the ara-C-treated murine bone marrow contains certain young-type cycling stromal stem cells which we refer to as pre-CFU-f. These stem cells produce huge fibroblastoid colonies in culture, indicating that they probably go through more cell cycles than CFU-f during the culture period. Alternatively, pre-CFU-f may have a higher self-replicative capacity than CFU-f.     

小鼠胎肝和骨髓造血基质细胞贴壁层对红系祖细胞生长的影响

本实验以Dexter培养体系作小鼠胎肝和骨髓造血基质细胞贴壁培养。在所获的基质细胞贴壁层上作红系造血祖细胞集落培养,观察两种来源造血基质细胞对红系集落生长的影响。实验结果表明,胎肝造血基质细胞贴壁层能明显促进早期红系造血祖细胞(BFU-E)形成集落,却不明显影响晚期红系造血祖细胞(CFU-E)的生长。成年小鼠骨髓造血基质细胞贴壁层对BFU-E和CFU-E均有刺激生长的作用;但对前者生长的刺激性影响较胎肝造血基质细胞贴壁层为弱。造血基质细胞贴壁层对红系集落生长的促进作用主要是通过体液因子实现的,细胞间短距离调节的影响亦不能除外。     

Reconstruction of hematopoietic inductive microenvironment after transplantation of VCAM-1-modified human umbilical cord blood stromal cells

The hematopoietic inductive microenvironment (HIM) is where hematopoietic stem/progenitor cells grow and develop. Hematopoietic stromal cells were the key components of the HIM. In our previous study, we had successfully cultured and isolated human cord blood-derived stromal cells (HUCBSCs) and demonstrated that they could secret hemopoietic growth factors such as GM-CSF, TPO, and SCF. However, it is still controversial whether HUCBSCs can be used for reconstruction of HIM. In this study, we first established a co-culture system of HUCBSCs and cord blood CD34(+) cells and then determined that using HUCBSCs as the adherent layer had significantly more newly formed colonies of each hematopoietic lineage than the control group, indicating that HUCBSCs had the ability to promote the proliferation of hematopoietic stem cells/progenitor cells. Furthermore, the number of colonies was significantly higher in vascular cell adhesion molecule-1 (VCAM-1)-modified HUCBSCs, suggesting that the ability of HUCBSCs in promoting the proliferation of hematopoietic stem cells/progenitor cells was further enhanced after having been modified with VCAM-1. Next, HUCBSCs were infused into a radiation-damaged animal model, in which the recovery of hematopoiesis was observed. The results demonstrate that the transplanted HUCBSCs were "homed in" to bone marrow and played roles in promoting the recovery of irradiation-induced hematopoietic damage and repairing HIM. Compared with the control group, the HUCBSC group had significantly superior effectiveness in terms of the recovery time for hemogram and myelogram, CFU-F, CFU-GM, BFU-E, and CFU-Meg. Such differences were even more significant in VCAM-1-modified HUCBSCs group. We suggest that HUCBSCs are able to restore the functions of HIM and promote the recovery of radiation-induced hematopoietic damage. VCAM-1 plays an important role in supporting the repair of HIM damage.     

Role of Nanog in the maintenance of marrow stromal stem cells during post natal bone regeneration

Post natal bone repair elicits a regenerative mechanism that restores the injured tissue to its pre-injury cellular composition and structure and is believed to recapitulate the embryological processes of bone formation. Prior studies showed that Nanog, a central epigenetic regulator associated with the maintenance of embryonic stem cells (ESC) was transiently expressed during fracture healing, Bais et al. In this study, we show that murine bone marrow stromal cells (MSCs) before they are induced to undergo osteogenic differentiation express ～50× the background levels of Nanog seen in murine embryonic fibroblasts (MEFs) and the W20-17 murine marrow stromal cell line stably expresses Nanog at ～80× the MEF levels. Nanog expression in this cell line was inhibited by BMP7 treatment and Nanog lentivrial shRNA knockdown induced the expression of the terminal osteogenic gene osteocalcin. Lentivrial shRNA knockdown or lentiviral overexpression of Nanog in bone MSCs had inverse effects on proliferation, with knockdown decreasing and overexpression increasing MSC cell proliferation. Surgical marrow ablation of mouse tibia by medullary reaming led to a ～3-fold increase in Nanog that preceded osteogenic differentiation during intramembranous bone formation. Lentiviral shRNA knockdown of Nanog after surgical ablation led to an initial overexpression of osteogenic gene expression with no initial effect on bone formation but during subsequent remodeling of the newly formed bone a ～50% decrease was seen in the expression of terminal osteogenic gene expression and a ～50% loss in trabecular bone mass. This loss of bone mass was accompanied by an increased ～2- to 5-fold adipogenic gene expression and observed increase of fat cells in the marrow space. In summary these data show that Nanog is expressed during surgically induced marrow bone formation and is functionally involved in post natal marrow stromal cell maintenance and differentiation.     

Bone marrow stromal cell regulation of B-lymphopoiesis. I. The role of macrophages, IL-1, and IL-4 in pre-B cell maturation

Our experiments have addressed regulation of B lymphocyte formation by bone marrow stromal cells. Stromal cells appear to produce a regulatory factor that acts at the pre-B cell stage to induce the expression of Ig L chains and surface Ig. Bone marrow stromal cell conditioned medium was found to contain this factor and the active component was partially purified by HPLC. This stromal cell-derived factor had a m.w. between 16,000 and 20,000, was specifically neutralized by anti-IL-4 mAb, 11B11, and enhanced the proliferation of anti-mu-stimulated B cells. We also found that rIL-4 induced B cell formation in culture. In our studies, IL-1 had no direct effect on pre-B cell maturation, however, IL-1 was found to stimulate the production of IL-4 by both heterogeneous bone marrow stromal cells and a cloned stromal cell line, SCL-160. These effects of IL-1 on factor production by stromal cells were duplicated by the addition of bone marrow-derived macrophages to SCL-160 cells. We conclude that stromal cell-derived IL-4 is a physiologic stimulator for B cell generation. In addition, macrophages appear to play a role in B cell formation by regulating the production of IL-4 by stromal cells via the secretion of IL-1.     

Rabies virus replication in primary murine bone marrow macrophages and in human and murine macrophage-like cell lines: implications for viral persistence.

To determine whether rabies viruses replicate in macrophage or macrophage-like cells, several human and murine macrophage-like cell lines, as well as primary cultures of murine bone marrow macrophages, were incubated with the Evelyn-Rokitnicki-Abelseth (ERA) virus and several different street rabies viruses (SRV). ERA rabies virus replicated well in human monocytic U937 and THP-1 cells and murine macrophage IC-21 cells, as well as primary cultures of murine macrophages. Minimal replication was detected in murine monocytic WEHI-3BD- and PU5-1R cells, and ERA virus did not replicate in murine monocytic P388D1 or J774A.1 cells. A tissue culture-adapted SRV of bat origin also replicated in IC-21 and U937 cells. Non-tissue culture-adapted SRV isolated from different animal species, particularly bats, replicated minimally in U937, THP-1, IC-21 cells and primary murine bone marrow macrophages. To determine whether rabies virus replication is dependent upon the state of differentiation of the macrophage-like cell, human promyelocytic HL-60 cells were differentiated with 12-O-tetradecanoylphorbol-13-acetate (TPA). ERA rabies virus replicated in the differentiated HL-60 cells but not in undifferentiated HL-60 cells. Persistent infections were established in macrophage-like U937 cells with ERA rabies virus and SRV, and infectious SRV was isolated from adherent bone marrow cells of mice that had been infected 96 days previously. Virus harvested from persistently infected U937 cells and the adherent bone marrow cells had specifically adapted to each cell. This specificity was shown by the inability of the viruses to infect macrophages other than U937 cells and primary bone marrow macrophages, respectively. Virus titers of the persistently infected U937 cells fluctuated with extended cell passage. After 30 passages, virus released from the cells had lost virulence as shown by its inability to kill intracranially inoculated mice. However, the avirulent virus released from the persistently infected cells was more efficient in infecting and replicating in naive U937 cells than the virus which was used to establish the persistent infection. These results suggest that macrophages may serve as reservoirs of infection in vivo, sequestering virus which may subsequently be activated from its persistent state, resulting in clinical infection and death.     

Morphology of interactions of the haematopoietic microenvironment with the haematopoietic cells in erythroid spleen colonies

The morphology of the interactions of the cells of the haematopoietic microenvironment with haematopoietic cells was studied in exogenous erythroid spleen colonies formed in mice, on the fifth and eighth day after their irradiation with a lethal dose of gamma rays and bone marrow transplantation. The characteristic type of stromal cell interacting with less mature cells of the erythroid series was a dark, branching reticular cell. The typical structural interaction of the reticular cells with erythroblasts was the formation of very long, fine cytoplasmic processes by the reticular cells. The processes were in close contact with the erythroblasts and formed a three-dimensional network stretching long distances from the nucleus of the reticular cells. Other cells of the haematopoietic microenvironment, in contact chiefly with poly- and ortochromic erythroblasts, were macrophages and the two together formed typical erythroblastic islands. In places, the macrophages and erythroblasts formed close, firm contacts by means of their cytoplasmic membranes. These morphological observations support the conception that close functional cooperation exists between the cells of the haematopoietic microenvironment and the haematopoietic cells in differentiation and proliferation processes in the haematopoietic tissue of spleen colonies.     

Morphologic analysis of long-term bone marrow cultures that support B-lymphopoiesis or myelopoiesis

Summary A comparative morphological analysis of the Whitlock-Witte long-term B-cell culture and the predominantly myeloid Dexter long-term bone marrow culture demonstrates that similarities and differences exist between the two systems. Cells from the long-term B-cell cultures have a characteristic lymphoid morphology, whereas those from the Dexter cultures are predominantly granulocytes and macrophages along with a few undifferentiated blast cells. A multilayered stromal cell layer is a common feature of both systems. Scanning electron micrographs show the cells in this layer to be large, irregularly shaped and flattened. The data further indicate that there are unique features in the relationship between developing B cells and stromal cells in the long-term B-cell cultures. Large, mononuclear cells are present that have numerous membrane infoldings within which numerous lymphoid cells lie. The relationship of these cells to macrophages and epithelial/reticular cells is discussed.Supported by the National Institutes of Health Grants AI2125601 (KD) and HD13704 (WHF)     

The stromal colony-forming cell (CFUf) count in the bone marrow of mice and the clonal nature of the fibroblast colonies they form

The clonal nature of FCFC-derived stromal colonies was tested by chromosomal analysis in mixed cultures of CBA and CBAT6T6 bone marrow cells depleted of macrophages and myeloid cells. Inoculation of the bone marrow cell suspensions in flasks coated with poly-l-lysine has revealed practically no stromal aggregates among the explanted cells. The coincidence of karyotypes within the stromal colonies in the mixed cultures proved that the FCFC-derived colonies were cell clones. It was shown by indirect immunofluorescence with antibodies to type 1 collagen that the mouse bone marrow FCFC-derived colonies consisted of stromal fibroblasts. The cloning efficiency of the bone marrow FCFS depends on the explantation density of cells; a stable colony-forming efficiency could be reached only in the presence of feeder cells (irradiated bone marrow). In the bone marrow cells suspensions obtained by trypsinization the amount of FCFC is markedly higher than in the suspensions of mechanically disaggregated bone marrow cells.     

Properties of primary murine stroma induced by macrophage colony-stimulating factor

The ability of purified human macrophage colony-stimulating factor (M-CSF) to accelerate the formation of stromal cells from murine bone marrow cells was investigated. The liquid culture of the marrow cells with M-CSF resulted in the formation of monolayers of macrophages on day 7. When the M-CSF was removed on that day and the residual adherent cells were cultured in the absence of M-CSF for an additional 7 days, many colonies appeared with cells that were morphologically distinguishable from M-CSF-derived macrophages. The appearance of the colonies was dependent on the concentration of M-CSF used at the beginning of the culture. Each colony was isolated as a single clone and analyzed. All clones were negative for esterase staining. These cells did not express M-CSF receptor mRNA and did not show a mitogenic response to M-CSF. On the contrary, these cells could be stimulated to proliferate by fibroblast growth factor and platelet-derived growth factor. The polymerase chain reaction analysis of these cells demonstrated constitutive expression of mRNA for M-CSF, stem cell factor, and interleukin (IL)-1, but not IL-3. Some clones expressed mRNA for granulocyte/M-CSF and IL-6. We also examined the ability of the cells to maintain murine bone marrow high proliferative potential colony-forming cells (HPP-CFC) in a coculture system. Most of the clones showed a significant increase in total HPP-CFC numbers after 2 weeks of coculture, although the extent of stimulation differed among clones. These results suggested that the colonies established by M-CSF were composed of functional stromal cells that were phenotypically different from macrophages. J. Cell. Physiol. 173:1–9, 1997. © 1997 Wiley-Liss, Inc.     

Effect of cytokines on the proliferation/differentiation of stroma-initiating cells

A culture system that identifies the precursor of murine bone marrow fibroblastic stromal cells (stroma-initiating cells, SIC) has been developed. In this system, mature fibroblasts are depleted by adherence to plastic dishes and the nonadherent cells are seeded at a low density, which results in the formation of colonies composed of fibroblastic cells. Macrophage colony-stimulating factor (M-CSF) has been shown to accelerate the colony formation in the system. In this study, we examined the stroma-inducing activity of a number of cytokines. Neither granulocyte-CSF, stem cell factor, interleukin (IL)-1, IL-6, transforming growth factor, epidermal growth factor, insulin-like growth factor, platelet-derived growth factor, nor fibroblast growth factor showed the activity. Similarly, tumor necrosis factor (TNF) did not show any stroma-inducing activity, but the factor inhibited the stromal colony formation induced by M-CSF. In this study, we found that granulocyte/macrophage-CSF (GM-CSF) and IL-3, as well as M-CSF had the stroma-inducing activity. Neither an additive nor synergistic effect was observed when the three factors were assayed in various combinations. The stroma-inducing activity of M-CSF, GM-CSF and IL-3 was observed even if lineage-negative bone marrow cells were used as target cells, suggesting that mature hematopoietic cells such as macrophages and granulocytes were not involved in the induction of stromal colony formation by these factors. Our results raise the possibility that GM-CSF and IL-3 as well as M-CSF stimulate the proliferation or differentiation of the precursor of bone marrow fibroblastic stromal cells.     

Haemopoietic long-term bone marrow cultures from adult mice show osteogenic capacity in vitro on 3–dimensional collagen sponges

Abstract. Adult murine bone marrow cells, cultured under conditions for long-term haemopoietic marrow cultures, produce bone matrix proteins and mineralized tissue in vitro , but only after the adherent stromal cells were loaded on a 3-dimensional collagen sponge. Provided more than 8 × 10⁶ cells are loaded, mineralization as measured by ⁸⁵Sr uptake from the culture medium, occurred in this 3-dimensional configuration (3-D) within 6 days. In contrast if undisrupted marrow fragments (containing more than 10⁷ cells) are placed directly on a collagen sponge, then it requires more than 10 days before significant mineralization can similarly be detected. The 2-dimensional (2-D) long-term marrow culture system allows prior expansion of the stromal cells and some differentiation in an osteogenic direction within the adherent stromal layer. This is suggested by the presence of type I collagen and alkaline phos-phatase positive cells. However, synthesis of osteonectin and a bone specific protein, osteocalcin, as well as calcification are only observed in 3-D cultures. Electron microscopy demonstrated hydroxyapatite mineral on collagen fibres, osteoblast-like cells, fibroblasts, cells which accumulated lipids, and macrophages which were retained on the collagen matrices. Irradiation of confluent long-term bone marrow cultures, prior to their loading on the collagen sponge showed that haemopoietic stem cells are not necessary for the mineralization.