After heterotopic (e.g. subcutaneous) transplantation of bone marrow, haemopoiesis in the graft ceases; reticular tissue develops instead, and later bone is formed (Denis, 1958). The result can be achieved by grafting either free pieces of bone marrow or those placed in diffusion chambers (Petrakova, Tolmacheva & Friedenstein, 1963; Rosin, Freiberg & Sajnek, 1963). In the case of free transplantation the bone formed is later filled with bone marrow. After transplantation in diffusion chambers haemopoiesis does not recur despite the development of a considerable mass of bone in the chambers (Friedenstein, 1965).

The population of bone marrow cells is very heterogeneous, including haemopoietic cells, reticular cells and endosteum elements. According to generally accepted views this population is a mixture of individual cell lines capable of mutual transformations within certain limits (Maximov, 1927; Burwell, 1964). After transplantation some of the pathways of differentiation open to bone marrow tissue (formation of reticular and bone tissues) are stimulated, while others (haemopoiesis) are arrested. This effect could be ascribed to the death of haemopoietic stem cells and to proliferation of cells responsible for the development of reticular and bone tissue. It may, however, depend upon transformation of haemopoietic cells to reticular and osteogenic cells.

An analysis of these possibilities encounters difficulties as there are no reliable data concerning the potentialities for transformation of different types of bone marrow cells, and especially of the stem cells. There is no convincing evidence for the dependence of various histogenetic pathways in bone marrow tissue (haemopoiesis, osteogenesis or the development of reticular tissue) upon individual differences of the stem cells concerned. All of them could be provided by one common line of stem cells whose differentiation is controlled by conditions within the population.

Osteogenesis appearing in bone marrow grafts may serve as a model of differentiation in a mixed cell population. Irrespective of whether all or some of the stem cells of bone marrow tissue are able to form bone, the following questions may be raised. In bone marrow transplants either bone and reticular tissue, or reticular tissue alone, develops. Does the difference depend upon whether or not osteogenic stem cells are included in the graft, their behaviour being unaffected by other members of the cell population? Or is differentiation towards osteogenesis a result of an interaction within a community of cells? These questions can be answered when bone marrow cell suspensions are transplanted in diffusion chambers. Under these conditions bone tissue is formed in the chambers. This technique makes it possible to vary the number of transplanted cells and the density of their initial population.

Bone marrow of adult C57BL mice was isotransplanted intraperitoneally to adult recipients in diffusion chambers made of Millipore HA filters (pore size 0·45 μ, thickness 150 μ) or of AUFS filters (pore size 0·6-0·9 μ, thickness 100 μ). The chambers were constructed according to the method of Algire (Algire, Weaver & Prehn, 1957) and were of two sizes: A, chambers with filter diameters of 14 and 10 mm; and B, with those of 7 and 3-2 mm, respectively (cf. Text-fig. 1). Chambers were sterilized in 70° alcohol for 15 min, washed in distilled water and placed into Hanks’s solution.

Text-fig. 1.

Construction of the diffusion chamber. A, Diffusion chamber type A; B, diffusion chamber type B; 1, millipore filter; 2, Plexiglass ring; 3, the cells.

Text-fig. 1.

Construction of the diffusion chamber. A, Diffusion chamber type A; B, diffusion chamber type B; 1, millipore filter; 2, Plexiglass ring; 3, the cells.

Bone marrow was extracted from femur and chopped into fragments of about 2 mm, which were placed into diffusion chambers, or bone marrow cell suspension was prepared in Hanks’s solution. After elimination of cell clumps by filtration through Capron net the suspension was diluted to 2 × 106 cells per ml.

To prepare lymphocyte suspensions cervical lymph nodes were used. After filtration the concentration of lymphocytes was adjusted to 107 per ml. To introduce cells into a chamber the larger filter was put on glass rails over a hollow-ground slide into which Hanks’s solution was poured until it touched the filter. A given volume of cell suspension was placed on the filter from above, the liquid passing across the filter and the cells precipitating on it. Filtration was performed in a Petri dish lined with cotton-wool soaked in saline. Then a filter of smaller diameter (without cells) was placed over the larger one (with cells) and the chamber was stuck. The number of viable cells in the remaining cell suspension was counted to determine that of the cells placed into chambers.

The chambers were fixed with 96° alcohol between 1 and 30 days after transplantation. They were freed from the surrounding tissue, the filters were separated and put into a cooled fixative. In most cases the Gomori reaction for alkaline phosphatase was performed on filters (Gomori, 1939) to reveal foci of osteogenesis. Filters then were counterstained with haematoxylin, dehydrated with alcohol, cleared with xylene and mounted in balsam as total preparations. Some filters were tested for calcium as the control to the Gomori reaction. After freeing from the plexiglass rings and fixation some chambers were embedded in paraffin and cut in serial sections that were stained by the Gomori method, for calcium, with haematoxylin-eosin, for PAS (counterstained with haematoxylin).

The following experiments were performed :

  1. Transplantation of a piece of bone marrow into chambers of types A and B.

  2. Transplantation of a suspension of 106 bone marrow cells into chambers of type A.

  3. Transplantation of a suspension of 105 bone marrow cells into chambers of type A.

  4. Transplantation of a mixture of bone marrow cells and lymphocytes (1:9), the total number of cells being 107, into chambers of type A.

  5. Transplantation of a suspension of 105 bone marrow cells into chambers of type B.

  6. Implantation of empty chambers.

In series 2 and 5 the initial density of nucleated cells on the filter was approximately 210 and in series 3 approximately 3 per 0·05 mm2.

(1) Morphology of transplanted pieces of bone marrow (HA chambers’)

When pieces of bone marrow were placed into chambers small blood vessels and fragments of spongy bone were included with them (Plate 1, fig. 1). Using the Gomori method both these tissues are clearly stained. The residual bone marrow tissue of mice showed a negative reaction for alkaline phosphatase. During the first days after transplantation the trabeculae of the bone necrosed. They ceased to show reaction for phosphatase, the osteocytes perished and empty cellular cavities remained in the bone. The surface of bone became denuded losing its covering layer of osteoblasts (Plate 1, fig. 2). At the same time a large number of cells migrated from the graft and a broad zone of haemopoietic and reticular cells arose in the chamber. The relative amount of the former rapidly declined, while that of the latter rose.

Plate 1

Fig. 1. Transplantation of a piece of bone marrow in a chamber with HA filters. Three days. Alcohol. Gomori. Total preparation. × 200. Blood vessels chamber with bone marrow.

Fig. 2. The same preparation. × 200. Fragment of dead bone in the chamber.

Figs. 3,4. Transplantation of a piece of bone marrow in a chamber with HA filters. Three days. Alcohol. Haematoxylin. Total preparation. × 400. Areas of the zone of growth. M, mitosis.

Fig. 5. Transplantation of a piece of bone marrow in a chamber with HA filters. Three days. Alcohol. Gomori. Total preparation. × 400. A focus of phosphatase activity in the zone of growth.

Fig. 6. Transplantation of a piece of bone marrow in a chamber with HA filters. Five days. Alcohol. Gomori. Total preparation. × 400. A focus of phosphatase activity in the zone of growth.

Plate 1

Fig. 1. Transplantation of a piece of bone marrow in a chamber with HA filters. Three days. Alcohol. Gomori. Total preparation. × 200. Blood vessels chamber with bone marrow.

Fig. 2. The same preparation. × 200. Fragment of dead bone in the chamber.

Figs. 3,4. Transplantation of a piece of bone marrow in a chamber with HA filters. Three days. Alcohol. Haematoxylin. Total preparation. × 400. Areas of the zone of growth. M, mitosis.

Fig. 5. Transplantation of a piece of bone marrow in a chamber with HA filters. Three days. Alcohol. Gomori. Total preparation. × 400. A focus of phosphatase activity in the zone of growth.

Fig. 6. Transplantation of a piece of bone marrow in a chamber with HA filters. Five days. Alcohol. Gomori. Total preparation. × 400. A focus of phosphatase activity in the zone of growth.

On the third day the filters were completely covered by cells usually arranged in several layers in total preparations (Plate 1, fig. 3). The majority of these cells were fibroblast-like elements, many in mitosis. Large spindle cells with dense cytoplasm, large nuclei and clear-cut nucleoli could be easily distinguished among them; in these cells mitoses were seen particularly often (Plate 1, fig. 4). Within such tissue osteogenic foci appeared on the third day. When tested by the Gomori reaction these foci looked like a phosphatase-positive network composed of elongated reticular cells with phosphatase-positive cytoplasm in a small amount of phosphatase-positive matrix (Plate 1, fig. 5). The osteogenic cells were bigger than the fibroblast-like cells on the filter. Meshes in the network were formed by one or, less frequently, two layers of adjacent cells between which phosphatase-negative fibroblast-like and haemopoietic cells were distributed, i.e. elements that covered the remaining area of the filter.

Foci of osteogenic tissue were easily distinguished from other phosphatasepositive structures that could occur in these preparations, such as fragments of marrow blood vessels and the remnants of dead bone. The vessels possessed a distinct wall, their branching tubes were of regular shape with their contours clearly outlined. The fragments of the spongy bone that got into the chamber during implantation lost all signs of viability by the third day. It is significant that in none of the cases investigated did foci of developing osteogenesis touch the fragments of old bone. These two were always observed at a distance from each other; fragments of necrotic bone were located within the initial graft, while foci of osteogenesis were found in the zone of outgrowth on the filter.

Between the third and the eighth day after transplantation the number of bone foci did not increase markedly but their structure changed : the amount of bone matrix rose, trabeculae became thicker and composed of a greater number of cells (Plate 1, fig. 6). The alveoli between trabeculae narrowed and the entire structure became more compact. The osteogenic foci increased in size; on the eleventh day they were quite distinct even when stained with haematoxylin only. They showed the structure of typical bone trabeculae with osteoblasts embedded in the bone matrix. In sections this bone tissue has also a peculiar appearance (Plate 2, fig. 7). By the twenty-fourth day bone occupied most of the chamber (Plate 2, figs. 8-10).

Plate 2

Fig. 7. Transplantation of a piece of bone marrow in a chamber with HA filters. Eleven days. Alcohol. PAS-haematoxylin. A section. × 200. A focus of osteogenesis in the chamber under the filter, a, Filter; b, bone tissue; c, a layer of osteoblasts.

Figs. 8, 9. Transplantation of a piece of bone marrow in a chamber with HA filters. Fifteen days. Alcohol. PAS-haematoxylin. × 400. Bone tissue adjoining the filter, a, Filter, b, bone tissue.

Fig. 10. Transplantation of a piece of bone marrow in a chamber with HA filters. Alcohol. Gomori. Section. × 200. Bone in the chamber, a, Filter, b, bone tissue.

Plate 2

Fig. 7. Transplantation of a piece of bone marrow in a chamber with HA filters. Eleven days. Alcohol. PAS-haematoxylin. A section. × 200. A focus of osteogenesis in the chamber under the filter, a, Filter; b, bone tissue; c, a layer of osteoblasts.

Figs. 8, 9. Transplantation of a piece of bone marrow in a chamber with HA filters. Fifteen days. Alcohol. PAS-haematoxylin. × 400. Bone tissue adjoining the filter, a, Filter, b, bone tissue.

Fig. 10. Transplantation of a piece of bone marrow in a chamber with HA filters. Alcohol. Gomori. Section. × 200. Bone in the chamber, a, Filter, b, bone tissue.

Table 1 presents the results of transplantation of pieces of bone marrow. At the time of transplantation such pieces consisted on the average of 8 × 106 cells.

Table 1.

Bone formation in chambers containing a piece of bone marrow

Bone formation in chambers containing a piece of bone marrow
Bone formation in chambers containing a piece of bone marrow

(2) Morphology of bone marrow transplanted in form of suspension of 106 cells into the chamber of type A (HA chambers’)

When a suspension of bone marrow cells was placed in a chamber the initial population consisted of: haemocytoblasts (11 %), myeloids (33 %), leucocytes (10 %), erythroids (8 %), monocytes (4 %), lymphocytes (34 %), reticular cells (0·5 %), only nucleated cells being taken into account. The cells were evenly distributed over the whole surface of the filters; no phosphatase-positive cells could be found.

On the second and third days (Plate 3, fig. 11) the filters were covered with a continuous layer of fibroblast-like cells and of haemopoietic elements, the majority of the latter showing signs of degeneration. Apart from this, solitary delimited foci consisting entirely of young haemopoietic cells of the myeloblast type occurred on filters on the third day (Plate 3, fig. 12). Cell counts in four such foci gave the following values, respectively, 81, 50, 72, 49 (mean = 63), and 4, 0, 2 and 5 mitoses, respectively, were seen in them. Each focus consisted entirely of cells of the same types. Large isolated cells with basophilic cytoplasm and large nuclei occurred among fibroblasts (Plate 3, fig. 13). The reaction for phosphatase remained negative in them all.

Plate 3

Fig. 11. Transplantation of 106 bone marrow cells in a chamber of type A with HA filters. Two days. Alcohol. Haematoxylin. Total preparation. × 400.

Fig. 12. The same experiment. Three days. Alcohol. Haematoxylin. Total preparation. × 200. A focus of 49 myeloid cells with five mitoses.

Fig. 13. The same experiment. Three days. × 200. Large basophilic cells. Mitosis in one.

Figs. 14-16. The same experiment. Five days. Alcohol. Haematoxylin. Total preparations. × 280. Bands of large polygonal cells.

Plate 3

Fig. 11. Transplantation of 106 bone marrow cells in a chamber of type A with HA filters. Two days. Alcohol. Haematoxylin. Total preparation. × 400.

Fig. 12. The same experiment. Three days. Alcohol. Haematoxylin. Total preparation. × 200. A focus of 49 myeloid cells with five mitoses.

Fig. 13. The same experiment. Three days. × 200. Large basophilic cells. Mitosis in one.

Figs. 14-16. The same experiment. Five days. Alcohol. Haematoxylin. Total preparations. × 280. Bands of large polygonal cells.

On the fifth day the composition of the cell population covering the filters underwent a change. The number of degenerating haemopoietic cells sharply declined, most of them being spindle-shaped and fibroblast like. Large polygonal cells with round oval nuclei and compact dense cytoplasm developed among them. These cells formed fine bands with close packing (Plate 3, fig. 14); mitoses were common (Plate 3, figs. 15,16) and they were positive for alkaline phosphates (Plate 4, figs. 17, 18) which distinguished them from the remaining tissue on the filters. Apart from such phosphatase-positive foci, large foci of myeloid cells (Plate 4, fig. 19) occurred on the filters on a network of small fibroblast-like cells connected by fine processes. Later, 7-12 days after transplantation, the bands of sharply phosphatase-positive cells became of typical, round, osteogenic foci (Plate 5, figs. 21, 22). Bone matrix appeared in their vicinity, the cells revealing characteristic features of osteoblasts (Plate 4, fig. 20). The rest of the filters were covered with fibroblasts. No myeloid foci could be revealed at this time; however, single haemopoietic cells might occur on filters often lying close to bone trabeculae (Plate 5, fig. 23). In sections bone foci had the structure typical of bone tissue (Plate 5, fig. 24).

Plate 4

Figs. 17, 18. The same experiment. Five days. Alcohol. Gomori. Total preparations. × 400.

Fig. 19. The same experiment. Five days. Alcohol. Haematoxylin. Total preparation. × 400. A focus of myeloid cells with mitoses.

Fig. 20. The same experiment. Seven days. Alcohol. Haematoxylin. Total preparation. × 400. A focus of osteoblasts with mitoses.

Plate 4

Figs. 17, 18. The same experiment. Five days. Alcohol. Gomori. Total preparations. × 400.

Fig. 19. The same experiment. Five days. Alcohol. Haematoxylin. Total preparation. × 400. A focus of myeloid cells with mitoses.

Fig. 20. The same experiment. Seven days. Alcohol. Haematoxylin. Total preparation. × 400. A focus of osteoblasts with mitoses.

Plate 5

Fig. 21. The same experiment. Seven days. Alcohol. Gomori. Total preparation. × 200. A focus of osteogenesis.

Fig. 22. The same experiment. Ten days. Alcohol. Gomori. Total preparation. × 200. A focus of osteogenesis.

Fig. 23. The same experiment. Eleven days. Alcohol. Haematoxylin. Total preparation. × 200. Bone trabeculae and haemopoietic cells.

Fig. 24. The same experiments. Twelve days. Alcohol. PAS. A section. × 200. Bone in the chamber, a, Filter, b, bone tissue.

Fig. 25. An empty chamber composed of AUFS chambers. Five days after implantation. Alcohol. Haematoxylin. Section. × 200. a, Filter; b, chamber contents.

Plate 5

Fig. 21. The same experiment. Seven days. Alcohol. Gomori. Total preparation. × 200. A focus of osteogenesis.

Fig. 22. The same experiment. Ten days. Alcohol. Gomori. Total preparation. × 200. A focus of osteogenesis.

Fig. 23. The same experiment. Eleven days. Alcohol. Haematoxylin. Total preparation. × 200. Bone trabeculae and haemopoietic cells.

Fig. 24. The same experiments. Twelve days. Alcohol. PAS. A section. × 200. Bone in the chamber, a, Filter, b, bone tissue.

Fig. 25. An empty chamber composed of AUFS chambers. Five days after implantation. Alcohol. Haematoxylin. Section. × 200. a, Filter; b, chamber contents.

(3) The effect of the number of cells and of their packing upon bone formation

Different numbers of cells were placed in chambers of different size and of each type. The results in terms of the proportion of chambers in which bone was formed are given in Table 2. The remaining chambers were filled with reticular tissue.

Table 2.

Bone formation in chambers with suspension of bone marrow cells

Bone formation in chambers with suspension of bone marrow cells
Bone formation in chambers with suspension of bone marrow cells

(4) Implantation of empty chambers

Empty chambers composed of HA filters sterilized for 15 min with 70° alcohol before intraperitoneal implantation proved to be impermeable to cells : in none of five empty chambers were cells found 7 days after implantation. No cells were detected in the filter material itself in sections of the many chambers composed of HA filters which were studied.

Chambers composed of AUFS filters were permeable to cells The average numbers of cells found after implantation of an empty chamber with 0·04 mm2 of filter surface were: 5·0 cells after 2 h; 18·5 cells after 6 h; 14·5 cells after 12 h; 24·5 cells after 24 h, and 50·0 cells after 48 h. During the first 5 h leucocytes predominated among the cells, later lymphocytes and fibroblast-like cells or histiocytes were predominant. Many cells were always found in the filter material in sections of chambers composed of AUFS filters (Plate 5, fig. 25).

When bone marrow cells are placed into diffusion chambers reticular or bone tissue is formed instead of haemopoietic elements. This considerable change of histogenesis occurs after free heterotopic transplantation of bone marrow as well (Denis, 1958). Therefore, it does not depend upon special cultivation conditions in diffusion chambers. The cells transplanted in a chamber are in novel conditions by comparison with those in situ. Their interaction with surrounding tissues (with bone in particular) and their mutual arrangement (tissue structure) are disturbed. In the case of transplantation of bone marrow fragments this holds true for the zone of cell outgrowth, while the original transplant itself usually degenerates. These disturbances seem to cause the main alteration in differentiation of transplanted cells, i.e. the extinction of haemopoiesis. Which of the causes mentioned is decisive remains unknown.

The fact that osteogenesis occurs in cell suspension shows that the osteogenic potency of the bone marrow cell population does not disappear after dissociation of cells.

It is evident that not all the transplanted bone marrow cells can act as osteogenic stem-cells. The results of the present work suggest that in a population of bone marrow cells cultured in diffusion chambers no cells form osteogenic foci individually. If this were otherwise, bone would form to the same extent in the chambers of different size after transplantation of the same number of cells (i.e. containing the same number of such precursor elements). However, the same number of bone marrow cells, say 105, behaves in a different way in the chambers of different size. In those of type B this number of cells formed bone, as a rule, while in the chambers with an area tenfold larger (type A) it did not. The area of the chamber was determined by the area of the smaller filter. Bone formation in larger chambers (type A) requires more cells, namely 106, i.e. ten times more. A tenfold dilution of 106 bone marrow cells placed in type A chamber (by the addition of 9 × 106 lymphocytes taken from lymph nodes) prevents osteogenesis. These data cannot be explained by the fact that general metabolic processes (e.g. those of metabolism, respiration, etc.) could themselves inhibit or stimulate bone formation in the chambers containing a different number of cells. When 106-108 bone marrow cells were placed into A-chambers osteogenesis usually occurred (Friedenstein, 1964). Yet, when 106 bone marrow cells + 9 × 106 lymphocytes were placed in such chambers bone was rarely formed. A lesser dilution of the bone marrow cells does not affect the result of transplantation : penetration of a small number of cells from outside into the chamber (when the chambers consisted of AUFS filters) did not change the frequency of osteogenesis when compared with the chambers composed of HA filters impermeable to cells.

The cells responsible for the formation of osteogenic foci seem to have a high mitotic activity. This is proved by the frequent occurrence of mitosis in the cells which, judging by morphological characters, are those which give rise to osteogenic structures in chambers.

It could be expected, therefore, that cells serving as precursors for osteogenic tissue at transplantation of 105 bone marrow cells in type B chambers can form considerable cell clones during 30-day cultivation when placed in type A chambers. Thus, if the population density of 105 cells in type A chambers is insufficient for osteogenesis, the cells of which bone foci could be formed either do not multiply or they proliferate, but differentiation of the corresponding clone does not proceed towards osteogenesis. This latter possibility seems to be most likely. At any rate, the results presented demonstrate that osteogenesis requires a certain initial density of bone marrow cells. Thus, the development of bone in a chamber differs from the formation of haemopoietic foci on the spleen of irradiated mice at transplantation of bone marrow cells (Till, McCulloch & Siminovitch, 1964; Becker, McCulloch & Till, 1963; Lewis & Trobaugh, 1964). A haemopoietic stem-cell, if in contact with an appropriate stroma (e.g. in irradiated spleen) is able to create a haemopoietic clone in isolation (Becker et al. 1963). On the other hand, when bone is formed in a chamber the differentiation of osteogenic stem-cells depends upon the interaction of cells at some initial period. It is characteristic that duration does not affect the results of transplantation, i.e. if the initial packing does not result in bone formation during eleven days, no osteogenesis develops in 30 days either.

As is well known, bone marrow cells are mobile and are expected to migrate within the chamber after transplantation. The fact that osteogenesis requires a certain density of the initial population of bone marrow cells implies that osteogenesis arises when necessary cells meet to form a corresponding structure. Apparently it is easily established when the initial density corresponds to 210 bone marrow cells per 0·05 mm2 being achieved but very rarely when the density equals thirty bone marrow cells per 0·05 mm2.

In the case of transplantation of a piece of marrow, even if it is sufficiently small to avoid rapid necrosis, bone formation proceeds only in the zone of outgrowing cells. This implies that the initial structure of marrow is inappropriate for bone formation, and the structure required for osteogenesis is created anew. Foci of proliferating haemopoietic cells are found on the third to seventh day in chambers with cell suspension. It seems very likely that these are clones originated from individual haemopoietic stem-cells similar to haemopoietic foci in the spleen of irradiated mice (Till et al. 1964). In chambers these foci disappear on the tenth day probably due to the fact that haemopoiesis proceeds only in the presence of a certain structure of haemopoietic tissue whose maintenance requires contact with either bone or spleen stroma (Friedenstein, 1965).

In the presence of bone (or spleen stroma) the stem cells of haemopoietic tissue differentiate towards haemopoiesis (Till et al. 1964). In the absence of bone, osteogenic potencies of marrow cell population are realized and bone is formed which in turn is necessary for the maintenance of haemopoiesis. This creates the mechanism supporting differentiation of haemopoietic tissue and controlling the behaviour of stem-cells in the bone marrow cell population. It should be borne in mind that bone tissue undergoes sustained reconstruction, i.e. it is resorbed in some places while being formed in others (Hattner, Epker & Frost, 1965). For this reason its influence upon the differentiation of stem cells can be expected in any individual area of marrow tissue.

There are grounds for believing that the formation of cells of marrow stroma and of blood cells is ensured by common stem cells, but no direct evidence is available. However, haemopoietic and stromal elements might have different stem cells while control over the composition of the whole population of marrow cells provides selective effects upon each of these categories of stem cells. At the present time it cannot be ruled out that bone and reticular tissue too have separate lines of stem cells in bone marrow, although it seems more likely that they have a common line.

  1. In bone marrow fragments and bone marrow cell suspensions isotransplanted to mice in diffusion chambers, reticular tissue develops and sometimes osteogenesis also occurs. No haemopoiesis takes place in the grafts.

  2. Bone formation in the chambers requires a certain density of the initial packing of bone marrow cells. This follows from the considerable differences in the frequency of bone formation in the case of transplantation of the same number of cells (105) in chambers of different size, and from cases in which different cell numbers (105 and 106) are transplanted in the chambers of the same size.

  3. The results obtained show that differentiation of the stem cells towards osteogenesis requires cell interaction within the cell community at a certain crucial moment.

The authors would like to express their sincere gratitude to Professor G. V. Lopashov for his valuable advice and interest in this paper.

Algire
,
G. A.
,
Weaver
,
J. M.
&
Prehn
,
K. T.
(
1957
).
Studies on tissue homotransplantation in mice using diffusion chamber method
.
Ann. N.Y. Acad. Sci
.
64
,
1009
12
.
Becker
,
A. J.
,
McCulloch
,
E. A.
&
Till
,
J. E.
(
1963
).
Cytological demonstration of the clonal nature of spleen colonies derived from transplanted mouse marrow cells
.
Nature, Lond
.
197
,
452
4
.
Burwell
,
R.
(
1964
).
Studies on the transplantation of bone
.
J. Bone Jt Surg
.
46 B
,
110
40
.
Denis
,
A.
(
1958
).
Etude de l’ossification dans les graftes de moelle osseuse. Bruxelles
.
Friedenstein
,
A. J.
(
1964
).
In Press (Russian)
.
Friedenstein
,
A. J.
(
1965
).
Factors influencing differentiation of hematopoietic tissue in diffusion chambers. Molecular and cellular basis of antibody formation
.
Praha
,
1965
,
321
-
8
.
Gomori
,
G.
(
1939
).
Microtechnical demonstration of phosphatase in tissue sections
.
Proc. Soc. exp. biol. Med
.
42
,
23
6
.
Hattner
,
R.
,
Epker
,
B.
&
Frost
,
H.
(
1965
).
Suggested sequential mode of control of changes in cell behaviour in adult bone remodelling
.
Nature, Lond
.
206
,
489
90
.
Lewis
,
J. P.
&
Trobaugh
,
F. E.
(
1964
).
Haemopoietic stem-cells
.
Nature, Lond
.
204
,
589
90
.
Maximov
,
A. A.
(
1927
).
Bindgewebe und blutbildende Gewebe
.
Handb. mikr. Anat. Menschen
.
2
,
232
426
.
Petrakova
,
K. V.
,
Tolmacheva
,
A. A.
&
Friedenstein
,
A. J.
(
1963
).
Bone formation occurring in bone marrow transplantation in diffusion chambers
.
Bull. exp. biol. med. (Russian)
,
56
(
12
),
87
91
.
Rosin
,
A.
,
Freiberg
,
H.
&
Sajnek
G.
(
1963
).
The fate of bone marrow, spleen and periosteum cultivated in vivo in the diffusion chamber with special reference to bone formation
.
Exp. Cell Res
.
29
,
176
87
.
Till
,
J. E.
,
McCulloch
,
E. A.
&
Siminovitch
,
L.
(
1964
).
A stochastic model of stem-cell proliferation based on the growth of spleen colony-forming cells
.
Proc. natn. Acad. Sci. U.S.A
.
51
,
29
36
.