J.CellSd. 84, 139-151 (1986)
139
Printed in Great Britain © The Company of Biologists Limited 1986
KINETICS AND DIFFERENTIATION OF MARROW
STROMAL CELLS IN DIFFUSION CHAMBERS IN VIVO
I. BAB 1 ' 2 , B. A. ASHTON 1 , D. GAZIT 2 , G. MARX 2 ,
M. C. WILLIAMSON 1 AND M. E. OWEN 1
X
MRC Bone Research Laboratory, Nuffield Department of Orthopaedic Surgery, Nuffield
Orthopaedic Centre, Oxford, 0X3 7LD, England
2
Division of Oral Pathology, Hebrew University-Hadassah
POBox 1172, Jerusalem 91010, Israel
School of Dented Medicine,
SUMMARY
Rabbit marrow cells inoculated into diffusion chambers (107 cells/chamber) were implanted
intraperitoneally into athymic mouse hosts and cultured in vivo for 20 days. A connective tissue
consisting of bone, cartilage and fibrous tissues is formed by the stromal fibroblastic cells of marrow
within the chambers. Cell kinetics and tissue differentiation have been studied using histomorphometric and biochemical analyses. Haemopoietic cell numbers decrease to less than 0-05%
of the initial inoculum during the 20-day period. At 3 days an average of 15 stromal fibroblastic cells
only are identifiable within the chambers. After 3 days there is a high rate of stromal cell
proliferation with a doubling time of 14-5 h during the period from 3 to 8 days and an increase in the
total stromal cell population by more than six orders of magnitude from 3 to 20 days. Thirteen to
fourteen population doublings occur before expression of the first observable differentiation
parameter, alkaline phosphatase activity. The data demonstrate that the mixture of connective
tissues formed within the chamber is generated by a small number of cells with high capacity for
proliferation and differentiation. This is consistent with the current hypothesis that stromal stem
cells are present in bone marrow.
INTRODUCTION
Stromal cell lines of marrow include cells of the fibroblastic-reticular network,
adipocytic and osteogenic cells. The hierarchy of these cell lines has not yet been
elucidated and they are often encompassed by the term 'fibroblastic' (Owen, 1985).
Studies of the potential for differentiation of marrow stromal cells were pioneered by
Friedenstein, who demonstrated that either marrow cell suspensions or fibroblastic
cells grown from them in vitro are able to form a calcified tissue in diffusion
chambers (DC) implanted in vivo (Friedenstein et al. 1970; Friedenstein, 1973,
1976). Recent investigations have shown that the bone, cartilage and fibrous tissues
formed by marrow stromal cells within DC are morphologically similar to their
counterparts in the skeleton (Ashton et al. 1980; Bab et al. 19846) and that there is an
increase in alkaline phosphatase activity and accumulation of mineral, parameters
that are indicative of osteogenic differentiation (Bab et al. 1984a; Ashton et al.
1984).
Current evidence for the existence of pluripotent stromal stem cells in the bone
marrow has recently been reviewed (Owen, 1985). Circumstantial evidence supports
Key words: marrow stromal cells, kinetics, diffusion chambers.
140
I. Bab and others
the hypothesis that a single stromal stem cell is able to generate several cell lines
(Friedenstein, 1980). Results from experiments in our laboratory have led us to
suggest that the mixture of fibrous and osteogenic tissues formed by marrow cells
within DC is likely to arise from a small number of cells with stem cell characteristics
(Owen, 1982, 1985; Babe/ al. 1984a).
Formation of osteogenic tissue within DC has many similarities to other bone
developmental systems: namely, rapid cell proliferation followed by production of a
fibrous anlage prior to osteogenesis. In the present experiments stromal cell kinetics
have been studied in relationship to the expression of parameters of differentiation
following implantation of rabbit marrow cell suspensions in DC in athymic mouse
hosts. The results obtained add support to the hypothesis that a small population of
cells with the characteristics of stem cells generate the tissues formed within the
chambers.
MATERIALS AND METHODS
Preparation of cell suspensions
Suspensions of marrow or spleen cells for inoculation into DC were prepared under sterile
conditions as described previously (Ashton et al. 1980, 1984). Male New Zealand White rabbits
weighing 570-630 g were used as cell donors in all experiments. Single cell suspensions of femoral
marrow cells were prepared in Minimal Essential Medium (MEM) by drawing the cells several
times through graded needles. Cell density was determined by counting in a haemocytometer.
Approximately 2X10 8 total marrow cells were obtained from each femur and suspended in 2 ml of
culture medium. Similarly dispersed spleen cells were used as non-osteogenic controls.
In vivo culture in diffusion chambers
DC were assembled from commercially available components (Millipore Corporation).
Chamber dimensions were: 13mm external diameter, 9mm internal diameter, 2mm thick,
approximately 127 mm 3 capacity. The pore size of membrane filters was 0-45 fim. Pairs of DC were
assembled by cementing their lucite rings edge to edge (Fig. 1) and sterilized with ethylene oxide.
Approximately 107 marrow or spleen cells in 100 /il samples were inoculated into the chambers
through a hole in the ring, which was then sealed with a tapered plastic plug coated with glue
(Fig. 1A). One donor rabbit and four host athymic mice (25 g male M F l / n u nu/Ola mice, Olac
1976 Ltd, Oxon., England) were used in each of three experiments with marrow cells and in one
control experiment with spleen cells. A single experiment consisted of 18 DC inoculated with
samples from the same suspension. Sixteen of the DC were implanted into the peritoneal cavity of
hosts, two pairs per mouse (Fig. 1A). The remaining pair of chambers was not implanted and
served as the 0-day controls. DC were harvested at 0, 3, 8, 12 and 20 days after implantation; the
contents of one chamber from each pair was prepared for histology and that of the other for
biochemical analysis (Fig. IB); four chambers were examined by each method at each time point.
Biochemical analysis of DC
The analysis was carried out as described previously (Ashton et al. 1984). The contents of one
DC from each pair (Fig. IB) was homogenized in 10 ml of triple-distilled water. Samples of
the homogenate were assayed for alkaline phosphatase activity and the results expressed as
jimolpNPh" 1 DC" 1 . The remainder of each homogenate was freeze-dried and resuspended in
0-4M-perchloric acid (500/il), and the samples were kept at 4°C for 2h and centrifuged for 15min
at 1800^. The amounts of calcium and phosphorus present in the supernatant were determined
by atomic absorption spectrophotometry and the colorimetric method of Chen et al. (1956),
1
respectively. The results were expressed as
Kinetics of marrow stromal cells
141
Host athymic mouse
Stages of
inoculation of DC
with cell suspension
Implantation
Empty
Full
Planes
of sectioning
Histology
Histomorphometry
I
Biochemical analysis
Alk. phosphatase
Fig. 1. Diagram illustrating: A, experimental procedure for inoculating and implanting
DC. DC-containing marrow cells (marrow DC) were cemented together in pairs,
plugged and implanted intraperitoneally into athymic mice, each pair on either side of the
abdominal mid-line; B, subsequent analysis of paired DC. One DC from each pair is
either processed for histology or the total contents prepared for biochemical analysis.
• • Numbering is arbitrary.
Histological processing of DC
Immediately after removal from the host animal, one DC from each pair (Fig. IB) was immersed
in cold 95% (v/v) ethanol. Dehydration with three changes of absolute ethanol and embedding in
glycol methacrylate were carried out in the cold to preserve alkaline phosphatase activity (Ashton et
al. 1980). Sections, 5/im thick, were cut with a glass-knife microtome (Auto-Cut, Reichert-Jung)
at three parallel, equally spaced planes perpendicular to the surface of the membrane filters
(Fig. 1B). At each plane, two sections were taken for each of the following stains: (1) haematoxylin
and eosin, (2) Von Kossa and Toluidine Blue, and (3) for alkaline phosphatase activity using an azo
dye method (Sigma kit no. 85), counterstained with Von Kossa and Meyer's haematoxylin.
Histomorphometry
The inner DC in a section is the area bounded by the membrane filters and lucite ring.
Measurements of the proportion of this area occupied by tissue and of differentia] counts of intact
cells and of alkaline phosphatase positive cells (see Results for criteria on cell typing) were carried
out using a computerized digitizing system (Bab et al. 1985). The light point tracer of the digitizer
was superimposed on the microscopical image via a drawing apparatus and quantification was
performed at X250 magnification. An average of 280 microscopical fields (34 mm2) was examined
for each DC and the following parameters were calculated according to Weibel (1963): volume of
tissue per DC and number of each cell type per DC.
142
/. Bab and others
The specific activity of alkaline phosphatase was expressed as the ratio between the total activity
per DC estimated biochemically, and the number of alkaline phosphatase positive cells in the
paired DC.
RESULTS
Histology
At day 0 the cell population in DC inoculated with marrow cells (marrow DC)
contains an abundance of rounded haemopoietic cells, 7-25 pm in diameter with a
regular perimeter, together with occasional adipocytes distinguishable by the large
lipid vacuole in their cytoplasm. The haemopoietic and fat cells occur singly or in
small clusters (Fig. 2B).
At 3 days haemopoietic and fat cells still predominate, distributed mainly close to
the Millipore membrane filters. Many of the haemopoietic cells are in different stages
of degeneration. Giant cells with phagocytosed nuclear elements within their
cytoplasm are also present (Fig. 2C). The central portion of the DC is occupied by a
loose fibrin clot.
At 8 days the occurrence of stromal cells other than adipocytes is first noted mainly
in proximity to the membrane filters, admixed with many intact or degenerating
haemopoietic cells (Fig. 2D). These stromal cells can be distinguished from haemopoietic cells by their shape and nuclear morphology. They have either elongated,
often fusiform fibroblastic morphology or are polygonal with at least one angle along
their perimeter that enables them to be distinguished from rounded haemopoietic
cells of a similar size (Fig. 2D). A large regular ovoid nucleus with prominent
nucleoli is also a consistent feature of these cells.
The majority of cells in DC at 12 days are stromal with many polygonal forms
(Fig. 2E, top) and appear slightly larger than those seen at 8 days (Fig. 2D). In
addition, at 12 days many stromal cells reside within large rounded lacunae of slightly
irregular shape in a fibrous matrix (Fig. 2E, bottom). Staining the tissue in DC at 12
days for alkaline phosphatase activity reveals scattered foci of these lacunar cells with
the reaction product localized at their plasma membrane (Fig. 3B).
At 20 days DC contain a mixture of bone, cartilage and fibrous tissues as has been
described in detail (Ashton et al. 1980; Bab et al. 19846). Initially, fibrous tissue is
formed adjacent to the membrane filters and then extends into the interior of the
chamber. Bone and cartilage develop within this fibrous anlage, cartilage mainly
towards the centre of the chamber and bone nearer to the filter (Figs 2F, 3C, 4). With
Toluidine Blue fibrous tissue stains weakly and cartilage strongly metachromatically
(Fig. 2F). Mineral (Von Kossa positive deposits) is seen at 20 days in regions
of developing bone in proximity to the cartilage (Figs 2F, 4). An osteoblast—
preosteoblast layer of cells with high alkaline phosphatase activity is consistently
found at the mineralizing front organized in short palisades; in some areas it is close
to the membrane filter (Fig. 3C) and in others fibrous tissue intervenes (Fig. 2F).
Small clusters of adipocytes can be seen in all chambers examined (Figs 2B,D, 4).
Kinetics of marrow stromal cells
143
5x10*
3
8
12
20
/
\
\
Time of incubation in DC cultures (days)
*
\
\
V-*W
: \i±
SSJ^0' n
t;.*£aa
• o.
Fig. 2. Changes with time after implantation in the cellular composition in marrow DC
inoculated with 107 marrow cells. A. Stromal (#
• ) , haemopoietic (A
A),
adipose (O
O), alkaline phosphatase positive cells (A
A). Each point is the
mean ± S.E.M. of four DC. B-F. Photomicrographs of sections from DC harvested at 0,
3, 8, 12 and 20 days, respectively, after implantation. B,C,E. Stained with haematoxylin
andeosin(H&E); X187. D. Stained with H &E; X120. F. Stained with Toluidine Blue
and Von Kossa; x 187. Membrane filter (tnf), adipocyte (ad), phagocytic giant cell (ph),
osteoblast (ob), cartilage (cr), fibrous tissue (ft), mineral deposits (mn), haemopoietic
cells (double arrows), stromal cells (single arrows), degenerating haemopoietic cells
(arrowheads).
144
/. Bab and others
In DC inoculated with spleen cells (control DC) from 8 days onwards a fibrous
tissue with some haemopoietic cells included in it is found (Fig. 5). This tissue fails
to stain positively for either alkaline phosphatase activity or with Von Kossa.
Morphometry and cells counts
In marrow DC the volume of tissue, consisting of cells and extracellular substance,
shows a highly significant linear increase (r=0-77;P<0-01) throughout the 20-day
experiment (Fig. 6). During this period there is a marked change in the cellular
composition of the tissue (Fig. 2A). Counts of stromal cells are for stromal cells
excluding mature adipocytes; the latter are counted separately. Adipocytic precursors, which are indistinguishable from other stromal precursors by the present
techniques, are included in the stromal cell count. During the first 3 days the total
120 r
0
Time of incubation in DC cultures (days)
Fig. 3. A. Changes with time after implantation in alkaline phosphatase activity in
marrow DC (•—-—•) and control DC (O
O). Each point is the mean ± s.E.M. of
four DC. B.C. Photomicrographs of sections from marrow DC. B, 12 days; X80; C, 20
days; X190. Stained for alkaline phosphatase activity and with Von Kossa. Membrane filter (mf), mineral deposits (mn), alkaline phosphatase positive cells (arrows). •,
-014; ••, P < O 0 0 1 .
Kinetics of marrow stromal cells
145
number of cells decreases to about 0-2% of the inoculated levels. Approximately
90% of those surviving are haemopoietic cells, with somewhat less than 10% being
adipocytic. The average number of stromal cells identified at day 3 in the four
chambers examined is 15 per DC (Fig, 2A), ranging from zero to 58.
Fig. 4. Photomicrograph of section from marrow DC at 20 days. Stained with Toluidine
Blue and Von Kossa; X160. Membrane filter (mf), fibrous tissue (ft), mineral deposits
(mn), cartilage (cr), adipocytes (ad).
Fig. 5. Photomicrograph of section from control DC at 20 days. Stained with H & E;
X 160. Membrane filter (mf), fibrous tissue (ft), haemopoietic cell (arrow).
/. Bab and others
146
100
80
D
40
20
0
3
8
12
20
Time of incubation in DC cultures (days)
Fig. 6. Tissue volume (cells + extracellular substance) in marrow DC plotted against
time after implantation. Each point represents the mean±S.E.M. of 4 DC. r = 0-77;
P<0'01.
From 3 days onwards the stromal cells show exponential growth with a 6 log
increase between day 3 and day 20. Over the same period there is a 1 log decrease in
the number of haemopoietic cells and a 1 log increase in the number of adipocytes
(Fig. 2A). The rate of stromal cell proliferation is very high between 3 and 8 days
with a doubling time of 14-5 h. Then the doubling time increases to 63-6 and 112-7h
for the 8- to 12- and 12- to 20-day periods, respectively (Table 1). The number of
cells added per day to the stromal cell population in the DC increases throughout the
20-day period (Table 1).
In control DC stromal cells grow rapidly and have reached 90% of their maximum
number by 8 days of incubation.
Table 1. Doubling time and growth rate of stromal cells in marrow DC during
different periods of implantation
Doubling
time
(h)
No. of cells added
DC" 1 day" 1
(X105)
3 to 8
14-5
8 to 12
12 to 20
63-6
0-7
1-4
2-8
Implantation
period
(days)
112-7
Kinetics of marrow stromal cells
147
Alkaline phosphatase activity
The total alkaline phosphatase activity measured in marrow DC on days 0, 3 and 8
is similar to that recorded in control DC (Fig. 3A). There are no alkaline phosphatase positive cells in the paired DC taken for histological examination. The number
of alkaline phosphatase positive cells and the total enzyme activity rises from day 8
onwards (Figs 2A, 3A). The two parameters approximately parallel each other up to
20 days (Figs 2A, 3A) so the specific activity of the enzyme does not change from 12
to 20 days (Table 2). The percentage of stromal cells positive for the enzyme also
increases rapidly during this period and constitutes more than 20% of the stromal
cell population by 20 days (Table 2).
Calcium content
The total calcium content of marrow DC shows a significant increase over control
DC only at the 20-day time point (Fig. 7). At this time Von Kossa positive areas are
first seen in the histological preparations. There is a wide variability in the calcium
content of the four DC measured, which is not unexpected at this early stage of
mineralization.
DISCUSSION
The formation of bone, cartilage and fibrous tissue, the increase in alkaline
phosphatase activity and accumulation of calcium within diffusion chambers inoculated with rabbit marrow cells are in agreement with previous results where rabbit
hosts were used (Ashton et al. 1980, 1984; Budenz & Bernard, 1980; Bab et al.
1984a). The use of athymic mice as hosts was based on preliminary experiments that
indicated better chamber-to-chamber reproducibility in these animals compared to
rabbits. This facilitated the present morphometric study where cellular changes prior
Table 2. Alkaline phosphatase activity (AlPase act.) in marrow DC after different
periods of implantation
Implantation
period
(days)
0
3
8
12
20
Total
AlPase act. DC" 1
(/imolpNPh" 1 )
5-49
4-84
2-15
18-68
95-39
±2-69
±3-23
±2-95
±3-051
±21-50§
No. of cells
positive for
AlPase act. DC" 1
(xlO 4 )
0
0
0
8-89 ±2-76
41-71 ±11-34)|
Values are mean ± S.E.M.
• ^ m o l p N P h " 1 10" 4 cells positive for AlPase act.
•(•NA, not applicable.
X Greater than at 0, 3 and 8 days, P < 0-014.
S Greater than at 12 days, P< 0-001.
if G reater than at 12 days, P < 0 • 025.
Specific
activity of
AlPase act.*
% Of stromal
cells positive
for AlPase act.
NAt
0
0
0
9-22 ± 1-53
23-41 ± 11-90
NA
NA
2 •10 ±0-27
2 •80 ±0-32
/. Bab and others
148
0
3
8
12
20
Time of incubation in DC cultures (days)
Fig. 7. Ca content in marrow DC ( •
• ) and control DC (O
O) plotted against
time after implantation. Each point represents the mean ± S.E.M. of 4 DC. # , P< 0-014;
••,P<0-001.
to and during osteogenesis are reported. Direct evidence has been obtained that the
fibrous—osteogenic tissues formed within the chambers are generated by a small
number of stromal cells. Implantation of diffusion chambers containing marrow cells
therefore provides a unique system for studying the differentiation of these tissues in
the adult from early cellular precursors.
There has been no previous study of the kinetics of stromal cell differentiation in
marrow DC culture, whereas there have been many studies (Steinberg & Robinson,
1985) of the kinetics of haemopoietic cells. The latter cannot be compared directly
with the present work for several reasons. Haemopoietic cells recovered from the
chambers were counted and identified morphologically after they had been brought
into suspension by limited proteolysis. Under these conditions it is unlikely that
stromal cells would be identified. The morphometric methods used here enable
necessary additional factors such as cell shape and extracellular matrix characteristics
to be invoked in the recognition of different cell populations and tissue differentiation. In many studies of haemopoiesis (Steinberg & Robinson, 1985) the host
animals were subjected to irradiation or other conditions that promote survival of
haemopoietic stem cells. Furthermore, the cell inoculum used here (107 cells DC" 1 )
is at least 10-fold higher than in studies of haemopoiesis. This may be important
Kinetics of marrow stromal cells
149
since, in a previous study in rabbits (Willemze et al. 1978), the number of haemopoietic cells harvested decreased as the inoculum was increased up to 106 cells per
DC.
In the present work there is a marked decrease in haemopoietic cell number
between implantation and day 3 as was also generally found in studies of haemopoietic differentiation in DC. In rats and mice, however, there was then exponential
growth of the surviving haemopoietic cells (Benestad, 1972; Steinberg & Robinson,
1985), whereas in rabbits (the present study; and Willemze et al. 1978) the initial
decrease is followed first by a modest increase in haemopoietic numbers to day 8,
and then a steady decrease to day 20. Over the same period, from 3 to 20 days, there
is a tremendous growth of the stromal population in our experiments. This is
reminiscent of what happened following transplantation of intact marrow tissue
under the renal capsule, where death of haemopoietic cells occurred in conjunction
with the survival and differentiation of stromal cells into bone and marrow stroma
(Tavassoli & Crosby, 1968; Friedenstein et al. 1978). The diminution of the
haemopoietic component may account for the absence of osteoclastic resorption of
the osteogenic tissue in the DC. It is widely accepted that osteoclasts are a
differentiated form of a haemopoietic cell line (Chambers, 1985; Marks, 1983) and
their absence has been considered a major advantage of the DC method as a system
for studying osteogenesisper se (Ashtonef al. 1980; Budenz & Bernard, 1980).
Marrow adipocytes are a cell line of the stromal system and are, apparently, a
specialized form of fat cells with different hormonal responses to adipocytes from
extramedullary tissues (Green & Meuth, 1974). The majority of mature adipocytes
are probably injected with the initial inoculum and their number underwent only
slight changes in the DC compared to other stromal cells. A close morphological
relationship between mature adipocytes and the mineralized tissue was often seen
and deserves further study.
The data obtained herein demonstrate the presence of a population of stromal cells
in post-natal marrow with a high potential for proliferation and differentiation. The
number of cells with stromal—fibroblastic morphology identified at day 3 in DC
ranged from zero to 58, and suggests that a small number of cells with stem cell
characteristics generates the mixture of fibrous and osteogenic tissues in the
chambers. The doubling time, initially 14-5 h between days 3 and 8, increased
during the subsequent weeks in culture in vivo, probably as a consequence of
increasing cellular differentiation. Given stromal numbers of approximately 50 cells
at day 3, some 13—14 population doublings occur between the onset of stromal cell
proliferation and the initial appearance of alkaline phosphatase positive cells between
days 8 and 12.
It is not possible to decide from the available evidence whether the cells must go
through a set number of divisions before alkaline phosphatase is acquired as a
differentiated cell product, or whether the expression of this cell surface marker is
secondary to other factors such as the local cell density. In vitro, the alkaline
phosphatase activity of osteoblast-like cells (MC 3T3-E1; Sudo et al. 1983; ROS
18/2.8; Majeska & Rodan, 1982) increases once the cells have reached confluence,
ISO
I. Bab and others
although a lower level of enzyme activity was expressed in the proliferative phase
(Rodan & Rodan, 1984). In vivo thymidine-labelling studies have shown that the
alkaline phosphatase-rich preosteoblast population in bone is a proliferating population (Owen, 1971), and epiphyseal chondrocytes in the proliferative zone of the
growth plate display the enzyme on the cell membrane. There is no evidence on the
proliferative capacity of the alkaline phosphatase positive cells in the marrow
diffusion chamber system. However, the continued increase in the stromal cell
population in the second and third weeks in DC culture indicates the persistence of a
proliferating cell pool during the differentiation phase.
When rabbit marrow cell suspensions are cultured in vitro the number of
fibroblastic colony-forming cells (FCFC or CFU-F) per number of cells inoculated
(colony-forming efficiency) is about 30 per 107 marrow cells (Eaglesom et al. 1980;
Ashton et al. 1984). This is of the same order of magnitude as the number of stromal
cells found in DC at 3 days. Although the methods used can give only a crude
estimate of the number of early stromal precursors, this result suggests that colonyforming efficiencies for CFU-F cells in vitro and in DC in vivo are similar.
Circumstantial evidence supports the current hypothesis of a stromal cell system
with stromal stem cells able to generate several cell lines present in bone marrow
(Friedenstein, 1980; Owen, 1980, 1985). The data in this paper are consistent with
the hypothesis, but conclusive proof of it must await the results of further
experiments.
Support from The Wellcome Trust, The Royal Society and The British Council for visits to the
MRC Bone Research Laboratory by I. Bab in the course of this study, is gratefully acknowledged.
Part of this study was presented as partial fulfilment of the requirement for a DMD degree at the
Hebrew University of Jerusalem.
REFERENCES
ASHTON, B. A., ALLEN, T. D., HOWLETT, C. R., EAGLESOM, C. C , HATTORI, A. & OWEN, M.
(1980). Formation of bone and cartilage by marrow stromal cells in diffusion chambers in vivo.
Clin. Orthop. rel. Res. 151, 294-307.
ASHTON, B. A., EAGLESOM, C. C , BAB, I. & OWEN, M. E. (1984). Distribution of fibroblaatic
colony-forming cells in rabbit bone marrow and assay of their osteogenic potential by an in vivo
diffusion chamber method. Calc. Tiss. Int. 36, 83-86.
BAB, I., ASHTON, B. A., SYFTESTAD, G. T. & OWEN, M. E. (1984a). Assessment of an in vivo
diffusion chamber method as a quantitative assay for osteogenesis. Calc. Tiss. Int. 36, 77—82.
BAB, I., GAZIT, D., MASSARAWA, A. & SELA, J. (1985). Removal of tibial marrow induces
increased formation of bone and cartilage in rat mandibular condyle. Calc. Tiss. Int. 37,
551-555.
BAB, I., HOWLETT, C. R., ASHTON, B. A. & OWEN, M. E. (19846). Ultrastructure of bone and
cartilage formed in vivo in diffusion chambers. Clin. Orthop. rel. Res. 187, 243-254.
BENESTAD, H. B. (1972). Cell kinetics in diffusion chambers: survival, resumption of proliferation
and maturation rate of murine haemopoietic cells. Cell Tiss. Kinet. 5, 421-431.
BUDENZ, R. W. & BERNARD, G. W. (1980). Osteogenesis and leukopoiesis within diffusion
chamber implants of isolated bone marrow subpopulations. Am.J. Anat. 159, 455-474.
CHAMBERS, T . J. (1985). The pathobiology of the osteoclast.J. clin. Path. 38, 241-252.
CHEN, P. S. JR, TORIBARA, T . Y. & WARNER, H. (1956). Microdetermination of phosphorus.
Analyt. Chem. 28, 1756-1758.
Kinetics of marrow stromal cells
151
EAGLESOM, C. C , ASHTON, B. A., ALLEN, T. D. & OWEN, M. E. (1980). The osteogenic capacity
of bone marrow cells. Cell Biol. Int. Rep. 4, 742.
FRIEDENSTEIN, A. J. (1973). Determined and inducible osteogenic precursor cells. In Hard Tissue
Growth, Repair and Remineralization, Ciba Fdn Symp. 11, pp. 169-185. North-Holland,
Amsterdam: Elsevier-Excerpta Medica.
FRIEDENSTEIN, A. J. (1976). Precursor cells of mechanocytes. Int. Rev. Cytol. 47, 327-355.
FRIEDENSTEIN, A. J. (1980). Stromal mechanisms of bone marrow: cloning in vitro and
retransplantation in vivo. In Intmunobiology of Bone Marrow Transplantation (ed. S.
Thienfelder), pp. 19-29. Berlin: Springer-Verlag.
FRJEDENSTEIN, A. J., CHAILAKHJAN, R. K. & LALYHNA, K. S. (1970). The development of
fibroblast colonies in monolayer cultures of guinea pig bone marrow and spleen cells. Cell Tiss.
Kinet. 3, 393-402.
FRIEDENSTEIN, A. J.,
IVANOV-SMOLENSKI, A. A.,
CHAILAKHJAN, R. K.,
GORSKEJA, U.
F.,
KURALESOVA, A. I., LATZINIK, N. V. & GERASIMOV, U. F. (1978). Origin of bone marrow
stromal mechanocytes in radiochimeras and heterotopic transplants. Expl Hemat. 6, 440-444.
GREEN, H. & MEUTH, M. (1974). An established pre-adipose cell line and its differentiation in
culture. Cell 3, 127-133.
MAJESKA, R. J. & RODAN, G. A. (1982). The effect of 1,25(OH)2D3 on alkaline phosphatase in
osteoblastic osteosarcoma cells. J. biol. Chem. 257, 3362-3365.
MARKS, S. C. (1983). The origin of osteoclasts. J. Path. 12, 226-256.
OWEN, M. E. (1971). Cell dynamics of bone. In The Biochemistry and Physiology of Bone, vol. I l l
(ed. G. H. Bourne), pp. 271-298. New York, London: Academic Press.
OWEN, M. E. (1980). The origin of bone cells in the postnatal organism. Arthritis Rheum. 23,
1073-1080.
OWEN, M. E. (1982). Bone growth at the cellular level: a perspective. Factors and Mechanisms
Influencing Bone Growth (ed. A. D. Dixon & B. G. Sarnat), pp. 19-28. New York: Alan R.
Liss.
OWEN, M. E. (1985). Lineage of osteogenic cells and their relationship to the stromal system.
In Bone and Mineral Research, vol. 3 (ed. W. A. Peck), pp. 1-25. Amsterdam, New York,
Oxford: Elsevier.
RODAN, G. A. & RODAN, S. B. (1984). Expression of the osteoblastic phenotype. In Bone and
Mineral Research, vol. 2 (ed. W. A. Peck), pp. 244-285. Amsterdam, New York, Oxford:
Elsevier.
STEINBERG, H. N. & ROBINSON, S. H. (1985). Laboratory methods in experimental hematology:
I. The diffusion chamber technique in experimental hematology. Expl Hemat. 13, lit—72..
SUDO, H., KODAMA, H.-A., AMAGAI, Y., YAMAMOTO, S. & KASAI, S. (1983). In
vitro
differentiation and calcification in a new clonal osteogenic cell line derived from newborn mouse
calvaria.J. Cell Biol. 96, 191-198.
TAVASSOLI, M. & CROSBY, W. H. (1968). Transplantation of marrow to extramedullary sites.
Science 161, 54-56.
WEIBEL, E. R. (1963). Principles and methods for the morphometric study of the lungs and other
organs. Lab. Invest. 12(2), 131-155.
WILLEMZE, R., WALKER, R., HERION, J. C. & PALMER, J. G. (1978). Marrow culture in diffusion
chambers in rabbits. I. Effects of mature granulocytes on cell production. Blood 51, 21-31.
{Received 17 February 1986-Accepted 19 March 1986)