Cell Tissue Res (1992) 268:17-30
Cell&Tissue Research 9 Springer-Verlag 1992
Multinucleated giant cells in primary cultures derived from canine bone marrow - evidence for formation of putative osteoclasts M.C. Bird 1, D. Garside
2,
and H.B. Jones 1
i Department of Pathology and z Department of Toxicology, Smith Kline and French Research Ltd., The Frythe, Welwyn, Hertfordshire AL6 9AR, UK Received June 17, 1991 / Accepted November 2, 1991
Summary. Mononucleated cells derived from canine bone marrow were maintained in vitro for up to 6 weeks. The culture characteristics and development of these cells were evaluated by histological, ultrastructural and histochemical methods. Within I week the cells had fused together to form flattened, multinucleated cells. Further fusing with one another and other mononucleated cells produced large (diameters more than 300 gin), multinucleated cells which frequently contained more than 50 nuclei per cell and exhibited ultrastructural and histochemical features that were strikingly similar to those displayed by osteoclasts. The confluent monolayer of mono- and multinucleated cells present at 4 weeks had, by the sixth week, become altered such that fibroblast overgrowth obliterated all other cells. During the development of the culture adipocytes became differentiated from mononuclear cells and frequently were located within spherical multicellular aggregates (spheroids). Functional assessments were employed to investigate whether the multinucleated ceils generated in this way, represented osteoclast-like cells, or alternatively, were related to macrophage polykarya as found in foreign body granulomata in vivo. Neither resorption pits on sperm whale dentine slivers (diagnostic of osteoclasts), nor formation of granulomata in vitro, were observed. We believe that the present results indicate that the multinucleated cells generated from canine bone marrow mononuclear precursors in vitro, merit designation as osteoclast-like cells. Definitive characterisation however, must await further functional assessments of hormone responsiveness.
The comparatively recent establishment of long-term tissue culture systems derived from canine bone marrow has made possible the study of component parts of important, but ill-understood, phenomena such as osteoclast formation (Allen et al. 1981 ; Severson 1983), bone resorption (Allen et al. 1981) and cellular responses to trophic factors (Suda et al. 1983). To date, these techniques have principally been used for investigations in two areas, namely, for the study of haemopoietic cell lines in a controlled in vitro environment, and for the study of the origin, function and fate of the bone-modelling osteoblasts and osteoclasts. The influence of imbalance in the complex interrelationships between the cells, and the effects of functional impairment by physiological or toxicological perturbations are important to an understanding of the pathophysiological processes associated with bone disease, e.g. osteoporosis and osteopetrosis. In the present study, an in vitro tissue culturing technique has been used to investigate the hitherto unrecorded establishment, maintenance and cellular characteristics of developing multinucleated cells derived from canine bone marrow. In addition, an assessment of the functional capability of the cultures has been made to throw some light on bone resorptive capacity and granuloma formation in vitro. It was felt that such a study was necessary to distinguish between osteoclasts and multinucleate giant cells (macrophage polykarya) present in vivo in inflammatory cell aggregates associated with poorly degradable particles (Kasahara et al. 1988; Kreipe et al. 1988; Walters and Schneider 1987).
Key words: Bone marrow - Osteoclasts - Giant cells - Macrophage polykarya Tissue c u l t u r e - Dog
Materials and methods Culture preparation and establishment
* Dr. Martin Bird died during the final phase of this work and it is to his memory that this paper is respectfully dedicated Offprint requests to: H.B. Jones, Safety of Medicines Department, ICI Pharmaceuticals, Alderley Park, Macclesfield, Cheshire, SK10 4TG, UK
Beagle dogs (employed as controls in safety evaluation studies of novel pharmaceutical agents) were killed by an intravenous injection of 150 mg/kg sodimn pentobarbitone (May and Baker, Dagenham, UK). Femurs were excised and, after cutting the epiphyses, the bone marrow was flushed out with Iscove's modified Dulbecco's medium (Gibco BRL, Paisley, UK) containing 20% isologous
18 dog serum, 10 -6 M hydrocortisone (Sigma, Poole, UK) and 50 gg/ ml gentamycin (growth medium). Dispersed cell suspensions were prepared by gentle trituration and pipetting onto 0-90% Percoll gradients, then centrifuged at approximately 500 g for 5 min, in order to remove all adipocytes and red blood cells. After washing, cultures were established by inoculating 2-5 x 107 nucleated cells in 10 ml of growth medium into sterile, 25 cm 2 (growth area) tissue culture flasks (Gibco Nunc), and incubated at 37~ C in 5% CO2/ 95% air. At weekly intervals they were fed by replacing half the spent growth medium (demidepopulation). Cells in suspension, removed together with the growth medium, were not replaced. Prior to microscopic examination, cultures were washed once with sterile saline (0.9% sodium chloride) and examined by phasecontrast microscopy using a Leitz Labovert inverted phase-contrast microscope. For examination and counting of nuclei in multinucleated cells, cells attached to the surface of the culture flask were washed once with saline, fixed with neutral buffered formalin and stained with haematoxylin and eosin. The numbers of nuclei in a total of 100 multinucleated cells, in randomly chosen fields, were counted in 7 and 21 day cultures. Three flasks were assessed in this way at both times.
Cytochemistry Monolayer cultures, sampled at intervals up to 21 days, were fixed by immersion in 10% neutral buffered formalin for 5 min, and processed cytochemically by standard methods (Hammarstrom et al. 1971 ; Pearse 1972; Bancroft and Stevens 1977) for the demonstration of the following tissue components: neutral lipid (oil Red O), calcium deposits (Alizarin Red S), and alkaline and acid phosphatases (tartrate resistant and unresistant). Counterstaining with haematoxylin was performed as appropriate.
Functional characteristics
Electron microscopy In order to determine the morphological characteristics of bone marrow mononucleated cells prior to culture establishment, cell suspensions were prepared as described above and depleted of erythrocytes by lysis with 0.8% (w/v) aqueous ammonium chloride. Their morphology was compared with that of canine peripheral blood mononucteated cells prepared by fractionation of whole blood over Ficoll Hypaque (sg of 1.077, Boyum 1983). These specimens, and established cultures (following decantation of the culture medium) sampled at 11, 28 and 42 days, were fixed by immersion in a solution of 4% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) for 2 h, then twice washed in the same buffer for 10 min. Secondary fixation by 1% osmium tetroxide in the same buffer for 1 h preceded separation of the samples into those designated for transmission (TEM) or scanning electron microscopy (SEM). For TEM, samples were dehydrated in an ascending ethanol series and 1, 2-epoxypropane. At this stage, cells adherent to the culture flask were carefully detached by means of a clean pith stick prior to infiltration with, and embedding in, Taab resin (Taab Equipment Laboratories Ltd., Aldermaston, UK). Semithin (1 Ixm thick), toluidine-blue-stained sections were examined by light microscopy (LM) in order to evaluate cell characteristics and selection of tissue areas for subsequent ultramicrotomy. Ultrathin sections (70 90 nm thick) were stained with uranyl acetate and lead citrate solutions, and examined in a JEOL, JEM 1200EX transmission electron microscope at an accelerating voltage of 80 kV. For SEM, a modification of the method of Wollweber et al. (1981) was employed. The osmicated samples were further immersed in a solution of 1% tannic acid in 0.1 M cacodylate buffer (pH 7.4) for 30 min and, after several washes in distilled water, placed into 1% aqueous uranyl acetate solution for the same length of time. Following dehydration in an ascending ethanol series, the specimens of cells adherent to substrata were critical-point dried and mounted on aluminium support stubs with silver paint (Agar Scientific Ltd, Stanstead, UK). All specimens were coated with a thin (8-10 nm thick) layer of gold prior to examination in a JEOL, JSM 840 scanning electron microscope at an accelerating voltage of 8 kV.
In order to elucidate multinucleated cell identity as either osteoclast or macrophage polykarya (multinucleated giant cells), 3 functional criteria were used for assessment: (1) Seven- or 10-day-old established cultures were maintained for up to 21 days with slivers (100 gm thick) of sperm whale dentine (Kindly supplied by Professor A. Boyde and Dr. S. Jones, Department of Anatomy and Embryology, University College, London). At the end of this period, adherent cells were removed from the slivers by gently rubbing the surface and, following light counterstaining with toluidine blue, the slivers were examined by light microscopy for the presence of resorption pits, a diagnostic feature of osteoclast activity. A few slivers were also prepared for examination by SEM. (2) Following the first demidepopulation, 7-day-old cultures were maintained for 14 days in medium to which fine, particulate fragments of dog bone (femur) were added. The particles were prepared from absolute-ethanol-sterilised pieces of femur by scraping the fractured end of the central portion of the diaphysis with a clean scalpel blade. These were then stored in ethanol until required. Prior to use, the ethanol was first removed by several changes of sterile isotonic saline. (3) Seven- or 14-day-old cultures were incubated for 14 days with fragments of sterile human hair, previously cleaned by immersion in absolute ethanol, or with powdered cellulose fibre. Cultures were examined daily by phase contrast microscopy to establish whether such relatively inert materials (which are normally poorly degraded by cell activity) were either, phagocytosed or, were capable of inducing granuloma formation, providing supportive evidence for the biogenesis of macrophage polykarya.
Results
Culture establishment, morphology and kinetics Cell suspensions consisted p r e d o m i n a n t l y o f single cells, b u t some cell clusters (of n o m o r e t h a n 5 m o n o n u c l e a t e d cells) were also present. G r a n u l o c y t e , e r y t h r o i d a n d meg a k a r y o c y t e lineages at all stages o f m a t u r a t i o n were p r e s e n t in the initial cell s u s p e n s i o n , as d e t e r m i n e d by light m i c r o s c o p i c a l e x a m i n a t i o n o f fixed specimens, stained with h a e m a t o x y l i n a n d eosin. After 24 h in culture, cells h a d b e g u n to a t t a c h to the surface o f the flasks a n d b y 48 h the a d h e r e n t cell
>
Fig. 1. a Two-day-old culture showing adherent cell layer of mono-
nucleated cells of various sizes and staining characteristics, b Eleven-day-old culture showing fusion of multinucleated cells. One nucleus (arrow)appears to be transferring from one multinucleated cell to the other, e Fusion of mono-, di- and oligonucleated cells with a large multinucleated cell in l 1-day-old culture, d Monoand multinucleated cells in ll-day-old culture showing variation in degree of acid phosphatase activity. Many mononucleated cells are intensely stained while others remain unstained. All multinucleated ceils are moderately positive, x 360. Bar: 30 gm
20
10~ 7 days
50 -G
2 25
0
3-9 10q9 20-29 30-39 40-49 50-59 60-69 70-79 80-89 90-99 100+ NO. of nuclei p e r cell
Fig. 2. Comparison of numbers of nuclei in 100 multinucleated cells per flask, at 7 and 21 days in culture: 3 flasks were assessed at each time. Bar: 1 SD. Note, that the high incidence of small, multinucleated cells containing few ( < 10) nuclei, which dominate the culture at 7 days, gradually falls with time, to be replaced by greater numbers of large, multinucleated cells containing abundant nuclei.
al differentiation to that (viz, tartrate-resistant acid phosphatase activity) held to be diagnostic of osteoclast identity. Large multinucleated cells were present in most cultures for up to 5 weeks, but in older cultures they rapidly disappeared from the monolayer, leaving a population consisting predominantly of isolated clusters of macrophages separated by fibroblast-like cells (Fig. 3). These latter rapidly overgrew the culture between 6 and 8 weeks, and became the dominant cell-type. From the first demidepopulation onwards, cells in suspension consisted of a relatively homogeneous population of mononucleated cells, with occasionally a few multinucleated cells also present. Large, macroscopically-visible multicellular aggregates (spheroids) composed of microscopically-similar mononucleated cells, and containing abundant lipid, were attached to the substrate and cells of the monolayer, and were also present in suspension (Fig. 4). Similar multicellular aggregates were also found attached to the cell monolayer as soon as multinucleated cells were established, and with which they were frequently associated. As demonstrated by phase contrast and TEM, in 5-week-old cultures, these attached multicellular aggregates often formed discrete foci of fibroblast-like and adipocyte-like cells in the cellmonolayer.
Influence of culture conditions population consisted predominantly of mononucleated macrophages (Fig. 1 a). After 4 days the monolayer consisted of a mixed population of macrophages and large mononucleated cells. Occasional bi- or multinucleated cells were also observed. Between days 5 and 9, a confluent monolayer of mononuclear and small multinucleated cells was established and the number and size of multinucleated cells in the culture medium increased. During the following 14 days, the number of multinucleated cells in the culture decreased, while their overall size and nuclear content increased (Figs. I b, 2). These giant multinucleated cells often attained a diameter of more than 300 gm. During the 2nd and 3rd week, cell fusion was frequently observed and often involved simultaneous fusions of mono-, di-nucleated and small and large multinucleated cells (Fig. 1 c). Mono-, bi- and multinucleated cells displayed positive acid phosphatase and tartrate-resistant acid phosphatase activities. Acid phosphatase activity appeared greatest in mono- and bi-nucleated cells and those multinucleated cells which contained relatively few nuclei. The larger multinucleated cells were the least positive (Fig. I d). Those mononucleated cells which displayed monocyte-like (round) or macrophage-like (multipolar) characteristics were stained to equal degrees. However, there remained within the culture, numerous, mainly small, mononucleated cells which were completely acid phosphatase-negative. Fewer cells demonstrated tartrate-resistant acid phosphatase activity, and it is noteworthy that these represented a large cohort of the acid phosphatase-positive cell population. These reactions indicated that a dual expression of these enzymes occurred within some cells of the population, suggesting function-
Cultures seeded at densities within the range 2-5 x 107 nucleated bone marrow cells per 25 cm 2 area of culture flask, developed in the manner already described. At seeding densities below 1 x 10v cells, however, multinucleated cells did not develop, and fibroblast overgrowth occurred between days 4-7. Growth of the cultures was rapid at 37~ C, but at 33~ C, the temperature at which murine long-term bone marrow cultures are normally maintained (Spooncer and Dexter 1984), growth was slower but somewhat more sustained. The presence of isologous dog serum was essential for the development of the multinucleated cells, whereas its substitution by foetal bovine, donor horse or donor sheep serum, resulted in the rapid development of a dense monolayer of fibroblast-like cells. Autologous serum was not, however, required. The presence of hydrocortisone in the culture medium resulted in retention of the multinucleated cells, whereas in its absence, multinucleated cells developed more rapidly, but also quickly disappeared from the culture.
UltrastructuraI characteristics Eleven days to 4 weeks. SEM showed that the cultures were characterised by many, very large cells with diameters often exceeding 250 ~tm (Fig. 5 a). The central area of their uppermost surface showed raised, closely packed, membranous projections, with a peripheral rim, 10-15 gm wide, making up a smooth marginal band (Fig. 5 b). In this area, small (80-200 nm diameter) punctate indentations of the membrane, reminiscent of pino-
21
Fig. 3. SEM of 6-week culture showing isolated groups of rounded mononucleated cells separated by sheets of overgrowing, fibroblast-like cells, x 790. Bar: 50 gm
Fig. 4. Phase contrast photomicrograph of 4-week culture showing 2 rounded spheroids (S). • 180. Bar: 60 gm
cytotic pits, were arranged in focal clusters. At the cell edge, m a n y fine filopodial processes made contact with similar processes on adjacent cells. In association with these large flattened cells were numerous smaller, rounded or spindle-shaped cells some 10-15 ~tm diameter; m a n y were attached at the periphery of the large cells (Fig. I b, c). The surface of these small cells showed
similar foldings of the p l a s m a l e m m a in conjunction with several relatively large spherical protrusions. In some instances the contact between the two cell-types was so close, with cytoplasmic processes derived f r o m the small cells appearing to merge with the surface of the larger cells, as to suggest fusion of one with the other (Fig. 5 c). By T E M , the mononucleated cells generally appeared
22
23 to be ultrastructurally similar to peripheral blood monocytes, except that they contained more lipid droplets, lysosomes, autophagosomes and Golgi complexes. The large multinucleated cells were strikingly similar to the mononucleated cells in their nuclear characteristics and cytoplasmic appearance, but had greater numbers of peripheral, electron-dense lipid droplets forming a circumferential band lying subjacent to the plasmalemma (Fig. 6 a). Other notable ultrastructural features of these cells were the increased numbers of autophagosomes and membrane whorls (Fig. 6b). Horizontal or tangential sections demonstrated the tortuous foldings of the ruffled membrane seen by SEM (Fig. 5 a). A feature c o m m o n to both mono- and multinucleated cells, was the presence of either, compact dense granules, or, moderately electron-dense, single membrane-bounded rounded organelles (diameter 0.2-0.8 l.tm), both of which contained tubular membrane elements arranged in a lattice-like array (Fig. 6b). The composition of these inclusions varied considerably, from small, single membrane-bounded vacuoles with flocculent content and a single cluster of loosely-associated tubular membranes, to large forms invested by, or containing, several membrane layers with a central, large (0.5-1.5 gm diameter), multilobed lattice array (Fig. 6 b). These inclusions were not seen in peripheral blood monocytes or mononucleated cells from fresh bone marrow of healthy dogs. A possible route of formation of these structures from effete organelles is indicated in Fig. 6b. At this time too, a few, rounded, multicellular aggregates of mononucleated cells were found to be attached to the adherent cell layer (Fig. 7). A t 4 weeks. The confluent monolayer consisted of larger
multinucleated cells interspersed with many smaller and intermediate-sized flattened cells, and was overlaid by many small, rounded monocytes which were no longer primarily associated with the periphery of the large flattened cells (similar to those cells shown in Fig. 5 a). Intercellular contact between adjacent cells of the monolayer was characterised by many short, interdigitating filopodial processes constituting a peripheral band some 25 gm wide. Occasionally, stouter processes interconnected cells at greater separations of up to 2-3 cell diameters.
Cell&Tissue Research 9 Springer-Verlag 1992
Multinucleated giant cells in primary cultures derived from canine bone marrow - evidence for formation of putative osteoclasts M.C. Bird 1, D. Garside
2,
and H.B. Jones 1
i Department of Pathology and z Department of Toxicology, Smith Kline and French Research Ltd., The Frythe, Welwyn, Hertfordshire AL6 9AR, UK Received June 17, 1991 / Accepted November 2, 1991
Summary. Mononucleated cells derived from canine bone marrow were maintained in vitro for up to 6 weeks. The culture characteristics and development of these cells were evaluated by histological, ultrastructural and histochemical methods. Within I week the cells had fused together to form flattened, multinucleated cells. Further fusing with one another and other mononucleated cells produced large (diameters more than 300 gin), multinucleated cells which frequently contained more than 50 nuclei per cell and exhibited ultrastructural and histochemical features that were strikingly similar to those displayed by osteoclasts. The confluent monolayer of mono- and multinucleated cells present at 4 weeks had, by the sixth week, become altered such that fibroblast overgrowth obliterated all other cells. During the development of the culture adipocytes became differentiated from mononuclear cells and frequently were located within spherical multicellular aggregates (spheroids). Functional assessments were employed to investigate whether the multinucleated ceils generated in this way, represented osteoclast-like cells, or alternatively, were related to macrophage polykarya as found in foreign body granulomata in vivo. Neither resorption pits on sperm whale dentine slivers (diagnostic of osteoclasts), nor formation of granulomata in vitro, were observed. We believe that the present results indicate that the multinucleated cells generated from canine bone marrow mononuclear precursors in vitro, merit designation as osteoclast-like cells. Definitive characterisation however, must await further functional assessments of hormone responsiveness.
The comparatively recent establishment of long-term tissue culture systems derived from canine bone marrow has made possible the study of component parts of important, but ill-understood, phenomena such as osteoclast formation (Allen et al. 1981 ; Severson 1983), bone resorption (Allen et al. 1981) and cellular responses to trophic factors (Suda et al. 1983). To date, these techniques have principally been used for investigations in two areas, namely, for the study of haemopoietic cell lines in a controlled in vitro environment, and for the study of the origin, function and fate of the bone-modelling osteoblasts and osteoclasts. The influence of imbalance in the complex interrelationships between the cells, and the effects of functional impairment by physiological or toxicological perturbations are important to an understanding of the pathophysiological processes associated with bone disease, e.g. osteoporosis and osteopetrosis. In the present study, an in vitro tissue culturing technique has been used to investigate the hitherto unrecorded establishment, maintenance and cellular characteristics of developing multinucleated cells derived from canine bone marrow. In addition, an assessment of the functional capability of the cultures has been made to throw some light on bone resorptive capacity and granuloma formation in vitro. It was felt that such a study was necessary to distinguish between osteoclasts and multinucleate giant cells (macrophage polykarya) present in vivo in inflammatory cell aggregates associated with poorly degradable particles (Kasahara et al. 1988; Kreipe et al. 1988; Walters and Schneider 1987).
Key words: Bone marrow - Osteoclasts - Giant cells - Macrophage polykarya Tissue c u l t u r e - Dog
Materials and methods Culture preparation and establishment
* Dr. Martin Bird died during the final phase of this work and it is to his memory that this paper is respectfully dedicated Offprint requests to: H.B. Jones, Safety of Medicines Department, ICI Pharmaceuticals, Alderley Park, Macclesfield, Cheshire, SK10 4TG, UK
Beagle dogs (employed as controls in safety evaluation studies of novel pharmaceutical agents) were killed by an intravenous injection of 150 mg/kg sodimn pentobarbitone (May and Baker, Dagenham, UK). Femurs were excised and, after cutting the epiphyses, the bone marrow was flushed out with Iscove's modified Dulbecco's medium (Gibco BRL, Paisley, UK) containing 20% isologous
18 dog serum, 10 -6 M hydrocortisone (Sigma, Poole, UK) and 50 gg/ ml gentamycin (growth medium). Dispersed cell suspensions were prepared by gentle trituration and pipetting onto 0-90% Percoll gradients, then centrifuged at approximately 500 g for 5 min, in order to remove all adipocytes and red blood cells. After washing, cultures were established by inoculating 2-5 x 107 nucleated cells in 10 ml of growth medium into sterile, 25 cm 2 (growth area) tissue culture flasks (Gibco Nunc), and incubated at 37~ C in 5% CO2/ 95% air. At weekly intervals they were fed by replacing half the spent growth medium (demidepopulation). Cells in suspension, removed together with the growth medium, were not replaced. Prior to microscopic examination, cultures were washed once with sterile saline (0.9% sodium chloride) and examined by phasecontrast microscopy using a Leitz Labovert inverted phase-contrast microscope. For examination and counting of nuclei in multinucleated cells, cells attached to the surface of the culture flask were washed once with saline, fixed with neutral buffered formalin and stained with haematoxylin and eosin. The numbers of nuclei in a total of 100 multinucleated cells, in randomly chosen fields, were counted in 7 and 21 day cultures. Three flasks were assessed in this way at both times.
Cytochemistry Monolayer cultures, sampled at intervals up to 21 days, were fixed by immersion in 10% neutral buffered formalin for 5 min, and processed cytochemically by standard methods (Hammarstrom et al. 1971 ; Pearse 1972; Bancroft and Stevens 1977) for the demonstration of the following tissue components: neutral lipid (oil Red O), calcium deposits (Alizarin Red S), and alkaline and acid phosphatases (tartrate resistant and unresistant). Counterstaining with haematoxylin was performed as appropriate.
Functional characteristics
Electron microscopy In order to determine the morphological characteristics of bone marrow mononucleated cells prior to culture establishment, cell suspensions were prepared as described above and depleted of erythrocytes by lysis with 0.8% (w/v) aqueous ammonium chloride. Their morphology was compared with that of canine peripheral blood mononucteated cells prepared by fractionation of whole blood over Ficoll Hypaque (sg of 1.077, Boyum 1983). These specimens, and established cultures (following decantation of the culture medium) sampled at 11, 28 and 42 days, were fixed by immersion in a solution of 4% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) for 2 h, then twice washed in the same buffer for 10 min. Secondary fixation by 1% osmium tetroxide in the same buffer for 1 h preceded separation of the samples into those designated for transmission (TEM) or scanning electron microscopy (SEM). For TEM, samples were dehydrated in an ascending ethanol series and 1, 2-epoxypropane. At this stage, cells adherent to the culture flask were carefully detached by means of a clean pith stick prior to infiltration with, and embedding in, Taab resin (Taab Equipment Laboratories Ltd., Aldermaston, UK). Semithin (1 Ixm thick), toluidine-blue-stained sections were examined by light microscopy (LM) in order to evaluate cell characteristics and selection of tissue areas for subsequent ultramicrotomy. Ultrathin sections (70 90 nm thick) were stained with uranyl acetate and lead citrate solutions, and examined in a JEOL, JEM 1200EX transmission electron microscope at an accelerating voltage of 80 kV. For SEM, a modification of the method of Wollweber et al. (1981) was employed. The osmicated samples were further immersed in a solution of 1% tannic acid in 0.1 M cacodylate buffer (pH 7.4) for 30 min and, after several washes in distilled water, placed into 1% aqueous uranyl acetate solution for the same length of time. Following dehydration in an ascending ethanol series, the specimens of cells adherent to substrata were critical-point dried and mounted on aluminium support stubs with silver paint (Agar Scientific Ltd, Stanstead, UK). All specimens were coated with a thin (8-10 nm thick) layer of gold prior to examination in a JEOL, JSM 840 scanning electron microscope at an accelerating voltage of 8 kV.
In order to elucidate multinucleated cell identity as either osteoclast or macrophage polykarya (multinucleated giant cells), 3 functional criteria were used for assessment: (1) Seven- or 10-day-old established cultures were maintained for up to 21 days with slivers (100 gm thick) of sperm whale dentine (Kindly supplied by Professor A. Boyde and Dr. S. Jones, Department of Anatomy and Embryology, University College, London). At the end of this period, adherent cells were removed from the slivers by gently rubbing the surface and, following light counterstaining with toluidine blue, the slivers were examined by light microscopy for the presence of resorption pits, a diagnostic feature of osteoclast activity. A few slivers were also prepared for examination by SEM. (2) Following the first demidepopulation, 7-day-old cultures were maintained for 14 days in medium to which fine, particulate fragments of dog bone (femur) were added. The particles were prepared from absolute-ethanol-sterilised pieces of femur by scraping the fractured end of the central portion of the diaphysis with a clean scalpel blade. These were then stored in ethanol until required. Prior to use, the ethanol was first removed by several changes of sterile isotonic saline. (3) Seven- or 14-day-old cultures were incubated for 14 days with fragments of sterile human hair, previously cleaned by immersion in absolute ethanol, or with powdered cellulose fibre. Cultures were examined daily by phase contrast microscopy to establish whether such relatively inert materials (which are normally poorly degraded by cell activity) were either, phagocytosed or, were capable of inducing granuloma formation, providing supportive evidence for the biogenesis of macrophage polykarya.
Results
Culture establishment, morphology and kinetics Cell suspensions consisted p r e d o m i n a n t l y o f single cells, b u t some cell clusters (of n o m o r e t h a n 5 m o n o n u c l e a t e d cells) were also present. G r a n u l o c y t e , e r y t h r o i d a n d meg a k a r y o c y t e lineages at all stages o f m a t u r a t i o n were p r e s e n t in the initial cell s u s p e n s i o n , as d e t e r m i n e d by light m i c r o s c o p i c a l e x a m i n a t i o n o f fixed specimens, stained with h a e m a t o x y l i n a n d eosin. After 24 h in culture, cells h a d b e g u n to a t t a c h to the surface o f the flasks a n d b y 48 h the a d h e r e n t cell
>
Fig. 1. a Two-day-old culture showing adherent cell layer of mono-
nucleated cells of various sizes and staining characteristics, b Eleven-day-old culture showing fusion of multinucleated cells. One nucleus (arrow)appears to be transferring from one multinucleated cell to the other, e Fusion of mono-, di- and oligonucleated cells with a large multinucleated cell in l 1-day-old culture, d Monoand multinucleated cells in ll-day-old culture showing variation in degree of acid phosphatase activity. Many mononucleated cells are intensely stained while others remain unstained. All multinucleated ceils are moderately positive, x 360. Bar: 30 gm
20
10~ 7 days
50 -G
2 25
0
3-9 10q9 20-29 30-39 40-49 50-59 60-69 70-79 80-89 90-99 100+ NO. of nuclei p e r cell
Fig. 2. Comparison of numbers of nuclei in 100 multinucleated cells per flask, at 7 and 21 days in culture: 3 flasks were assessed at each time. Bar: 1 SD. Note, that the high incidence of small, multinucleated cells containing few ( < 10) nuclei, which dominate the culture at 7 days, gradually falls with time, to be replaced by greater numbers of large, multinucleated cells containing abundant nuclei.
al differentiation to that (viz, tartrate-resistant acid phosphatase activity) held to be diagnostic of osteoclast identity. Large multinucleated cells were present in most cultures for up to 5 weeks, but in older cultures they rapidly disappeared from the monolayer, leaving a population consisting predominantly of isolated clusters of macrophages separated by fibroblast-like cells (Fig. 3). These latter rapidly overgrew the culture between 6 and 8 weeks, and became the dominant cell-type. From the first demidepopulation onwards, cells in suspension consisted of a relatively homogeneous population of mononucleated cells, with occasionally a few multinucleated cells also present. Large, macroscopically-visible multicellular aggregates (spheroids) composed of microscopically-similar mononucleated cells, and containing abundant lipid, were attached to the substrate and cells of the monolayer, and were also present in suspension (Fig. 4). Similar multicellular aggregates were also found attached to the cell monolayer as soon as multinucleated cells were established, and with which they were frequently associated. As demonstrated by phase contrast and TEM, in 5-week-old cultures, these attached multicellular aggregates often formed discrete foci of fibroblast-like and adipocyte-like cells in the cellmonolayer.
Influence of culture conditions population consisted predominantly of mononucleated macrophages (Fig. 1 a). After 4 days the monolayer consisted of a mixed population of macrophages and large mononucleated cells. Occasional bi- or multinucleated cells were also observed. Between days 5 and 9, a confluent monolayer of mononuclear and small multinucleated cells was established and the number and size of multinucleated cells in the culture medium increased. During the following 14 days, the number of multinucleated cells in the culture decreased, while their overall size and nuclear content increased (Figs. I b, 2). These giant multinucleated cells often attained a diameter of more than 300 gm. During the 2nd and 3rd week, cell fusion was frequently observed and often involved simultaneous fusions of mono-, di-nucleated and small and large multinucleated cells (Fig. 1 c). Mono-, bi- and multinucleated cells displayed positive acid phosphatase and tartrate-resistant acid phosphatase activities. Acid phosphatase activity appeared greatest in mono- and bi-nucleated cells and those multinucleated cells which contained relatively few nuclei. The larger multinucleated cells were the least positive (Fig. I d). Those mononucleated cells which displayed monocyte-like (round) or macrophage-like (multipolar) characteristics were stained to equal degrees. However, there remained within the culture, numerous, mainly small, mononucleated cells which were completely acid phosphatase-negative. Fewer cells demonstrated tartrate-resistant acid phosphatase activity, and it is noteworthy that these represented a large cohort of the acid phosphatase-positive cell population. These reactions indicated that a dual expression of these enzymes occurred within some cells of the population, suggesting function-
Cultures seeded at densities within the range 2-5 x 107 nucleated bone marrow cells per 25 cm 2 area of culture flask, developed in the manner already described. At seeding densities below 1 x 10v cells, however, multinucleated cells did not develop, and fibroblast overgrowth occurred between days 4-7. Growth of the cultures was rapid at 37~ C, but at 33~ C, the temperature at which murine long-term bone marrow cultures are normally maintained (Spooncer and Dexter 1984), growth was slower but somewhat more sustained. The presence of isologous dog serum was essential for the development of the multinucleated cells, whereas its substitution by foetal bovine, donor horse or donor sheep serum, resulted in the rapid development of a dense monolayer of fibroblast-like cells. Autologous serum was not, however, required. The presence of hydrocortisone in the culture medium resulted in retention of the multinucleated cells, whereas in its absence, multinucleated cells developed more rapidly, but also quickly disappeared from the culture.
UltrastructuraI characteristics Eleven days to 4 weeks. SEM showed that the cultures were characterised by many, very large cells with diameters often exceeding 250 ~tm (Fig. 5 a). The central area of their uppermost surface showed raised, closely packed, membranous projections, with a peripheral rim, 10-15 gm wide, making up a smooth marginal band (Fig. 5 b). In this area, small (80-200 nm diameter) punctate indentations of the membrane, reminiscent of pino-
21
Fig. 3. SEM of 6-week culture showing isolated groups of rounded mononucleated cells separated by sheets of overgrowing, fibroblast-like cells, x 790. Bar: 50 gm
Fig. 4. Phase contrast photomicrograph of 4-week culture showing 2 rounded spheroids (S). • 180. Bar: 60 gm
cytotic pits, were arranged in focal clusters. At the cell edge, m a n y fine filopodial processes made contact with similar processes on adjacent cells. In association with these large flattened cells were numerous smaller, rounded or spindle-shaped cells some 10-15 ~tm diameter; m a n y were attached at the periphery of the large cells (Fig. I b, c). The surface of these small cells showed
similar foldings of the p l a s m a l e m m a in conjunction with several relatively large spherical protrusions. In some instances the contact between the two cell-types was so close, with cytoplasmic processes derived f r o m the small cells appearing to merge with the surface of the larger cells, as to suggest fusion of one with the other (Fig. 5 c). By T E M , the mononucleated cells generally appeared
22
23 to be ultrastructurally similar to peripheral blood monocytes, except that they contained more lipid droplets, lysosomes, autophagosomes and Golgi complexes. The large multinucleated cells were strikingly similar to the mononucleated cells in their nuclear characteristics and cytoplasmic appearance, but had greater numbers of peripheral, electron-dense lipid droplets forming a circumferential band lying subjacent to the plasmalemma (Fig. 6 a). Other notable ultrastructural features of these cells were the increased numbers of autophagosomes and membrane whorls (Fig. 6b). Horizontal or tangential sections demonstrated the tortuous foldings of the ruffled membrane seen by SEM (Fig. 5 a). A feature c o m m o n to both mono- and multinucleated cells, was the presence of either, compact dense granules, or, moderately electron-dense, single membrane-bounded rounded organelles (diameter 0.2-0.8 l.tm), both of which contained tubular membrane elements arranged in a lattice-like array (Fig. 6b). The composition of these inclusions varied considerably, from small, single membrane-bounded vacuoles with flocculent content and a single cluster of loosely-associated tubular membranes, to large forms invested by, or containing, several membrane layers with a central, large (0.5-1.5 gm diameter), multilobed lattice array (Fig. 6 b). These inclusions were not seen in peripheral blood monocytes or mononucleated cells from fresh bone marrow of healthy dogs. A possible route of formation of these structures from effete organelles is indicated in Fig. 6b. At this time too, a few, rounded, multicellular aggregates of mononucleated cells were found to be attached to the adherent cell layer (Fig. 7). A t 4 weeks. The confluent monolayer consisted of larger
multinucleated cells interspersed with many smaller and intermediate-sized flattened cells, and was overlaid by many small, rounded monocytes which were no longer primarily associated with the periphery of the large flattened cells (similar to those cells shown in Fig. 5 a). Intercellular contact between adjacent cells of the monolayer was characterised by many short, interdigitating filopodial processes constituting a peripheral band some 25 gm wide. Occasionally, stouter processes interconnected cells at greater separations of up to 2-3 cell diameters.
