0013-7227/86/1185-1843$02.00/0 Endocrinology Vol. 118, No. 5 Copyright © 1986 by The Endocrine Society Printed in U.S.A.

Growth Hormone Potentiates Colony Formation of Epiphyseal in Suspension Culture*

ANDERS LINDAHL, JORGEN ISGAARD, ANDERS NILSSON, AND

OLLE G. P. ISAKSSON Downloaded from https://academic.oup.com/endo/article/118/5/1843/2540080 by guest on 30 September 2021 Department of Physiology, University of Goteborg, S-400 33 Goteborg, Sweden

ABSTRACT. The effect of human GH (hGH) on colony for- centrations of hGH (80 and 160 ng/ml) showed reduced poten- mation of rat epiphyseal plate chondrocytes was studied in tiation of colony formation. The stimulatory effect of hGH was suspension culture. Chondrocytes were isolated enzymatically expressed at 10-15% of NCS at 14 days of culture. There was a from epiphyseal plates of the proximal of 28-day-old normal linear relation between the number of seeded cells and the male rats, and were cultured in a suspension stabilized with number of colonies formed, both in the absence and presence of 0.5% agarose. After approximately 7 days of culture in the hGH. presence of 10% newborn calf serum (NCS), colo- These results show that GH potentiates colony formation in nies developed consisting of varying numbers of cells in matrix. chondrocytes of the epiphyseal growth plate, providing further No colonies developed in the absence of NCS, and the number support for the contention that GH exert a direct stimulatory of formed colonies was proportional to the concentration of NCS effect on epiphyseal and thus stimulates longitudinal (5-20%) in the medium. growth directly. The finding that GH preferentially poten- hGH potentiated the formation of large size colonies (diam- tiated the formation of large size colonies suggests that GH eter >90 nm) after a culture period of 10 or 14 days. The lowest promoted the differentiation of early proliferative chondrocytes effective concentration of hGH was 10 ng/ml, while 40 ng/ml or stem cell chondrocytes with an inherent high capacity to hGH gave a maximal stimulatory effect (40-50%). Higher con- proliferate. (Endocrinology 118: 1843-1848, 1986)

REVIOUS experiments have shown that local ad- sequently these cells mature, enter the hypertrophic Pministration of GH into the epiphyseal growth plate zone, become calcified, and are eventually incorporated of the tibia of hypophysectomized rats stimulates longi- into metaphyseal bone as (8, 9). Therefore, tudinal bone growth on the injected side (1-4). Further- the production of new cells in the growth plate is the more, cultured chondrocytes from rabbit ear and epiphy- result of both cell differentiation and proliferation. seal plate have specific binding sites for GH (5). GH Earlier studies have shown that chondrocytes isolated stimulates DNA synthesis in cultured chondrocytes from from various cartilage sources form colonies when cul- rabbit ear and rat rib epiphyseal plates (6), as well as tured in suspension (10, 11). In contrast to monolayer proteoglycan synthesis in cultured chondrocytes from rat culture, suspension culture of chondrocytes makes it rib growth cartilage (7). These observations suggest that possible to study the growth characteristics of individual GH directly interacts with cells in the growth plate and chondrocytes from a heterogenous cell population. that this interaction ultimately results in an increased To address the question if GH influences the growth cell proliferation. of epiphyseal chondrocytes, we have adopted the tech- The chondrocytes in the growth plate are strictly nique of culturing epiphyseal growth plate chondrocytes in a suspension stabilized with agarose (10). We now organized in a fascicular pattern according to the stage report that GH potentiates the formation of large size of maturation. During the process of longitudinal bone chondrocyte colonies, which suggests that GH stimulates growth, cells from the stem cell area bordering the bony differentiation of epiphyseal chondrocyte stem cells or differentiate and enter the proliferative zone early proliferative chondrocytes. where the cells undergo limited clonal expansion. Sub-

Received July 17, 1986. Materials and Methods Address requests for reprints and all correspondence to: Dr. Anders Lindahl, Department of Physiology, University of Goteborg, P.O. Box Animals 33031, S-400 33 Goteborg, Sweden. * This work was supported by grants from the Swedish Medical Male Sprague-Dawley rats (Alab Laboratory Services Ltd., Research Council (14X-04250), The Goteborg Medical Society, and the Stockholm, Sweden), 28-30 days of age, were used in all exper- Faculty of Medicine, University of Goteborg. iments. They were housed under controlled conditions with

1843 1844 GH AND CHONDROCYTE COLONY FORMATION Endo • 1986 Volll8«No5 constant temperature (24-26 C) and humidity (50-60%), and KabiVitrum (Stockholm, Sweden) and the ovine PRL (oPRL, with a 14-h light, 10-h dark cycle. The animals were given tap lot NIH-P-S-3) was supplied by NIADDK (Bethesda, MD). water and pellet food ad libitum. The cultures were maintained at 37 C in air containing 5% CO2, for 6, 10, 14, or 21 days. Isolation of chondrocytes Clonal assay Animals were killed by cervical dislocation, the distal part of the body was soaked in 70% ethanol, and both tibiae were Cultures were terminated by fixation in buffered formalde- dissected free. The isolated legs were soaked in 0.1% chlorhex- hyde (4%) and stained with alcian blue (0.5% in 0.04 M hydro- idine for 10 min and then rinsed twice in Ham's F-12 medium chloric acid) to identify colonies producing giycosaminoglycans (Grand Island Biological Co., Paisley, Scotland). Subsequently, (12). Each experiment consisted of pooled growth plates from the tibial bone was trimmed free of soft tissue, and the epiphy- 7 to 8 animals, from which duplicate cultures were made. Downloaded from https://academic.oup.com/endo/article/118/5/1843/2540080 by guest on 30 September 2021 seal cartilage was dissected from the proximal tibia under the Colonies were optically counted in 100 squares (2-mm grid) for microscope, avoiding any loss of cartilage bordering the bony each cultured dish, counts being made separately by 2 investi- epiphysis. gators, and the total number of colonies were calculated from 415 grids. A cell colony was defined as a cluster of cells with Pooled epiphyseal growth plates from seven to eight animals matrix stained by alcian blue with a diameter of more than 30 were digested in a spinner bottle for 4 h at 37 C in 0.12% (wt/ fim. Serial sections from colonies of this size showed that they vol) clostridium collagenase (batch 4A196, Worthington Bio- contained 5-15 chondrocytes. The amount of growth stimula- chemicals, Freehold, NJ) and 0.02% DNAse I (batch 74F-9670, tion was expressed as percent stimulation of control value, of Sigma Chemical Co., St Louis, MO) in F-12 medium. After as an increase in number of colonies per dish. isolation the cells were washed three times in serum-free me- dium, and cell counts were made in a hemocytometer. Cell viability at the end of the incubation was at least 90%, as Colony size determination determined by the trypan blue exclusion technique. Subse- 5 Cell cultures from 5 different experiments were assayed for quently, cells were diluted to a concentration of 5 X 10 cells/ colony size. The diameter of each colony was measured by ml (F-12 medium, 37 C) and were suspended in the agarose gel microscope using a calibrated eyepiece, and a total of 974 culture. colonies was counted, 423 in the control group and 551 in the GH group. The colonies were arranged in 5 different groups Culture of chondrocytes in a suspension stabilized with agarose according to size before statistical analysis, the group bound- aries representing subdivisions of the measurement scale: 40- Cells were cultured in agarose according to Benya and Shaf- 56, 64-80, 88-104, 112-128, and 136-152 /an. fer (10) in 60-mm diameter Petri dishes (Falcon Plastics, Los Angeles, CA) coated with a bottom film of 1% standard low RIA of hGH in culture medium (SL) agarose (catalog no 162-0100, Bio-Rad, Richmond, CA) in water. Before coating, the agarose was autoclaved (110 C, 45 RIA of hGH in culture medium was performed in duplicate min). The gel was allowed to solidify at room temperature. samples of each culture at 1, 7, and 14 days after start of Low gel temperature (LGT) agarose (Bio-Rad, catalog no. culture. hGH was determined by a double antibody RIA as 162-0020) was autoclaved, as above, and mixed with an equal described earlier (13), the sensitivity of this assay being defined volume of concentrated (x2) F-12 medium to give a final by 2 SD above a zero dose response in the standard curve, i.e. agarose concentration of 1%. The cells were diluted with F-12 below 5 hg/ml. containing newborn calf serum (NCS; Grand Island Biological Co.) 2 times the final culture concentration, at 37 C to a final Statistical procedure concentration of 80,000 cells/ml. This cell suspension was then mixed with 1% LGT agarose to give a final cell concentration Values are given as mean ± SEM. The significance of differ- of 40,000 cells/ml in 0.5% agarose. In cell cultures 3 different ences between means was calculated by Student's paired t test cell densities were used (10,000, 20,000, and 40,000 cells/ml). in Figs. 2B and 5B and Student's t test in Figs. 4 and 2A. Dishes precoated with SL agarose were heated to 37 C before 2 Differences in distribution of colony size were determined by 2 ml of the LGT cell suspension were added. Dishes were kept at X -test. Two-way analysis of variance was used to evaluate the 37 C for 10 min before gelation at 4 C (10 min). Subsequently, effect of hGH on colony formation at different cell densities. P 3 ml F-12 medium supplemented with various serum concen- values less than 0.05 were considered significant. trations (5, 10, 15, or 20% NCS) with and without added hormones were added on top of the gelated agarose. Thereafter, Results the cultures were screened for adherent cell clusters of more than 3 cells. No such clusters were seen at the start of culture Rat epiphyseal chondrocytes cultured in a suspension in any experiments presented in this study. Medium was further stabilized with agarose formed small colonies with few supplemented with HEPES (10 HIM), trimethylamino ethane cells during 7 days in culture. The colonies continued to sulfonic acid (10 mM), and 100 fig/m\ with gentamicin sulfate grow and, after 14 days in culture, they were easily (Gentamycin, Sigma) and 50 Mg/ml L-ascorbic acid (50 Mg/ml). bac identified (Fig. 1). Colonies consisted of varying numbers The bacterially produced hGH (hGH ) was a gift from of chondrocytes embedded in a matrix containing pro- GH AND CHONDROCYTE COLONY FORMATION 1845

15O-i D---ohGH 40 ng/ml 140- **

i- £ i icH

100-

9°-. n Downloaded from https://academic.oup.com/endo/article/118/5/1843/2540080 by guest on 30 September 2021 80- 5 10 15 20 SERUM CONCENTRRTION V.

lOOOn • DhGH 40 ng/ml .rol 800-

" GOO- If) LJ

O 400-

200- FIG. 1. Chondrocyte colonies from epiphyseal cartilage in suspension culture. Chondrocytes were isolated from epiphyseal growth plates of B the proximal tibia of 28-day-old male rats and cultured for 14 days in 0 J suspension culture stabilized with agarose in F-12 medium with 10% 0 5 10 15 20 NCS. Cultures were fixed in buffered 4% formaldehyde. A, Chondrocyte SERUM CONCENTRRTION V. colonies (X50); B, chondrocyte colony (X200). FIG. 2. Effect of hGH on the colony formation of rat tibia epiphyseal TABLE 1. Effect of hGH on the colony formation of rat tibia epiphyseal chondrocytes at various serum concentrations. Chondrocytes were iso- chondrocytes at two different culture periods. lated from epiphyseal growth plates of the proximal tibia of 28-day-old male rats and cultured in suspension at a density of 80,000 cells per % stimulation dish. The suspension culture was stabilized with agarose in F-12 Colonies per dish (n) Culture time of control value medium with various concentrations of NCS, with or without 40 ng/ (days) ml hGH. Cultures were incubated for 14 days, and the number of Control hGH (40 ng/ml) P Control (100%) P colonies were counted using inverted light microscope. A, Effect of 10 503 ± 64 758 ±119 <0.05 152 ± 14.8 <0.01 hGH on colony formation expressed as percent stimulatory effect of 14 618 ± 56 899 ± 27 <0.01 143 ± 13.8 <0.01 control value at 5, 10, 15, or 20% serum concentration. Control value indicated by solid line. B, Effect of hGH on colony formation at 5, 10, Chondrocytes were isolated from rat epiphyseal growth plates of the 15, and 20% serum concentration expressed as total number of colonies proximal tibia of 28-day-old male rats and cultured in suspension at a per dish. Values are means ± SEM from 11 (5 and 10% serum) and 7 density of 80,000 cells per dish. The suspension culture was stabilized (15 and 20% serum) different cell isolations using different rats. (*, P with agarose in F-12 medium containing 10% NCS, with or without 40 < 0.05; **, P < 0.01 control vs. hGH in each serum concentration ng/ml hGH. Colonies were counted using an inverted light microscope. calculated by Student's t test in panel A and paired t test in panel B). Values are means ± SEM of 5 different cell isolations using different rats. tion times (Table 1). The presence of 40 ng/ml hGH teoglycans as identified with alcian blue staining (12). In increased the number of colonies above the control group the present study, aggregates of cells with a diameter of to approximately the same degree for both culture times, 30 /um or more were classified as colonies. Serial sections (Table 1). However, since the colonies were more easily of colonies of this size showed that they contained 5-15 identified after 14 days, this culture period was used chondrocytes (data not shown). When chondrocytes were subsequently for studying GH effects. cultured for 10 or 14 days, the total number of colonies NCS caused a dose-dependent stimulation of colony in the control group increased between the two observa- formation (Fig. 2B). No colony formation was observed 1846 GH AND CHONDROCYTE COLONY FORMATION Endo • 1986 Volll8«No5

8OO-1 160-1 o ahGH A AoPRL 150- sntro1

-ahGH 40 ng/ml •OContro 130- 600- •• ;§ 120

X 110- if)

I—( Downloaded from https://academic.oup.com/endo/article/118/5/1843/2540080 by guest on 30 September 2021 a \ 100- LJ 400- 90-1 I-H 2 0 20 40 60 80 100 120 140 160 O HDDED HORMONES (ng/ml) _J o u FIG. 4. Dose-response effect of hGH and oPRL on rat tibia chondro- cyte colony formation. Chondrocytes were isolated from epiphyseal growth plates of the proximal tibia of 28-day-old male rats and cultured at a density of 80,000 cells per dish in suspension. The suspension 200- culture was stabilized with agarose in F-12 medium containing 10% NCS with and without various concentrations of hGH or oPRL. Cul- tures were incubated for 14 days, and the number of colonies were counted using inverted light microscope. The stimulation is expressed as percent stimulation of control value (control value = 100%). Values are means ± SEM of 8 different cell isolations using different rats. (**, P < 0.01 calculated by Student's t test). oPRL but not hGH caused a 1 0- I I I significant stimulatory effect at 160 ng/ml (##, P < 0.01). 0 20 40 GO 80 NUMBER OF CELLS/DISH *1000 seeded cells (cloning efficiency), compared to 8.6 colonies FIG. 3. Effect of hGH on the colony formation of rat tibia epiphyseal chondrocytes at various cell densities. Chondrocytes were isolated from per 1000 cells in the GH group. The stimulatory effect epiphyseal growth plates of the proximal tibia of 28-day-old male rats of GH on colony formation was thus 41%. The dose- and cultured at various cell densities (10,000,20,000,40,000, and 80,000 response effect of GH on the colony formation of epi- cells per dish) in suspension culture stabilized with agarose in F-12 physeal chondrocytes is shown in Fig. 4, the threshold medium containing 10% NCS, with or without 40 ng/ml hGH. Cultures were incubated for 14 days, and the number of colonies were counted dose of hGH being 10 ng/ml and the maximal effective using inverted light microscope. Values are means ± SEM of 7 different dose being 40 ng/ml. The stimulatory effect of GH de- cell isolations using different rats. The effect of hGH at the 4 indicated creased as concentrations were increased and, at a con- cell densities was significant as calculated by 2-way analysis of variance centration of 160 ng/ml, no stimulatory effect was ap- parent (Fig. 4). oPRL failed to stimulate colony forma- in the absence of serum. A stimulatory effect of hGH tion at low concentrations (10-80 ng/ml) but caused an was apparent at a serum concentration of 10% or 15% effect at 160 ng/ml, probably due to intrinsic GH activity after 14 days of culture. However, after 21 days of culture of oPRL or ovine GH (oGH) contamination of this in the presence of 5% NCS, hGH (40 ng/ml) caused a preparation. The hGH concentration (50 ng/ml added to significant stimulatory effect (135 ± 4.2% over control medium at start of culture) showed no significant de- value, P < 0.01). These observations suggest that the stimulatory effect of GH on colony formation was de- crease during a culture period of 14 days (day 1, 50.4 ± pendent upon the simultaneous presence of serum factors 9.5 ng/ml; day 7, 39.4 ± 1.8 ng/ml; day 14, 40.9 ± 3.9 ng/ and that the expression of the stimulatory effect of GH ml. Values represent means ± SEM of 3 different exper- at low serum concentrations was dependent upon the iments). time of culture. The effect of 40 ng/ml hGH on colony size distribution The relationship between the number of colonies was also determined. As seen in Fig. 5 the size distribu- formed and the number of seeded cells is shown in Fig. tion of chondrocyte colonies varied between the control 3. These two parameters correlated well both in the and the GH-treated groups. The number of large colonies absence and presence of hGH, suggesting that hGH (diameter 90 nm or more) was overrepresented in the effects on colony formation was not dependent upon the GH group. When expressed as percent stimulatory effect cell concentration. The slope of the regression line in the of total numbers of colonies of the same colony size, the control group corresponds to 6.12 colonies per 1000 number of colonies in the GH group was 55-65% of the GH AND CHONDROCYTE COLONY FORMATION 1847

cytes and provides further support for the contention 200- CH--DhGH 40 ng/ml that GH stimulates longitudinal bone growth directly (1- 180- O OContro 1 3, 6, 14). Interestingly, direct effects of GH on sulfate

_ONIE S 160- incorporation, but not cell replication, has earlier been O 140- observed in isolated sternal chondrocytes from 13-day u. o 120- old chick embryos (15). 100- Colony numbers are frequently estimated when growth 80- and differentiation of hematopoietic stem cells are stud- ied (16). Colony counting assays and determination of 60- NUMBE R 3 . _i [ H]thymidine incorporation show a good correlation CE 40- (17), validating the use of optical colony counting for Downloaded from https://academic.oup.com/endo/article/118/5/1843/2540080 by guest on 30 September 2021 O t- 20- n determination of cell proliferation. Stimulatory effects 0- of GH on colony formation of cells from nonskeletal 40 60 80 100 120 140 160 tissues have been demonstrated previously by Golde and COLONY SIZE (urn) co-workers (16, 18, 19). These investigators showed that GH potentiated the colony formation of erythropoietin- U 70-, stimulated, hematopoietic, progenitor cells and stimu- M lated colony formation of human peripheral blood T m progenitor cells when studied in a suspension culture 3o 60-i stabilized with methyl cellulose. Interestingly, they (16) found the same type of biphasic dose-response effect with GH as that observed in the present study using 50-' growth plate chondrocytes. We have no satisfactory ex- planation for the reduced potentiation of colony forma- tion with high concentrations of GH, but one plausible 40- explanation might be that the effect of GH is dependent on the stage of cell differentiation. Since chondrocytes were isolated from whole growth plates in the present 30J study, it is likely that the isolated chondrocytes were 40 BO 80 100 120 140 160 heterogenous in terms of development and differentia- COLONY SIZE (urn) tion. FIG. 5. Effect of hGH on distribution of chondrocyte colony size. By dissecting chondrocytes from the different matu- Chondrocytes were isolated from rat epiphyseal growth plates of the ration zones of rabbit epiphyseal growth plates and cul- proximal tibia of 28-day-old male rats and cultured at a density of 80,000 cells per dish in suspension. The suspension culture was stabi- turing them in a monolayer, Corvol and co-workers (20) lized with agarose in F-12 medium containing 10% NCS with and found that cells from different zones of the growth plate without 40 ng/ml hGH. Chondrocyte colony size (diameter in microm- showed different patterns of growth. Thus, chondrocytes eters) was measured using an inverted light microscope with calibrated isolated from the proliferative zone formed colonies con- ocular. A total of 974 colonies were counted (423 control and 551 hGH) sisting of a large number of cells growing in several from 5 different cell isolations using different rats. These colonies were arranged into 5 different groups according to size. A, Colony size layers, in contrast to chondrocytes from the resting zone distribution of the total number of colonies. The size distribution of that developed as a monolayer. The finding in the present colonies was significantly different between the control and the hGH- study that GH potentiated colony formation and pref- treated groups, as calculated by x2 test (P < 0.05). B, Distribution of erentially increased the formation of large size colonies colonies in the control and hGH-treated groups expressed as percent suggests that GH stimulated stem cell chondrocytes or of the total number of colonies in each size group. Values are means ± early proliferative chondrocytes with an inherent high SEM. *, P < 0.05; **, P < 0.01 calculated by paired t test. capacity to proliferate. The subsequent clonal growth total number compared to 35-45% in the control group might be dependent upon the production of insulin-like at a colony size of 90 /im or more. growth factors in the chondrocytes themselves that pro- mote the clonal growth by paracrine or autocrine mech- Discussion anisms. Indeed, it has recently been demonstrated that neutralization of endogenously produced somatomedin- The present investigation showed that GH potentiated like peptides inhibits replication of cultured human fi- the colony formation of cultured chondrocytes isolated broblasts and smooth muscle cells giving direct experi- from rat epiphyseal growth plates. This observation dem- mental evidence for a paracrine/autocrine mechanism of onstrates that GH acts directly on growth plate chondro- action for insulin-like growth factors (21). 1848 GH AND CHONDROCYTE COLONY FORMATION Endo • 1986 Vol 118* No 5 The discovery by Green and co-workers (22, 23) that rabbit ear and rat rib growth cartilage. Nature 304:545 7. Madsen K, Makower AM, Friberg U, Eden S, Isaksson O 1985 GH stimulates the differentiation of preadipose cells Effect of human on proteoglycan synthesis in from a cloned 3T3 preadipose line and multinucleated cultured rat chondrocytes. Acta Endocrinol (Copenh) 108:338 muscle cells from myoblasts of a cloned muscle line 8. Sissons HA 1971 The growth of bone. In: Bourne GH (ed) The Biochemistry and Physiology of Bone, ed 2. Academic Press, New suggests that cell differentiation might be an important York, vol 3:145 regulatory mechanism for the stimulatory effect of GH. 9. Kember NF 1978 Cell kinetics and the control of growth in long According to this hypothesis, GH does not stimulate cell . Cell Tissue Kinet 11:477 10. Benya PD, Shaffer JD 1982 Dedifferentiated chondrocytes reex- proliferation directly but does so indirectly by making press the differentiated collagen phenotype when cultured in aga- cells at an early stage of development competent to rose gels. Cell 30:215 proliferate (24). We have proposed a similar mechanism 11. Yasui N, Osawa S, Ochi T, Nakashima H, Ono K 1982 Primary

culture of chondrocytes embedded in collagen gels. Exp Cell Biol Downloaded from https://academic.oup.com/endo/article/118/5/1843/2540080 by guest on 30 September 2021 for the stimulatory effect of GH on longitudinal bone 50:92 growth (2, 3, 25). The present study does not exclude the 12. Kiernan JA 1981 Histological and Histochemical Methods: Theory possibility that insulin-like growth factors, somatome- and Practice. Pergamon Press, Toronto, p 162 dins, play a role for the clonal growth of chondrocytes in 13. Jansson J-0, Albertsson-Wikland K, Eden S, Thorngren K-G, Isaksson 0 1982 Effect of frequency of growth hormone adminis- colonies in vitro or the proliferation of chondrocytes in tration on longitudinal bone growth and body weight in hypophy- the growth plate in vivo. In fact, we have recently ob- sectomized rats. Acta Physiol Scand (Copenh) 114:261 served, using immunohistochemistry, that proliferative 14. Isaksson OGP, Eden S, Jansson J-0 1985 Mode of action of pituitary growth hormone on target cells. Annu Rev Physiol 47:483 chondrocytes in rat epiphyseal growth plates produce 15. Meier S, Solursh M 1972 Stimulation of sulfate incorporation by IGF-like immunoreactive material and that GH regulates growth hormone treatment of cultured chick embryo chondrocytes. the number of chondrocytes in the growth plate that Gen Comp Endocrinol 18:89 16. Golde DW, Bersh N, Li CH 1977 Growth hormone: Species-specific contain IGF-I-like immunoreactivity (26). Furthermore, stimulation of erytropoiesis in vitro. Science 196:1112 cells in cultured chondrocyte colonies contain IGF-I-like 17. Jones CA, Tsukamoto T, O'Brien PC, Uhl CB, Alley MC, Lieber immunoreactivity (Lindahl, A., A. Nilsson, and 0. Isaks- MM 1985 Soft agarose culture human tumour colony forming assay for drug sensitivity testing (3H)-thymidine incorporation vs colony son, to be published), directly demonstrating that prolif- counting. Br J Cancer 52:303 erating chondrocytes are capable of producing IGF-I-like 18. Golde DW 1979 In vitro effects of growth hormone. In: Pecile A, substances in vitro. These observations give further sup- Miiller EE (eds) Growth Hormone And Other Biologically Active Peptides. Excerpta Medica, Amsterdam, vol 495:52 port for the theory that insulin-like growth factors, sup- 19. Golde DW 1980 Growth factors. Ann Intern Med 92:650 ports clonal chondrocyte growth via autocrine or para- 20. Corvol MT, Dumontier MF, Rappaport R 1975 Culture of chon- crine mechanisms, as suggested earlier (3, 27). drocytes from the proliferative zone of epiphyseal growth plate cartilage from prepubertal rabbits. Biomedicine 23:103 21. Clemmons DR, Van Wyk JJ 1985 Evidence for a functional role of endogenously produced somatomedinlike peptides in the regulation References of DNA synthesis in cultured human fibroblasts and porcine smooth muscle cells. J Clin Invest 75:1914 1. Isaksson 0, Jansson J-0, Gause IAM 1982 Growth hormone stim- 22. Morikawa M, Nixon T, Green H 1982 Growth hormone and the ulates longitudinal bone growth directly. Science 216:1237 adipose conversion of 3T3 cells. Cell 29:783 2. Isgaard J, Nilsson A, Lindahl A, Jansson J-0, Isaksson 0 1986 23. Nixon BT, Green H 1984 Growth hormone promotes the differ- Effect of local administration of GH and IGF-I on longitudinal entiation of myoblasts and preadipocytes generated by azacytidine bone growth in the rat. Am J Physiol 250 (Endocrinol Metab:l3) treatment of 10T V2 cells. Proc Natl Acad Sci USA 81:3429 3. Isaksson 0, Eden S, Albertsson-Wikland K, Jansson J-0, Friberg 24. Green H, Morikawa M, Nixon T 1985 A dual effector theory of U, Madsen K, Direct action of growth hormone on cartilage me- growth hormone action. Differentiation 29:195 tabolism. In: Raiti S (ed) Human Growth Hormone Symposium. 25. Isaksson OGP, Eden S, Jansson J-O, Lindahl A, Isgaard J, Nilsson Plenum Press, New York, in press A, Sites of action of growth hormone on somatic growth. J Anim 4. Russel SM, Spencer EM 1985 Local injections of human or rat Sci, in press growth hormone or of purified human somatomedin-C stimulate 26. Nilsson A, Isgaard J, Lindahl A, Dahlstrom A, Isaksson OGP 1985 unilateral tibial epiphyseal growth in hypophysectomized rats. Growth hormone increases the number of chondrocytes containing Endocrinology 116:2563 IGF-1/Sm-C immunoreactive material in the rat epiphyseal growth 5. Eden S, Isaksson OGP, Madsen K, Friberg U 1983 Specific binding plate. Acta Physiol Scand 124 [Suppl 542]:323 (Abstract) of growth hormone to isolated chondrocytes from rabbit ear and 27. D'Ercole AJ, Stiles AD, Underwood LE 1984 Tissue concentration epiphyseal plate. Endocrinology 112:1127 of somatomedin C: further evidence for multiple sites of synthesis 6. Madsen K, Friberg U, Roos P, Eden S, Isaksson 0 1983 Growth and paracrine or autocrine mechanisms of action. Proc Natl Acad hormone stimulates the proliferation of cultured chondrocytes from Sci USA 81:935