Retinoic acid modulation of mrna levels in malignant, nontransformed, and immortalized osteoblasts

JOURNAL OF BONE AND MINERAL RESEARCH Volume 6, Number 7, 1991 Mary Ann Liebert, Inc., Publishers

Retinoic Acid Modulation of mRNA Levels in Malignant, Nontransformed, and Immortalized Osteoblasts

HONG ZHOU,’ R. GLENN HAMMONDS, JR.,2 DAVID M. FINDLAY,’ PETER J . FULLER,3 T. JOHN MARTIN,’ and KONG WAH NG’

ABSTRACT

Clonal cell lines presumably “arrested” at a particular stage of differentiation are useful models to study the processes of differentiation in osteoblasts. UMR-201 is a presumptive preosteoblastic nontransformed rat clonal cell line with a limited life span in culture. Two immortalized cell lines, UMR-201-1OA (10A) and UMR-201-1OB (lOB), were derived from UMR-201 by stable transfection with simian virus (SV) 40 large T antigen. This study compares the growth and profile of gene expression of the immortalized cell lines with those of UMR-201 and UMR-106-06, a rat clonal cell line with well-defined osteoblast-like phenotypic characteristics. All four cell lines constitutively expressed the mRNA for the y, a, and /3 receptors for retinoic acid (RA), the growth hormone receptor, pro-al(1) collagen, osteonectin, bone proteoglycan I, and bone morphogenetic proteins (BMP) 1 and 2A. Alkaline phosphatase mRNA was absent in the preosteoblast cell lines but was induced by treatment with M RA, which also increased the steady-state levels of mRNA for osteopontin and BMP1. mRNA for matrix gla protein was constitutively present and further induced by RA in UMR-201 and 10B only. Messenger RNA for bone sialoprotein and bone morphogenetic protein 3 were constitutively expressed in UMR-106-06 and UMR-201 but absent in the immortalized cell lines. None of the cell lines expressed measurable mRNA for bone gla protein or bone proteoglycan 11. 10B grew more rapidly than UMR-201, but unlike UMR-201, it was also able to proliferate in serum-free medium and ex- hibit anchorage-independent growth. In summary, this study identifies novel retinoic acid effects on gene expression in these cells. Differences noted in the expression of mRNAs between UMR-106-06 and the other cell lines may provide some insight into the sequence of expression of these phenotypic characteristics as osteoblasts differentiate.

INTRODUCTION

URRENT VIEWS of cellular interactions in bone have de- C veloped from in vitro studies of the actions of osteo- tropic hormones on several cell lines with various properties expected of osteoblasts. The same approach has been useful in the study of osteoblastic differentiation, where it is presumed that there is a developmental sequence leading from uncommitted precursor cells to committed preosteoblasts, osteoblasts, and possibly osteocytes. An important step in the study of the osteoblast lineage was the establish-

ment in culture of clonal lines from cells isolated from bone or bone turn or^.^'-^) This enabled investigators to de- fine more precisely the heterogeneity of bone cell popula- tions and to determine the influence of regulatory factors, such as transforming growth factor P I s ) and 1,25-dihydroxyvitamin D, [ 1 ,25-(OH),D,],(6) on the expression of characteristics associated with the osteoblast phenotype using the clonal cell lines as models representing different stages of differentiation. This is based on the assumption that the clonal malignant cell lines represent cells “arrested” at a particular point in the developmental sequence. 1’)

~~

‘Department of Medicine, University of Melbourne and St. Vincent’s Institute of Medical Research, St. Vincent’s Hospital, Mel-

’Genentech, Inc., San Francisco, California. ’Institute of Medical Research, Prince Henry’s Hospital, Melbourne, Australia.

bourne, Australia.

767

ZHOU ET AL. 768

UMR-201 cells are nontransformed clonal cells derived from noeonatal rat calvariae with features suggestive of preosteoblasts. For example, they have a very low basal alkaline phosphatase activity, which is increased significantly by retinoic a ~ i d . ' ~ , ~ ) Because of the finite life span of UMR-201 cells in culture, it was considered advantageous to have immortalized UMR-201 cells for future studies provided that the phenotypic characteristics of interest were not substantially altered in the process. Immortalization was carried out by transfecting the cells with SV40 large T antigen, which has been widely used to immortalize primary cells from a wide variety of tissues and species. In those studies, there have been mixed results with respect to the maintenance of the original phenotype. A different approach was used in the present work in which a well-characterized clonal cell line, UMR-201, was used to obtain immortalized cells. This offered the dual advantages of a less heterogeneous starting point than mixed primary cell cultures and further allows careful comparison of phenotypic characteristics between the parent and immortalized cell lines.

The present study was designed to compare, in several osteoblast model systems, the growth, profile of gene expression of osteoblast-related phenotypic characteristics, and their response to retinoic acid (RA). The cell lines comprised two immortalized cell lines, 10A and 10B, their parent cell line, UMR-201, and UMR-106-06, a clonal cell line derived from a rat osteogenic sarcoma with well-characterized osteoblast-like features,' lo) representing perhaps a later stage of osteoblast differentiation. These studies of mRNA expression and effects of RA were embarked upon because of the growing evidence that RA treatment is capable of promoting marked changes in differentiated properties of osteoblasts. The effects of RA on growth inhibi- tion, cytoskeletal structures, morphology, and some differentiated functions in malignant osteoblast-like cells have been and subsequent observations in presumptive preosteoblasts identified very substantial changes in the expression of specific gene^."^.^^) Although the production of type 1 collagen and elevated alkaline phosphatase (ALP) activity are generally accepted as typi- cal properties of osteoblasts, the development of nonenzy- matic procedures to isolate the organic constituents of bone matrix has resulted in the identification of a number of noncollagenous proteins synthesized by bone cells (for review, see Ref. 16). Even though the functions of these noncollagenous proteins remain speculative, they neverthe- less must collectively play an important role in osteogene- sis. It was considered important to study the profile of expression of osteoblast-associated mRNAs in preosteoblastic cell lines and malignant osteoblasts and to determine whether their production could be modulated by RA. The ultimate aim of these studies is to enable the specific expression of various bone proteins to be related to their presumed functions in bone.

MATERIALS AND METHODS Transfections with SV40 large T antigen

The plasmid containing DNA coding for the SV40 large

T antigen, pRSVTag, was constructed by Dr. C. Gorman, Genentech Inc., San Francisco. The long terminal repeat of the Rous sarcoma virus serves as a strong eukaryotic promoter for this plasmid, which was derived from the plasmid pRSV-CAT by exchanging the appropriate protein-coding region. (I7) The plasmid containing DNA coding for the neomycin resistance gene, pRSVNeo, was constructed similarly. Both plasmids were gifts of Dr. Andrew Baker, Genentech. UMR-201 cells in passage 8 were grown to approximately 70% confluence in 100 mm plastic culture plates with a 5O:SO (vol/vol) mixture of Dulbecco's modified Eagle's medium (DMEM) and Ham's F12 medium containing 10% fetal bovine serum (FBS) and freshly supplemented with glutamine to 2 mM. Medium was changed 1 h before the transfection and the cells reequili- brated at 37°C. The cells in each plate were transfected by the calcium phosphate method with 1 pg pRSVNeo plus 10 pg pRSVTag. After 4 h, each plate was glycerol shocked and incubated overnight at 37°C. Cells were subcultured and incubated with medium containing 0.8 mg/ml geneticin (G418, GIBCO, Grand Island, NY). All medium used subsequently contained geneticin. After 8 days, there were no viable cells remaining in control cultures that were not transfected with pRSVNeo. Plates containing single colonies of G418-resistant cells were grown to confluence and subcultured sequentially for 5 weeks before aliquots were frozen in liquid nitrogen. We obtained four clonal cell lines, designated UMR-201-1OA (lOA), UMR-201-1OB (lOB), UMR-201-1OC (lOC), and UMR-201-11B (llB), which had been cotransfected with pRSVNeo and pRSVTag.

Cell cultures

UMR-201 and UMR-106-06 cells were routinely grown in a-modified MEM (a-MEM) containing 10% FBS. 10A and 10B cells were grown in the same medium with the addition of 0.8 mg/ml of (3418. Incubation was carried out at 37°C and equilibrated with 5% C 0 2 in air.

In the growth studies, UMR-201 and 10B were subcultured into 9.6 cm' six-plate multiwell dishes in a-MEM containing 10% FBS. In one set of experiments, the medium was changed 24 h later to a-MEM with 2% vitamin A-depleted FBS with or without M RA. In another set of experiments, the medium was a-MEM containing 0.1070 bovine serum albumin (BSA). Medium was changed every second day, with fresh RA added when appropriate. Cell count was performed on days 1, 3, and 5 on replicate wells.

Anchorage-independent growth

Single-cell suspensions prepared from UMR-201, lOA, and 10B were plated at 2000 cells per 35 mm dish in a double-layer nutrient agar system.(L8) The medium was a- MEM with a final concentration of 10% FBS, and the dishes were incubated for 14 days at 37°C in sealed plastic boxes gassed with 5% oxygen, 10% CO,, and 85% nitrogen. At the end of the incubation period, the colonies were stained with 2-(p-iodophenyl)-3-(p-nitrophenyl)-5-

EFFECTS OF RETINOIC ACID ON IMMORTALIZED PREOSTEOBLASTS 769

phenyltetrazolium chloride hydrate (1 mg/ml; 200 p1 per dish) and left overnight at 37°C. Stained colonies were counted with a Bioquant I1 image analyzer system (R & M Biometrics, Nashville, TN). Only colonies with diameters exceeding 50 pm were counted. Statistical analyses were carried out using Student’s t-test.

ALP activity

ALP activity was measured by the hydrolysis of p-nitro- phenol phosphate as previously described,‘8) and the specific activity of ALP was expressed as pmoles of p-nitro- phenol released per hour per mg protein. In the experiments, cells were subcultured into six-plate multiwell dishes in a-MEM containing 10% FBS. After 24 h medium was changed to a-MEM with 2% vitamin A-depleted FBS with increasing doses of RA. Medium was changed every second day and fresh RA added. Cells from replicate wells were assayed for ALP specific activity on days 1, 3, and 5.

Northern blot analysis

Cells were subcultured into 175 cm2 tissue culture flasks in a-MEM with 10% FBS until they were almost conflu- ent. The medium was then changed to a-MEM with 2% vitamin A-depleted FBS and M RA added to the appropriate flasks for a further 24 h incubation. Total RNA was isolated with guanidine thiocyanate as described,(I9) separated in a 1.5% agarose-formaldehyde gel, and trans- ferred to nylon filters.(2o) Filters (Amersham, United King- dom) were hybridized overnight in buffer containing 50% formamide, 5-fold SSPE (SSPE contains 0.15 M NaCI, 0.01 M NaH2P04, and 0.001 M EDTA), 5-fold Denhardt’s solution, 0.5% sodium dodecyl sulfate (SDS), and 1 % skimmed milk powder. The filters were washed sequentially in 2-fold SSPE with 0.1% SDS at 42°C for 15 minutes, SSPE with 0.1% SDS at 65°C for 30 minutes, and finally 0.1-fold SSPE with 0.1% SDS at room temperature for 15 minutes. Specifically bound probe was visualized by autoradiography.

Complementary DNA probes were nick translated with [32P]a-dCTP to a specific activity of 1 x lo9 dpm/pg DNA according to the manufacturer’s instructions (Boehr- inger Mannheim GmbH, Mannheim, Germany). The full- length 2.4 kb rat cDnA for alkaline phosphatase was a gift from Dr. G. Rodan (Merck, Sharp and Dohme Research Laboratories, West Point, PA). The human cDNAs for bone sialoprotein (BSP), osteopontin (OP), biglycan (bone proteoglycan I, PG I) were gifts of Dr. L. Fisher (National Institute of Dental Research, NIH, Bethesda, MD). Dr. J . Wozney (Genetics Institute, Cambridge, MA) generously provided the human cDNAs for bone morphogenetic proteins 1, 2A, and 3 (BMP1, BMP2A, and BMP3), as well as the rat cDNA for bone gla protein (BGP). Rat cDNA for matrix gla protein (MGP) was the gift of Dr. P. Price (University of San Diego, La Jolla, CA). Dr. W. Wood (Genentech Inc., San Francisco) provided the rabbit cDNA for growth hormone receptor (GHR). Bovine osteonectin cDNA and chicken actin cDNA were the gifts of Dr. M. Young (National Institute of Dental Research,

NIH, Bethesda, MD), and Dr. J . Bateman (Royal Chil- dren’s Hospital, Melbourne, Australia) provided the rat cDNA for procrl(1) collagen. Professor P. Chambon (In- stitut de Chimie Biologigue, Strasbourg, France) generously provided the mouse cDNAs for the a, p , and y retinoic acid receptor (aRAR, PRAR, and yRAR).

RESULTS Growth in monolayer culture and soft agar

The growth of UMR-201 in a-MEM containing 2% vitamin A-depleted FBS in the presence or absence of M RA was compared to that of the transformed counterpart 10B (Fig. 1). At the same time, a comparison was also made of the ability of both cell lines to grow in a-MEM containing 0.1% BSA (Fig. 1, inset). Several differences were noted between the growth characteristics of 10B and UMR-201. First, 10B cells grew more rapidly than UMR-

1 1 3 5

Days

FIG. 1. Comparison of the growth of UMR-201 and 10B in monolayer culture. Cells were subcultured into 9.6 cm2 6-plate multiwell dishes in a-MEM containing 10% FBS for 24 h before medium was changed to a-MEM with 2% vitamin A-depleted FCS f M RA. (0) Untreated 10B cells; (W) 10B cells treated with M RA; (0) untreated UMR-201 cells; ( 0 ) UMR-201 cells treated with M RA. Medium was changed every second day and fresh RA added. Seeding density was 2.5 x lo4 cells per well. Repli- cate wells were harvested on the days indicated. (Inset) Growth in medium containing 0.1% BSA of 10B cells (0) compared to UMR-201 cells (0). Bars = SEM. Points are mean of four replicate wells. These experiments have been performed three times with similar results and a representative result is shown.

770 ZHOU ET AL.

201 cells in FBS-containing medium but, like UMR-201 cells, were inhibited in their growth by RA. Second, 10B cells but not UMR-201 cells were able to grow in serum- free medium. Third, we were able to grow 10B cells continuously in culture for a period of 10 months (over 80 passages), whereas the nontransformed UMR-201 cells undergo senescence after 12 passages in culture. Finally, the immortalized cell lines 10A and 10B formed significantly more colonies compared with UMR-201 in the stringent conditions imposed by soft agar (Table 1).

ALP specific activity

The results show that the immortalized cell lines 10A and 10B demonstrate a time- and dose-dependent induction of alkaline phosphatase activity in response to RA in a manner similar to the parent cell line UMR-201 (Fig. 2).

TABLE 1. ANCHORAGE-INDEPENDENT GROWTH OF UMR-201 COMPARED WITH 10A AND lOBa

No. of colonies Significance

UMR-201 8 2 2 1 OA 318 2 18 p < 0.001 10B 267 k 2 p < 0.001

aA total of 2000 cells from each clonal cell line were subcultured in triplicate into each 35 mm Petri dish in a double-layer nutrient agar system. The medium was a-MEM with a final concentration of 10% FBS. The dishes were incubated for 14 days, stained, and counted with a Bioquant I1 image analyzer system. Only colonies exceeding 50 pm were counted. Statistical analysis comparing the number of colonies formed by 10A and 10B with UMR-201 was carried out using Student's 1-test.

~

UMR201 { 0 D1

D3 0 D5

Northern blot analyses

Retinoic Acid Receptor (RAR): The effects of RA on growth and differentiation in normal and malignant osteoblasts is well established, but to date the RAR subtypes in bone cells have not been characterized. Using cDNA probes for the a, /3, and yRAR mRNA, all three species were found in each of the osteoblast-like cell lines but we were particularly interested in the demonstration of the y- RAR in cells of the osteoblast lineage following recent reports showing a specific spatial and temporal distribution of the yRAR during mouse embryogenesis.'2'.22) Our results show the presence of a 2.7 kb mRNA for the y receptor for RA in all the osteoblastic cell lines (Fig. 3) using a 32P-labeled plasmid containing a mouse yRAR cDNA as probe.(23) The addition of M RA for 24 h did not alter the steady-state level of yRAR mRNA, and this is similar to a previous report on the effects of RA on yRAR mRNA expression in F9 teratocarcinoma cells.(241

Alkaline Phosphatase and Osteopontin: Only UMR- 106-06 cells constitutively expressed the 2.4 kb mRNA for ALP.(2s) After 24 h treatment with M RA, ALP mRNA was induced in UMR-201, 10A, and 10B (Fig. 4) consistent with an increase in enzymatic activity of ALP in the same cells (Fig. 2). These results thus indicate that 10A and 10B have retained this particular phenotypic charac- teristic of UMR-201. A labeled 1.5 kb XhoI-XbaI fragment of human OP C D N A ( ~ ~ I hybridized to a 1.6 kb OP mRNA in all four cell lines. The steady-state level of OP mRNA in all cell lines was increased upon treatment with

M RA for 24 h (Fig. 4).

B"l -Log[M] retinoic acid

100

/ I 4'

1-Q / / -",o-c ,-

FIG. 2. Time- and dose-dependent response of ALP specific activity of UMR-201, 10A, and 10B to RA. Cells were subcultured into six-plate multiwell dishes as described in Materials and Methods. Seeding density was 5 x lo4 cells per well. Points are mean of six replicate wells. Bars = SEM. (0) Day 1; ( 0 ) Day 3; (0) Day 5. Results shown are those of a representative experiment that was performed three times.


FIG. 3. Detection of mRNA for y RAR in UMR-106-06, UMR-201, 10A, and 10B. Untreated cells (-); cells treated with M RA for 24 h (+). Total RNA (20 pg) was loaded into each lane. The filter was hybridized with a plasmid containing a mouse cDNA probe for y RAR. The arrowheads point to the 28s and 18s ribosomal bands, respectively. The filter was reprobed with 32P-labeled chicken actin cDNA to normalize the amount of RNA present in each lane.

FIG. 4. Detection of mRNA for ALP, OP, and MGP in the clonal cell lines. Untreated cells (-); cells treated with

M RA for 24 h (+). Total RNA (20 pg) was loaded into each lane. One filter was probed with a plasmid containing a 2.4 kb rat cDNA for ALP, washed, and reprobed with a 3aP-labeled chicken actin cDNA for normalization. Another filter was probed initially with a 1.5 kb XhoI- XbaI fragment of human OP cDNA, washed, and reprobed with a plasmid containing a rat MGP cDNA, washed again, and finally probed with a cDNA for chicken actin. The arrowheads point to the 28s and 18s ribosomal bands, respectively.

Matrix gla Protein and Bone gla Protein: A 700 bp MGP mRNA(6) was constitutively expressed by 10B and, to a much lesser extent, by UMR-201. Treatment with M RA resulted in a substantial increase in the steady-state level of MGP mRNA in 10B and UMR-201 (Fig. 4). The influence of RA on the gene expression of MGP mRNA has not been reported previously. Messenger RNA for MGP was not detected in UMR-106-06 or 10A either in the basal or in the RA-stimulated states. Messenger RNA for BGP was not detected in any of the cell lines (data not shown).

Pro-d(I) Collagen and Bone Morphogenetic Protein I : All the cell lines expressed the mRNA for pro-crl(1) collagen upon probing with a 1.6 kb PstI insert of rat pro-crl(1) collagen cDNA.Iz7) The steady-state levels of mRNA for pro-crl(1) were higher in UMR-106-06 and 10B. Treatment with M RA resulted in a decrease in the steady-state levels of pro-crl(1) collagen mRNA in UMR-106-06 but yielded an increase in UMR-201 (Fig. 5 ) . Each of the four cell lines constitutively expressed a minor 4 kb mRNA and a major 2.8 kb mRNA for BMPl on probing with a 2.5 kb XbaI-SalI fragment of human BMPl CDNA.(*~) The steady-state level of BMPl mRNA was raised after treat-

FIG. 5. Detection of mRNA for pro-crl(1) collagen and BMP 1 in the clonal cell lines. Untreated cells (-); cells treated with M RA for 24 h (+). The filter was initially probed with a 1.6 kb PstI insert of human pro-crl(1) collagen cDNA, washed, and reprobed with a 2.5 kb XbaI-SalI insert of human BMP 1 cDNA, washed again, and finally probed with a plasmid containing a chicken actin cDNA for normalization. The arrowheads point to the 28s and 18s ribosomal bands, respectively.

772 ZHOU ET AL.

ment for 24 h by M RA in UMR-201 and 10A, but a similar change was not seen in RA-treated UMR-106-06 or 10B (Fig. 5).

Changes in steady-state levels of mRNA in response to RA were observed only for the expression of ALP, pro- al(1) collagen, OP, MGP, and BMPl genes, respectively. The expression of the remaining genes to be discussed later was not influenced by RA.

Bone Morphogenetic Proteins 2A and 3: Filters were probed with a 1.6 kb EcoRI insert of human BMP2A cDNA and a 1.5 kb EcoRI-XBaI insert of human BMP3 cDNA, respectively.(281 UMR-201, 10A, and 10B expressed a single 4.0 kb mRNA species for BMP2A. UMR-106-06, on the other hand, expressed a minor 3.5 kb and a major 1.6 kb mRNA species (Fig. 6). In the case of BMP3, UMR- 106-06 and UMR-201 expressed two species of BMP mRNAs with major bands at 3.6 and 1.2 kb, respectively. mRNA for BMP3 was not detected in 10A or 10B (Fig. 6).

Growth Hormone Receptor, Bone Sialoprotein, Osteo- nectin, and Biglycan: All four cell lines expressed mRNA for GHR, hybridizing to two major bands (3.5 and 1.2 kb)

FIG. 6. Detection of mRNA for BMP 2, BMP 3, GHR, BSP, ON, and biglycan (PG I) in the clonal cell lines. No change was detected in the steady-state levels of the respective mRNAs following treatment with RA, and the corre- sponding lanes were omitted from the figure. The filters were individually probed with the following 32P-labeled cDNAs: a 1.6 kb EcoRI insert of human BMP 2A; a 1.5 kb EcoRI-XbaI insert of human BMP 3; a 638 bp BamH1- EcoRI fragment of rabbit GHR; a 1.6 kb EcoRI insert of bovine osteonectin; a plasmid containing a 1.2 kb insert of human BSP and a 1.1 kb XbaI-NaeI insert of human biglycan. The arrowheads point to the 28s and 18s ribosomal bands, respectively.

and one minor band at 2.4 kb (Fig. 6). The probe used was a 638 bp BamHI-EcoRI fragment of the coding region of the rabbit GHR C D N A . ‘ ~ ~ ’ Similar multiple mRNA species for GHR have been reported in mouse liver.(3o1 UMR-106- 06 and UMR-201 expressed multiple mRNA species for BSP when probed with the plasmid containing a 1.2 kb fragment of human BSP cDNA. Both cell lines hybridized to a major 1.2 kb and a minor 3.2 kb band. Messenger RNA for BSP was not detected in 10A and 10B. A single 2 kb mRNA species for ON was constitutively expressed by all four cell lines. The filter was probed with a 1.6 kb EcoRI insert of the bovine osteonectin cDNA.13” A 1.1 kb XbaI-NaeI insert of human P G I (biglycan) CDNA‘~*’ hybridized with a single 2.6 kb mRNA species in UMR-201, 10A, and 10B. In this particular filter, no P G I mRNA was detected in UMR-106-06. In other northern blot analyses using UMR-106-06, however, a faint 2.6 kb biglycan mRNA band can be detected provided the autoradiograph is allowed to develop for a t least 10 days, suggesting a low level of expression for biglycan mRNA in UMR-106-06 (data not shown).

The constitutive expression of the various mRNAs and their response to RA in the four clonal cell lines is summa- rized in Table 2.

DISCUSSION

Many transformed cell lines derived from primary cultures have maintained some differentiated characteristics of the cell population from which they were established. However, loss of differentiated phenotype has also been reported in many instances following SV40 large T antigen irnrnortalization.(33-3sl In situations in which SV40 immortalized cell lines were established from primary cultures, such as hepatocytes, diversity in gene expression and regulation of the immortalized cell lines has been well documented.(36.37’ In such cases, diversity could already have been present in the heterogeneous cell population from which the immortalized cells were derived. Our approach was to immortalize a well-characterized, nontransformed clonal cell line, thus ensuring a less heterogeneous starting material and also enabling us to compare phenotypic characteristics of the parent and immortalized cells. Altered growth requirement was the most obvious change observed in the immortalized cell lines when compared with UMR-201. The ability of 10A and 10B t o form colonies in the stringent conditions imposed by soft agar is a good indicator of their transformed state. In addition, the ability of these transformed counterparts to grow in serum-free medium is an obvious advantage since it makes it possible to use the cells to test systematically the effects of various agents on the expression of osteoblast-related phenotypic characteristics without interference from serum and to identify and quantitate substances secreted by the cells. We have grown 10B cells in culture continuously for 10 months, and they have not shown any qualitative or quantitative change in their ALP response to RA or their resistance to G418 (data not shown), thus suggesting that stable transfections have occurred. Some diversity was


TABLE 2. SUMMARY OF THE CONSTITUTIVE EXPRESSION OF OSTEOBLAST-RELATED MRNAs IN CLONAL RAT OSTEOBLASTIC CELL LINES AND THEIR RESPONSE TO RETINOIC ACID^

UMR- 106-06 UMR-201 IOA IOB

C RA C RA C RA C RA

Alkaline phosphatase + I ND 1 ND t ND 1 Pro-al(1) collagen Osteopontin Matrix gla protein Bone gla protein

- + - + 1 + 1 + + 1 + t + t + t

ND ND + t ND ND + t ND ND ND ND ND ND ND ND

~~ - Bone morphogenetic protein 1 + - + t + t + Bone morphogenetic protein 2A + - + - + - + - Bone morphogenetic protein 3 + - + - ND ND ND ND Osteonectin Bone proteoglycan I Bone sialoprotein

- + ND ND + - + - +

- + - + - +

+ - + - ND ND ND ND - - + - Growth hormone receptor + - + - + - Receptor for retinoic acid + - + - + - +

aC, untreated cells; RA, treatment with M retinoic acid for 24 h; +, the constitutive expression of mRNA (by northern blots); ND, mRNA not detected; the following symbols characterize the change in steady-state levels of mRNA in response to treatment with RA: f , increase; I , decrease; -, no change.

noted in the phenotypic characteristics when compared to UMR-201, such as the absence of mRNA for BSP and BMP3 in the immortalized cell lines. There were also cer- tain differences between 10A and 10B, for example in the expression of mRNA for MGP. It is possible that such diversity arose as a result of the process of immortalization and less likely that it could have been due to diversity already present in the population of clonal UMR-201 cells from which they were derived. Indeed, genetic studies involving karyotypic analysis and Southern analysis of inte- grated viral sequences in clonal SV40-immortalized human diploid fibroblasts have demonstrated both random and nonrandom alterations.(38)

Retinoic acid has been implicated as the natural morphogen responsible for pattern formation in developing chick limb It is now recognized that the retinoic acid receptors are members of the steroid and thyroid receptor superfamily that acts primarily through the direct modulation of gene transcription.(42) Recent studies have demonstrated the existence of three distinct retinoic acid receptors designated aRAR, PRAR, and yRAR.(Z3.43-46) The almost complete interspecies conservation of the amino acid sequences of the three RAR as well as their differential distribution in mouse tissue has raised the strong possibility that each member of the RAR subfamily may play a specific role during development and in the adult animal.(z31 Although aRAR and PRAR are present in each of the osteoblast-like cell lines examined (data not shown), we were particularly interested in determining the expression of yRAR in osteoblasts because of recent reports of specific spatial and temporal distribution of the y RAR during embryogenesis. In the mouse embryo, yRAR transcripts were demonstrated by in situ hybridization in the

presomitic posterior region as early as 8 days postcoitum. By 13.5 days the transcripts were localized in cartilage, dif- ferentiating squamous epithelia and teeth.”’.’’’ These data persuasively indicated a distribution of y RAR mRNA transcripts restricted to areas related to bone and teeth formation or squamous epithelial and augment previous studies that showed that expression of yRAR was limited to skin, the F9 teratocarcinoma, and breast cancer T47D cell

These findings imply an important role for y- RAR in the transduction of RA signals at the level of gene expression during morphogenesis and differentiation of squamous epithelia. In the context of this paper, the demonstration of the yRAR in chondrogenesis provides the es- sential link between this receptor and endochondral bone formation. Our demonstration, for the first time, that the osteoblast cell lines express the y RAR has the important implication that it plays an equally vital role in osteoblast differentiation.

Osteopontin is a phosphorylated glycoprotein found in bone, kidney, and placental de~idua.‘~’) The amino acid sequence is characterized by the presence of a cell attachment RGD (Arg-Gly-Asp) sequence as well as a stretch of nine aspartate residues that may enhance binding to hydroxyapatite There has not been a previous report on the regulation of OP gene expression by RA; other investigators have reported the regulation of OP gene expression by transforming growth factor /3,(s) 1,25- (OH)2D3,(49) parathyroid hormone,(s0) H-r~s,(~‘~ and phorbol ester.(sZ)

BSP is the other noncollagenous protein expressed by the cell lines that has a RGD sequence in its peptide thought to be involved in cell attachment through the vitronectin r e ~ e p t o r . ( ~ ~ . ~ ~ ) Unlike OP, BSP mRNA was not

774 ZHOU ET AL.

stimulated by RA in our cell lines. However, it has been reported that mRNA for BSP was increased in a rat osteogenic sarcoma cell line (ROS 17/23) in response to dexa- m e t h a ~ o n e . ( ~ ~ ) The species of BSP mRNA expressed by our cell lines is different from the 2.0 kb mRNA species that was previously d e ~ c r i b e d ( ~ ~ , ~ ~ ) thus suggesting the differential expression of more than one species of mRNA for BSP in separate cell lines.

Matrix gla protein and BGP are two evolutionarily related proteins characterized by the presence of vitamin D-dependent y-carboxyglutamic residues postulated to play a role in the regulation of hydroxyapatite crystal growth.(S6) They differ in structure, physical properties, tissue distribution, and the sequence in which they appear in development. In the rat femur, for instance, MGP is present at birth but BGP reaches adult levels only after 3 weeks of thus suggesting different functions in developing bone. Steroid hormones obviously play important roles in the regulation of gene transcription of MGP and BGP. 1 ,25-(OH),D3 increases steady-state levels of mRNA for both MGP and BGP in various rat osteogenic sarcoma cell line^,(^^.^^) and the secretion of BGP in vivo is diminished by glucocorticoids. ( 6 0 ) Although it was previously reported that the expression of MGP and BGP was mutually exclusive in rat osteosarcoma cell lines, ( 6 ) current evidence suggests that this is no longer the case, at least in the rat ROS 17/2 cells.(58) Our demonstration that MGP mRNA is induced by RA should add impetus to investigations of the interplay between steroid hormones in the regulation of these genes, particularly since it has recently been shown that a common responsive element in the human BGP gene is recognized by the oncogene product Jun- Fos as well as receptors for vitamin A and D.(61)

The mRNA for osteonectin was constitutively expressed, although in some experiments, as reported previously, a slight increase in ON mRNA level was observed with RA treatrnent.(ls) Biglycan mRNA was much more abundant in UMR-201, lOA, and 10B compared to UMR-106-06.

The regulation of ALP gene transcription is compli- cated. In the present study and in published reports, RA stimulated ALP enzymatic activity and induced the ALP mRNA in clonal osteoblast cell lines that have very low or undetectable ALP activity.(9.’4) In contrast, RA inhibited the constitutive ALP activity in cells generally regarded as representatives of differentiated osteoblasts, such as UMR-106-06, ROS 17/23, and late-passage calvarial cells.(8.62) A similar paradoxical pattern was observed in our study with the expression of pro-crl(1) collagen mRNA. RA treatment resulted in a decrease in the steady- state levels of pro-a(I) collagen mRNA in UMR-106-06 but yielded a slight increase in UMR-201. The decrease in the steady-state levels of pro-al(1) collagen mRNA in UMR- 106-06 following treatment with RA is similar to that reported for skin fibroblast A rise in the steady- state level of pro-al(1) collagen mRNA in response to RA was reported in immortalized RCT-1 cells, a clonal osteoblast cell line with characteristics very similar to those of UMR-201 .(14) This could imply that transcription of the ALP and pro-crl(1) collagen genes may be influenced directly by the binding of RAR to the promoter elements of the re-

spective genes or indirectly by nuclear factors induced by RA. The presence of these unknown factors may in turn be determined by the state of differentiation of the cells.

The isolation, cloning, and expression of a family of BMP from a human osteosarcoma cDNA library was a sig- nificant breakthrough following the pioneering work of Urist, who showed that endochondral bone formation can be stimulated in vivo by protein extracts derived from bone.(z8.64-66) Each of the three polypeptides expressed (BMP1, 2A, and 3) was capable of inducing cartilage formation in vivo, and BMP2A has subsequently been shown to stimulate bone formation after the stage of cartilage formation.(67) BMPl is a novel protein whose structure is different from that of BMP2A and BMP3, which have been identified as members of the transforming growth factor p supergene family. We now report that the steady- state level of BMPl mRNA is increased by RA, thus suggesting a way in which RA may directly influence endochondral bone formation. This is the first report of modulation of gene expression of any of the BMP genes that have been cloned. The different mRNA species for BMP2A expressed by UMR-106-06 and the other cell lines could be due to alternative splicing of the gene or to the expression of a closely homologous mRNA species, such as BMP2B.(28) Whether this is related to the different states of differentiation or to different sites of origin of the cells is not known.

The mRNA for GHR was constitutively expressed by all four cell lines. We have reported the presence of a single class of receptors for growth hormone in UMR-106-06 with an affinity constant of 1.2 + 0.4 x lo9 M-*. These receptors mediate a proliferative response to growth hormone in UMR-106-06 and UMR-201 cells and is thus consistent with a role for growth hormone in osteoblast func- tion,168) either directly, or indirectly acting through the regulation of the synthesis of other factors, such as the in- sulin-like growth factor I . ( 6 9 )

In the present study, we have added OP, MGP, and BMPl to the list of genes whose expressions are regulated by retinoic acid. In the remaining genes, no change was observed in the steady-state levels of the respective mRNAs, and further investigations involving detailed time courses and dose responses are currently being undertaken to determine possible variations in their response to RA.

In conclusion, we have succeeded in establishing in culture immortalized cell lines capable of the continued expression of most of the differentiated characteristics typi- cal of the clonal cells from which they were derived. Dif- ferences noted in the profile of gene expression between UMR-106-06 and the other cell lines may provide some insight into the coordinate appearance of those phenotypic characteristics as osteoblasts differentiate. Finally, we have broadened the profile of genes whose expression is regulated by RA. The expression of mRNA for y R A R in these bone-derived cell lines points to its importance as a trans- ducer of the actions of RA in bone. That we were able to show that expression of a substantial number of osteoblast-related genes was regulated by RA points to its cen- tral role in osteoblast differentiation.


ACKNOWLEDGMENTS

The authors gratefully acknowledge the generosity of various investigators listed in Materials and Methods who provided the cDNAs that enabled us to carry out this work. This project was supported by grants from the Na- tional Health and Medical Research Council, Australia and the Anti-Cancer Council of Victoria.

REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

Aubin JE, Heersche JNM, Merrilees MJ, Sodek J 1982 Isola- tion of bone cell clones with differences in growth, hormone responses, and extracellular matrix production. J Cell Biol

Partridge NC, Alcorn D, Michelangeli VP, Ryan GB, Martin TJ 1983 Morphological and biochemical characterization of four clonal osteogenic sarcoma cell lines of rat origin. Cancer Res 43:4308-4314. Sudo H, Kodama H, 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 %:191-198. Majeska RJ, Rodan GA 1984 Culture and activity of osteoblasts and osteoblast-like cells. In: Butler WT (ed) The Chemistry and Biology of Mineralized Tissues. Ebsco Media, Inc., Birmingham, Alabama, pp. 279-285. Noda M, Yoon K, Prince CW, Butler WT, Rodan GA 1988 Transcriptional regulation of osteopontin production in rat osteosarcoma cells by type beta transforming growth factor. J Biol Chem 263:13916-13921. Fraser JD, Otawara Y, Price PA 1988 1,25-Dihydroxyvita- min D, stimulates the synthesis of matrix gla protein by osteosarcoma cells. J Biol Chem 263:911-916. Greaves MR 1986 Differentiation-linked leukomogenesis in lymphocytes. Science 234:697-704. Ng KW, Gummer PR, Michelangeli VP, Bateman JF, Mas- cara T, Cole WG, Martin TJ 1988 Regulation of alkaline phosphatase expression in a neonatal rat clonal calvarial cell strain by retinoic acid. J Bone Miner Res 353-61. Ng KW, Hudson PJ, Power BE, Manji SS, Gummer PR, Martin TJ 1989 Retinoic acid and tumour necrosis factor-a act in concert to control the level of alkaline phosphatase mRNA. J Mol Endocrinol 357-64. Forrest SM, Ng KW, Findlay DM, Michelangeli VP, Livesey SA, Partridge NC, Zajac JD, Martin TJ 1985 Characteriza- tion of an osteoblast-like clonal cell line which responds to both parathyroid hormone and calcitonin. Calcif Tissue Int

Thein R, Lotan R 1982 Sensitivity of cultured human osteosarcoma and chondrosarcoma cells to retinoic acid. Cancer Res 42:4771-4775. Livesey SA, Ng KW, Collier GR, Kubota M, Steiner AL, Martin TJ 1985 Effect of retinoids on cellular content and human parathyroid hormone activation of cyclic adenosine 3’:5’-monophosphate-dependent protein kinase isoenzymes in clonal rat osteogenic sarcoma cells. Cancer Res 455734- 5740. Ng KW, Livesey SA, Collier F, Gummer PR, Martin TJ 1985 Effect of retinoids on the growth, ultrastructure, and cytoskeletal structures of malignant rat osteoblasts. Cancer Res

Heath JK, Rodan SB, Yoon K, Rodan GA 1989 Rat calvarial cell lines immortalized with SV-40 large T antigen: Constitu- tive and retinoic acid-inducible expression of osteoblastic

92:452-461.

37~51-56.

495106-51 13.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25,

26.

27.

28.

29.

30.

3 1 .

features. Endocrinology 124:3060-3068. Ng KW, Manji SS, Young MF, Findlay DM 1989 Opposing influences of glucocorticoid and retinoic acid on transcriptional control in preosteoblasts. Mol Endocrinol 3:2079- 2085. Robey PG 1989 The biochemistry of bone. Endocrinol Metab Clin North Am 18:859-902. Gorman CM, Merlin0 GT, Willingham MC, Pastan I, Howard BH 1982 The Rous sarcoma virus long terminal repeat is a strong promoter when introduced into a variety of eukaryotic cells by DNA mediated transfection. Proc Natl Acad Sci USA 79:6777-6781. Bradley TR, Hodgson GS, Rosendaal M 1978 The effect of oxygen tension on haemopoietic and fibroblast cell prolifera- tion in vitro. J Cell Physiol 97517-522. Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acid guanidium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156-159. Maniatis T, Fritsch EF, Sambrook J 1982 Molecular Clon- ing: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, p. 545. Dolle P, Ruberte E, Kastner P, Petkovich M, Stoner CM, Gudas LJ, Chambon P 1990 Differential expression of genes encoding a, p and y retinoic acid receptors and CRABP in the developing limbs of the mouse. Nature 342:702-705. Ruberte E, Dolle P, Krust A, Zelent A, Morris-Kay G, Chambon P 1990 Specific spatial and temporal distribution of retinoic acid receptor gamma transcripts during mouse embryogenesis. Development 108:213-222. Zelent A, Krust A, Petkovich M, Kastner P, Chambon P 1989 Cloning of murine retinoic acid receptor a and cDNAs and of a novel third receptor y predominantly expressed in skin. Nature 339:714-717. Hu L, Gudas LJ 1990 Cyclic AMP analogs and retinoic acid influence the expression of retinoic acid receptor a, 0, and y mRNAs in F9 teratocarcinoma cells. Mol Cell Biol 10:391- 396. Thiede MA, Yoon K, Golub EE, Noda M, Rodan GA 1988 Structure and expression of rat osteosarcoma (ROS 17/2.8) alkaline phosphatase: Product of a single gene. Proc Natl Acad Sci USA 83319-323. Young MF, Kerr JM, Termine JD, Wewer UM, Wang MG, McBride OW, Fisher LW 1990 cDNA cloning, mRNA distribution and heterogeneity, chromosomal location, and RFLP analysis of human osteopontin/SSP-1 . Genomics 7:491-502. Genovese C, Rowe D, Kream B 1984 Construction of DNA sequences complementary to rat alpha I and alpha 2 collagen mRNA and their use in studying the regulation of type 1 collagen synthesis by 1,25-dihydroxyvitamin D. Biochemistry

Wozney JM, Rosen V, Celeste AJ, Mitsock LM, Whitters MJ, Kriz RW, Hewick RM, Wang EA 1988 Novel regulators of bone formation: Molecular clones and activities. Science

Leung DW, Spencer SA, Cachianes G, Hammonds RG, Collins C, Henzel WJ, Barnard R, Waters MJ, Wood WI 1987 Growth hormone receptor and serum binding protein: Purification, cloning and expression. Nature 330537-543. Smith WC, Linzer DIH, Talamantes F 1988 Detection of two growth hormone receptor mRNAs and primary translation products in the mouse. Proc Natl Acad Sci USA 899576- 9519. Young MF, Bolander ME, Day AA, Ramis CI, Robey PG, Yamada Y, Termine JD 1986 Osteonectin mRNA: Distribu- tion in normal and transformed cells. Nucleic Acids Res 14:

23:6210-6216.

242: 1528-1534.

4483-4497.

ZHOU ET AL. 776

32. Fisher LW, Termine JD, Young MF 1989 Deduced protein sequence of bone small proteoglycan 1 (biglycan) shows ho- mology with proteoglycan I1 (decorin) and several noncon- nective tissue proteins in a variety of species. J Biol Chem

33. Cherington V, Morgan B, Spiegelman BM, Roberts TM 1986 Recombinant retroviruses that transduce individual polyoma tumor antigens: Effects on growth and differentiation. Proc Natl Acad Sci USA 83:4307-4311.

34. Tooze J 1981 DNA tumor viruses. In: Tooze J (ed) DNA Tumor Viruses. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp. 61-124.

35. Topp WC, Lane D, Pollack R 1981 Transformation by SV40 and polyoma virus. In: Tooze J (ed) DNA Tumor Viruses. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp. 205-2%.

36. Woodworth CD, Isom HC 1987 Regulation of albumin gene expression in a series of rat hepatocyte cell lines immortalized by simian virus 40 and maintained in chemically defined medium. Mol Cell Biol 7:3740-3748.

37. Woodworth CD, Kreide JW, Mengel L, Miller T, Meng Y, Isom HC 1988 Tumorigenicity of simian virus 40-hepatocyte cell lines: Effect of in vitro and in vivo passage on expression of liver-specific genes and oncogenes. Mol Cell Biol 8:4492- 4501.

38. Neufield DS, Riple S , Henderson A, Ozer HD 1987 Immor- talization of human fibroblasts transformed by origin-defec- tive simian virus 40. Mol Cell Biol 7:2794-2802.

39. Thaller C, Eichele G 1987 Identification and spatial distribution of retinoids in the developing chick limb bud. Nature

40. Thaller C, Eichele G 1990 Isolation of 3.4-didehydroretinoic acid, a novel morphogenetic signal in the chick wing bud. Nature 345:815-819.

41. Wagner M, Thaller C, Jessell T, Eichele G 1990 Polarizing activity and retinoid synthesis in the floor plate of the neural tube. Nature 345:819-822.

42. Evans RM 1988 The steroid and thyroid hormone receptor superfamily. Science 240:889-895.

43. de The H, Marchi A, Tiollais P, Dejean A 1987 A novel steroid thyroid hormone receptor-related gene inappropriately expressed in human hepatocellular carcinoma. Nature 330:

44. Giguere V, Ong ES, Segui P, Evans RM 1987 Identification of a receptor for the morphogen retinoic acid. Nature 330:

45. Petkovich M, Brand MJ, Krust A, Chambon P 1987 A human retinoic acid receptor which belongs to the family of nuclear receptors. Nature 312:724-729.

46. Krust A, Kastner P, Petkovich M, Zelent A, Chambon P 1989 A third human retinoic acid receptor, hRAR y. Proc Natl Acad Sci USA 86:5310-5314.

47. Butler WT 1989 The nature and significance of osteopontin. Connect Tissue Res 23:123-126.

48. Oldberg A, Franzen A, Heinegard D 1986 Cloning and sequence analysis of rat bone sialoprotein (osteopontin) cDNA reveals an Arg-Gly-Asp cell-binding sequence. Proc Natl Acad Sci USA 83:8819-8823.

49. Prince CW, Butler WT 1987 1,25-Dihydroxyvitamin D, regu- lates the biosynthesis of osteopontin, a bone-derived cell attachment protein, in clonal osteoblast-like osteosarcoma cells. Collagen Relat Res 7:305-313.

50. Noda M, Rodan GA 1989 Transcriptional regulation of osteopontin production in rat osteoblast-like cells by para-

264:4571-4576.

327:625-628.

667-670.

624-629.

thyroid hormone. J Cell Biol 108:713-718. 51. Craig AM, Nemir M, Mukherjee BB, Chambers AF, Den-

hardt DT 1988 Identification of the major phosphoprotein secreted by many rodent cell lines as 2ar/osteopontin: En- hanced expression in H-rus-transformed 3T3 cells. Biochem Biophys Res Commun 157: 166- 173.

52. Craig AM, Smith JH, Denhardt DT 1989 Osteopontin, a transformation-associated cell adhesion phosphoprotein, is induced by 12-0-tetradecanoylphorbol 13-acetate in mouse epidermis. J Biol Chem 264:9682-9689.

53. Fisher LW, McBride W, Termine JD, Young MR 1990 Hu- man bone sialoprotein. Deduced protein sequence and chromosomal localization. J Biol Chem 265:2347-2351.

54. Oldberg A, Franzen A, Heinegard D, Pierschbacher M, Rouslahti E 1988 Identification of a bone sialoprotein receptor in osteosarcoma cells. J Biol Chem 263:19433-19435.

55. Oldberg A, Jirskog-Hed B, Axelsson S , Heinegard D 1989 Regulation of bone sialoprotein mRNA by steroid hormones. J Cell Biol 109:3183-3186.

56. Price PA, Williamson MK, Haba T, Dell RB, Jee WSS 1982 Excessive mineralization with growth plate closure in rats on chronic warfarin treatment. Proc Natl Acad Sci USA 79:

57. Otawara Y, Price PA 1986 Developmental appearance of matrix gla protein during calcification in the rat. J Biol Chem 261: 10828- 10832.

58. Fraser JD, Price PA 1990 Induction of matrix gla protein synthesis during prolonged 1,25-dihydroxyvitamin D, treatment of osteosarcoma cells. Calcif Tissue Int 46:270-279.

59. Price PA, Fraser JD, Metz-Virca G 1987 Molecular cloning of matrix gla protein: Implications for substrate recognition by the vitamin D-dependent y-carboxylase. Proc Natl Acad Sci USA 8423335-8339.

60. Reid IR, Chapman GE, Fraser TRC, Davies AD, Surus AS, Meyer J, Huq NL, Ibbertson HK 1986 Low serum osteocal- cin levels in glucocorticoid-treated asthmatics. J Clin Endo- crinol Metab 62:379-383.

61. Schule R, Umesono K, Mangelsdorf DJ, Bolado J, Pike JW, Evans RM 1990 Jun-Fos and receptors for vitamins A and D recognize a common response element in the human osteocal- cin gene. Cell 61:497-504.

62. Imai Y, Rodan SB, Rodan GA 1988 Effects of retinoic acid on alkaline phosphatase messenger ribonucleic acid, cate- cholamine receptors, and G proteins in ROS 17/2.8 cells. En- docrinology 122:456-463.

63. Oikarinen H, Oikarinen Al, Tan EML, Abergel RP, Meeker CA, Chu M-L, Prockop DJ, Uitto J 1985 Modulation of procollagen gene expression by retinoids. J Clin Invest 75:

64. Urist MR 1%5 Bone: Formation by autoinduction. Science

65. Reddi AH, Huggins C 1972 Biochemical sequences in the transformation of normal fibroblasts in adolescent rats. Proc Natl Acad Sci USA 69:1601-1605.

66. Wang EA, Rosen V, Cordes P, Hewick RM, Kriz MJ, Luxenberg DP, Sibley BS, Wozney JM 1988 Purification and characterization of other distinct bone-inducing factors. Proc Natl Acad Sci USA 85:9484-9488.

67. Wang EA, Rosen V, DAlessandro JS, Bauduy M, Cordes P, Harada T, Israel DI, Hewick RM, Kerns KM, LaPan P, Luxenberg DP, McQuaid D, Moutsatsos IK, Nove J, Wozney JM 1990 Recombinant human bone morphogenetic protein induces bone formation. Proc Natl Acad Sci USA 87: 2220-2224.

7734-7738.

1545-1553.

150:893-899.

EFFECTS OF RETINOIC ACID ON IMMORTALIZED PREOSTEOBLASTS

68. Barnard R, Ng KW, Martin TJ, Waters MJ 1991 Growth hormone receptors in clonal osteoblast-like cells mediate a mitogenic response to growth hormone. Endocrinology 128:

69. Canalis E, McCarthy TL, Centrella M 1989 The role of growth factors in skeletal remodeling. Endocrinol Metab Clin North Am 18:903-918.

1459- 1464.

Address reprint requests to: Dr. K.W. Ng

University of Melbourne Department of Medicine

St . Vincent’s Hospital Melbourne, Victoria, 306s Australia

777

Received in original form September 14, 1990; in revised form January 22, 1991; accepted February 6, 1991.

Retinoic acid modulation of mrna levels in malignant, nontransformed, and immortalized osteoblasts

Documents