JOURNAL OF BONE AND MINERAL RESEARCH Volume 5, Number 2, 1990 Mary Ann Liebert, Inc., Publishers

Monoclonal Antibodies to ROS 17/2.8 Cells Recognize Antigens, Some of Which Are Restricted to Osteoblasts and Chondrocytes JUDITH PERRY, MOIRA GILLIGAN, ELAINE GREEN, HILARY DOCHERTY, and DAVID HEATH

ABSTRACT We have raised a panel of 15 monoclonal antibodies (MAbs) recognizing cell surface antigens of the rat osteoblast-like cell line ROS 17/2.8. The MAbs were selected on the basis of preferential binding to ROS 17/2.8 cells compared to ROS 25/1 cells. Immunohistochemical studies of antigen localization on cryostat sections of rat calvaria, long bone, and soft tissues demonstrated that five of these MAbs, UBIM 1,2,3,12, and 17, recognize antigens that are restricted to normal rat osteoblasts and chondrocytes. The antigens appear to be localized to the cell surface of the osteoblast, with no apparent staining of bone matrix in either undecalcified or decalcified sections. In vitro, these MAbs recognize cell surface antigens present on two additional cell lines, ROS 24/1 and Rat 2 cells, and on the adherent cell population cultured from rat long bone marrow. Of these MAbs, three (UBIM 1, 2, and 3) recognize high-molecular-weight antigens of M,200,OOo225,000. This study has also identified cell surface antigens of ROS 17/2.8 cells that are not expressed by osteoblasts in vivo. MAbs UBIM 9 and 21 bind to marrow cells in long bone sections, to the 7-day-old nonadherent cell population from cultured marrow, and to lymphoid tissue in sections of spleen. Another four MAbs (UBIM 10, 11, 14, and 22) bind to a variety of cells and tissues both in vitro and in vivo. Studies of the interactions of this panel of MAbs with osteogenic tissues and cell lines may have an important impact on the understanding of osteoblast physiology.

INTRODUCTION HE OSTEOBLAST (OB) is a key cell in bone formation, synthesizing and secreting the organic matrix of bone, and is involved in the regulation of matrix mineralization. The study of OB function and differentiation has been hampered by the inability to purify pure OB populations. The availability of specific markers of the cell surface of the OB would be of great value in this task. Properties associated with the OB plasma membrane include alkaline phosphatase (AP) activity and the presence of parathyroid hormone (PTH) receptors,") although neither of these is unique to the OB. Monoclonal antibodies (MAbs) have been used successfully to characterize the cell surface of


other cell types, such as o s t e o ~ l a s t s (and ~ ~ ~lymphoid ) and myeloid cells, and to study the lineage of cells of the hematopoietic system.(4.5)The object of our study was to raise MAbs to antigens restricted to the plasma membrane of the normal rat OB and to use these MAbs to further characterize the cell surface phenotype of the OB. Other workers have attempted to raise OB-specific MAbs using cells isolated from calvarial digests as immunogen. This technique has produced a MAb specific to the cell surface of the chick osteOcyte,'6) to the murine PTH receptor,") and to OB-restricted antigens.'*) In this study we have raised MAbs directed against the rat clonal OB-like cell line ROS 17/28. These cells exhibit elevated AP activity modulated by PTH and dexamethasone, receptors for


Department of Medicine, University of Birmingham, Queen Elizabeth Hospital, Birmingham, B15 2TH, England.



188 PTH and 1,25-&hydroxyvitamin D3 11,25-(OH)2D,), synthesize and secrete matrix proteins, such as collagen type 1 and osteocalcin, and are osteogenic when implanted in v~vo.(~-") The antibody selection procedure employed utilized a differential screening assay involving positive recognition of ROS 17/2.8 cells but not ROS 25/1 cells. This latter cell line is derived from the same rat osteosarcoma but is fibroblastic in morphology and expresses none of the OB-like characteristics of ROS 17/2.8 cells. In this way it was hoped to select antibodies to antigens associated specifically with the OB phenotype and eliminate those of housekeeping or transformation antigens. We have generated a panel of 15 MAbs to ROS 17/2.8 cells, 5 of which recognize antigens restricted to normal rat OBs and chondrocytes in vivo.

MATERIALS AND METHODS Materials Cell culture media and trypsin were from Flow Laboratories (Hertfordshire, UK). Sera and methionine-free medium (Selectamine Kit) were purchased from GIBCO (Paisley, Scotland). A mouse monoclonal antibody to rat major histocompatibility complex (MHC) class 1 antigens, MRC 0 x 1 8 , was purchased from Serotec (Oxfordshire, UK). Polyclonal rabbit antimouse immunoglobulins (from Dako, Buckinghamshire, UK) were iodinated by the chloramine-T method. I l l ) Peroxidase-conjugated rabbit antimouse immunoglobulins were also obtained from Dako. Radioisotopes were obtained from Amersham International plc (Buckinghamshire, UK) and protein A-Sepharose CL-4B from Pharmacia (Milton Keynes, UK). A mouse monoclonal typing kit was obtained from Serotec. All other chemicals were purchased from Sigma Chemical Company (Poole, Dorset, UK) and were of analytic grade.

Cell culture ROS 17/2.8, ROS 25/1 and ROS 24/1 cells(11)(all kindly provided by Dr. G. Rodan, Merck, Sharp and Dohme Laboratories, West Point, PA) were maintained in monolayer culture in Ham's F-12 medium containing 10% (v/v) fetal bovine serum (FBS) in an atmosphere of 5% COl in air. UMR-106 cells(13)were grown in Ham's F-12 medium containing 5% (v/v) FBS. Rat 2 cells(14)and cell cultures derived from baby rat retina, rat embryonic muscle, rat embryonic brain, and baby rat kidney were kindly provided by Prof. P. Gallimore, Department of Cancer Studies, University of Birmingham, UK. These cell lines and rat hepatoma cells (2sFou)(") were grown in Dulbecco's modified Eagle's medium containing 10% (v/v) FBS. Rat pituitary tumor cells (GH3)(I6)were grown in Ham's F-12 medium containing 10% (v/v) horse serum and 5% (v/v) FBS. Rat insulinoma cells (RIN)117)were grown in RPMI-1640 medium containing 10% (v/v) FBS. The rat thyroid cell line FRTL5'") (kindly provided by Dr.

L. Kohn, NIH, Bethesda, MD, and Miss E.G. Black, Dept. Medicine, University of Birmingham, UK) was grown in Coon's modified Ham's F-12 medium plus 5% (v/v) newborn bovine serum. Medium was replaced on all cultures every 3-4 days. For binding assays cells were seeded in 24-well plates and grown to confluence prior to fixation. The mouse plasmacytoma cell line (NSO)(") and hybridomas were maintained in RPMI-1640 medium plus 10% (v/v) FBS.

Production of monoclonal antibodies ROS 17/2.8 cells were seeded at a density of 7 x lo' per cm2 and cultured for 14 days. Four different protocols were employed in the preparation of cells for immunization. (1) The cell monolayer was washed with Dulbecco's phosphate-buffered saline (DPBS) containing 0.1 Yo (w/v) CaCl, and 0.1% (w/v) MgCL and then harvested into DPBS on ice by scraping with a rubber policeman. The cells were pelleted at 100 x g for 5 minutes, resuspended in DPBS, and a single cell suspension obtained by passing through a 19G needle. (2) Confluent ROS 17/2.8 cells were treated with 30 nM dexamethasone in Ham's F-12 medium plus 10% (v/v) FBS for 5 days with medium replacement daily. The cells were harvested by limited trypsinization in 0.125% (w/v) trypsin per 1 mM Na2EDTA in DPBS at room temperature until the cells detached. The reaction was terminated by the addition of FBS and the cells washed three times with DPBS. (3) and (4) ROS 17/2.8 cells, harvested by trypsinization as described, were coated prior to injection with a mouse serum obtained from animals immunized with ROS 25/1 cells. Similarly, dexamethasone-treated ROS 17/2.8 cells were precoated with a mouse polyclonal antiserum raised against untreated ROS 17/2.8 cells. Cells (loe) were incubated with 300 pl undiluted serum at 4°C for 1 h and then washed in DPBS.l2") Cell viability was assessed by trypan blue exclusion, and all cell preparations were resuspended at a final concentration of 5 x 10' cells per ml for immunization. Balb/c mice were injected intraperitoneally with lo7 cells at 14 day intervals. Sera were tested for an immune response 10 days later, and animals showing a good serum titer were selected for tail vein injection with lo' cells 3 days prior to fusion. Spleen cells from immune mice were fused with NSO cells using 50% (w/v) polyethylene glycol 6OOO (Koch-Light Ltd. Suffolk, UK).(ll)

Screening and selection of secreting hybridomas Hybridomas were selected for preferential binding to ROS 17/2.8 cells compared to ROS 25/1 cells. Confluent cell monolayers in 24-well plates were washed in DPBS containing 0.1% CaCl, (w/v) and 0.1% MgC12 (w/v) and fixed using 0.25% (w/v) glutaraldehyde in DPBS for 5 minutes at room temperature. Nonspecific antibody binding was blocked with 10% (w/v) bovine serum albumin (BSA), 0.1% (w/v) Na azide, and 100 mM glycine in DPBS, and plates were stored at -20°C without loss of


MONOCLONAL ANTIBODIES TO ROS 17/2.8 CELLS immunoreactivity.[22)At assay, the plates were thawed at room temperature, washed three times in DPBS and 0.5% (w/v) BSA, and incubated with hybridoma supernatant (200 pl per well) for 1 h at 4°C. After washing, [1251]rabbit antimouse immunoglobulin was added (200,000 cpm per well) and incubated at 4°C for 1 h. The cells were washed and harvested into DPBS and radioactivity determined using a Nuclear Enterprises NE 1600 gamma counter. MRC 0 x 1 8 , a mouse MAb recognizing rat MHC class 1 antigen^,^'^) was used as the positive control. Nonspecific binding (NSB) was determined by incubation with cell culture medium alone during the first incubation period. This procedure was also used to assess the binding of MAbs to cultured cell lines. Significant binding to the cell monolayer was determined by applying a two-tailed t-test to the difference between mean cpm NSB and mean cpm of wells incubated with hybridoma supernatant. Selected hybridomas were cloned twice by limiting dilution.

Competition binding studies MAbs were radioactively labeled by incubating hybridomas in methionine-free medium with the addition of L[35S]methionine (specific activity > 1000 Ci/mmol: 100 pCi per lo7cells) for 3 h at 37°C. The supernatant was collected by centrifugation, and unlabeled L-methionine added at a final concentration of 2 mM. ROS 17/2.8 cells were incubated with [35S]MAbplus or minus saturating amounts of nonradioactive hybridoma supernatant for 1 h at 4°C. After washing with DPBS (3 x 1 ml), the cells were harvested and radioactivity determined by scintillation counting.

rotated at 4°C for 3 h. The mixture was then loaded onto a discontinuous sucrose gradient of 0.5 ml 15% (w/v) sucrose and 0.5 ml 7.5% (w/v) sucrose in 10 mM Tris-HCI, pH 8, 50 mM NaCI, 0.5% (v/v) NP40, 0.5% (w/v) sodium deoxycholate, and 0.05% (w/v) sodium dodecyl sulfate (SDS) and centrifuged at 7000 x g for 5 minutes. The beads were treated with three consecutive washes (1 ml) consisting of (1) 10 mM Tris-HC1, pH 8, 500 mM NaCI, 0.5% (v/v) NP40, and 0.05% (w/v) SDS; (2) 10 mM TrisHCI, pH 8, 150 mM NaCI, 0.5% (v/v) NP40, 0.5% (w/v) Na deoxycholate, and 0.05% (w/v) sodium dodecyl sulfate (SDS); (3) 10 mM Tris-HCI, pH 8, and 0.05% (w/v) SDS. ( 2 5 ) Immunoprecipitates were released from the beads by boiling in SDS-polyacrylamide gel electrophoresis (SDSPAGE) sample buffer and analyzed by SDS-PAGE, followed by fluor~graphy."~) MRC OX18 was used in immunoprecipitations as a positive control (two polypeptide bands of M, 50,000 and 12.000 on SDS-PAGE).

Culture of rat long bone marrow Marrow was mechanically removed from long bones of 6-week-old Wistar rats(2*)and inoculated at 8 x 104 cells per cm2 in a-MEM containing 10% (v/v) FBS. The cells were grown for 7 days, after which the nonadherent cells were removed and collected by centrifugation at 100 x g for 5 minutes. The adherent cell population was maintained through three passages. The nonadherent cells were washed three times in DPBS and inoculated into poly-L-lysine coated %well plates at 1.5 x lo5 cells per well.(291 After 1 h at room temperature, excess cells were tipped from the wells and the plates fued, blocked, and stored as described. Binding of MAbs was assessed using radiolabeled second antibody, as described.

Immunoprecipitation of antigens recognized by MAbs ROS 17/2.8 cells (3 x lo5)were metabolically labeled by incubation for 4 h in 2 ml methionine-free medium containing 125 pCi L-["5S]methionine (specific activity > 1000 Ci/mmol). The cells were washed three times with DPBS containing 2 mM unlabeled L-methionine and solubilized at 4°C for 1 h in 1 mI 10 mM Tris-HC1, pH 8, 0.5% (v/v) NP40, 150 mM Na,EDTA, 5 mM iodoacetamide, 2 mM L-methionine, 0.5% (w/v) BSA, 150 mM NaCI, and 1 mM phenylmethylsulfonyl fluoride. Nuclei were removed by centrifugation at 7000 x g for 5 minutes. To preclear 0.1 ml MAb to Fucus sperm(a4)was added to the cell lysate and incubated at 4°C for 1 h with mixing. A 10% (v/v) suspension (0.1 ml) of protein A-Sepharose CL-4B previously saturated with rabbit antimouse immunoglobulins (protein A-RAM) was added, mixed at 4°C for 30 minutes, and then the beads removed by centrifugation at 7000 x g. A further 0.1 ml protein A-RAM was added to the cell lysate for another 30 minutes. The beads were again removed. Hybridoma supernatants (0.05 ml) were added to 0.25 ml aliquots of the cell lysate and rotated for 1 h at 4°C. Protein A-RAM (0.1 ml) was added to each tube and

Immunohistochemistry of cryostat sections of rat calvaria, long bone, and soft tissues Rat calvaria (2 days old) and 2- and 6-day-old rat long bone were embedded in cryo-M-bed freezing gel (Bright Instrument Co., Cumbria, UK), snap frozen in liquid nitrogen, and stored at -70°C. A number of rat soft tissues were obtained from 6-week-old Wistar rats and treated in the same way. Cryostat sections (5 pm) were cut and air dried for 1 h. The slides were foil wrapped and stored at -20°C prior to use. Calvaria and long bone were decalcified by incubation for 10 minutes in 0.3 M Na2EDTA in DPBS. Sections were incubated with hybridoma supernatants in a humidified chamber at room temperature for l h and then washed for 15 minutes in DPBS. Control sections were incubated with RPMI plus 10% (v/v) FBS. The second step consisted of incubation for 1 h with rabbit antimouse immunoglobulin conjugated to horseradish peroxidase [1:100 in DPBS containing 10% (v/v) normal rabbit serum] followed by incubation with diaminobenzidine [0.5 mg/ml in DPBS plus 0.3% (w/v) hydrogen peroxide] for





ROS 17/2.8

ROS 2411

ROS 25/1


Pituitary GH3

Thyroid FRTLS

Baby rat kidney (BRK) ~



1 2 3 12 17

11.9b 13.1b 12.5b 3.7b 12.8b

2.8b 2.0b 2.4b 2.9b 5.6b

1.o 0.9 1.1 1.3 1.1

0.7 0.9 0.7 0.8 1.3

0.6 0.5 0.6 1.3 0.8

0.7 0.8 0.6 1.1 1.1

1.1 1.3 0.9 0.9 0.8



8 13 15 19

12.8b 13.7b 7.6b 7.6b

2. l b 2.0b 1.3 1.1

1.2 1.3 1.1 1.5

0.9 1.4 1.2 1.8c

1.6 0.8 1.1 3 9

0.8 1.1 1.1 1.5

1.7 0.9 0.9 1.o



14.6b 2. lb

1.3 1.1

1.3 1.2

1.o 1.8c

0.7 1.1

0.8 1.3

1.6 1.1



22.7b 25.9b 12.8b

1.o 1.3 1.5 1.5

1.5 1.3 0.9 1.4

0.9 0.9 1.8c 1.9c

14.2b 1O.Ob 8.5b 9.3b

0.9 0.7 1.2 1.o

1.6 1.3 0.9 1.6



Hepatoma 2sFOU

Rat insulinoma (RW)

Baby rat retina (BRR)

Rat embryonic muscle (REM) ~

Rat embryonic brain (REB)

Fibroblasts RAT2




1 2 3 12 17

1.1 0.9 1.2 0.9 1.o

0.5 0.4 0.6 0.7 1.o

0.8 0.6 0.5 0.7 0.9

0.6 0.8 1.o 1.2 0.8

0.4 0.6 0.5 1.o 0.8

3.6b 2.3b 2.2b 1.7d 5.4b



8 13 15 19

1.6 1.o 0.9 1.4

1.1 1.o 1.o 3.3b

1.2d 1.o 1.1 1.1

1.2 1.o 0.9 1.3

1.5 0.9 1.o 1.1

1.3 2.0b 1.1 1.4



1.3 0.9

1.2 1.6d

1.o 0.6

1.o 1.3d

1.2 1.o

1.6 0.9


1.7 1.3 1.3 1.7

2 9 3.2b 4.8b 5.6b

1.2 1.1 1.2 1.2

4.2b 4.8b 12.8b 15.1b

2.4c 2.2c 2.5d 3.2b

36.7b 39.7b 17.0b 19.0b


10 11 14 22

aCells were fixed with 0.25% glutaraldehyde in PBS and incubated with MAbs for 1 h at 4°C followed by incubation for a further 1 h with ['Y]rabbit antimouse immunoglobulins. Results are expressed as a binding ratio = mean cpm with hybridoma supernatant divided by mean cpm control wells. Statistical analysis of the difference between cpm control wells and cpm with hybridoma supernatant by the two-tailed Students r-test is indicated. bP < 0.001. CP < 0.01. d P < 0.05.




Nonadherent marrow



Primary adherent marrow

Passage I

Passage 2

Passage 3



0.7-1.2 0.9-1.2 0.7-1.1 0.8 0.9

1.4d-2.4c 1.4-2.2C 1.2-1.8d 1.0-2.3c 2.4“-4.4”

5.2”-6.5” 4.3b-5.4b 4.84.5” 1.3-3.2” 2.3d-4.6”

1.6-5.8” 1.5-4. l b 1.3-4.2” 1.7-3.0” 6.8”-8.OC

2.ld-3.2c 1.3-2.lC 1.3-2.1C 1.8”-2.Y 3.4”-5.5”



1.3-1.7” 0.9 1.o 0.8

1.4-1.9 1.6”-3.4 1.3 1.0-1.3

3.4-5.9b 1.0-29 1.2-1.6“ 0.8-1.1

1.2-3.8” 1.4-5.Ob 1.4-2.4C 0.8-1.4

1.5-2.2d 1.9”-2.Y 1.5C-l.6C 1.0-1.3



1.5-2.3” 1.6“-1.7”

0.9-1.4 0.9-1.5

1.5-1.9“ 0.8-1.2

0.9-1.4“ 0.8-1.5

1.0-1.3d 1.2-1.3



1.4-1.5“ 1.2-1.3 1.2-3.4C 1.1-1.5

9.8b-23.9” 9.5”- 13.9” 8.7”-10.6” 9.9b- 16.8”

10 11 14 22

19.9b-22.3” 19.6b-22.5” 5.3”- 10.6” 5.7b

22.3” 22.5” 8.2”- 18.5b 8.3b

23.7” 24.6b 8.4”-2 1.2” 8.9

aMarrow from rat long bones was cultured in vitro for 7 days, followed by separation of nonadherent cells. The adherent population was maintained through three passages. Binding of MAbs to glutaraldehyde-fixed cells was assessed by binding of [”’I]rabbit antimouse immunoglobulins. Results are expressed as a binding ratio = mean cpm with hybridoma supernatant divided by mean cpm of control wells. Statistical analysis of the difference between cpm control wells and cpm with hybridoma supernatant by the two-tailed Student’s 1-test is indicated. bP

< 0.001.

CP < 0.01. dP c 0.05.

10 minutes at room temperature. Sections were washed in tap water and counterstained with Mayer’s hematoxylin.


Monoclonal antibodies A total of 15 MAbs of interest were selected from the products of five fusions using four different immunization techniques. MAbs were selected by their ability to bind to ROS 17/2.8 cells but not to ROS 25/1 cells. MAbs UBIM 1,2, and 3 were raised against ROS 17/2.8 cells harvested from the monolayer by scraping. Dexamethasone-treated ROS 17/2.8 cells were used as immunogen for the production of MAbs UBIM 8,9, 12, 13, 14, 15, 17, and 19. MAbs UBIM 10 and 11 were raised against ROS 17/2.8 cells precoated with ROS 25/1 antiserum and MAbs UBIM 21 and 22 raised against dexamethasone-treated ROS 17/2.8 cells precoated with untreated ROS 17/2.8 antiserum. The limited trypsinization procedure employed was shown to result in retention of cell surface antigens. Approximately 50% of A P activity and PTH receptor-linked adenylate cy-

clase activity was retained (data not presented). Trypsinized cells also retained the ability to bind MRC 0 x 1 8 , which recognizes plasma membrane MHC class 1 antigens. All MAbs selected were of subclass IgGI, except UBIM 9 (IgG2b) and UBIM 10 and 11 ( I g G d . On the basis of the in vitro binding assays, immunohistochemistry, and competition binding analysis, MAbs were divided into four groups: UBIM 1, 2, 3, 12, and 17 constituted group A; UBIM 8, 13, 15, and 19, group B; UBIM 9 and 21. group C; and UBIM 10, 11, 14, and 22, group D.

Binding of MAbs to cultured cell lines Binding of MAbs to the following rat cell lines was assessed: ROS 24/1, which expresses a phenotype intermediate between ROS 17/2.8 and ROS 25/1“’); UMR-106, a clonal rat osteosarcoma cell line; baby rat retina (BRR); rat embryonic muscle (REM); rat embryonic brain (REB); fibroblastic Rat 2 cells; baby rat kidney (BRK); 2sFou, rat hepatoma; GH3, rat pituitary tumor; RIN, rat insulinoma; and FRTL5, rat thyroid cells. All the MAbs recognized antigens that are restricted to certain cell types (Table 1).








194 In addition to ROS 17/2.8 cells, MAbs in group A (UBIM 1 , 2, 3, 12, and 17) all bound to ROS 24/1 and Rat 2 cells. Group B MAbs bound variously to a small number of the cell lines tested. UBIM 8 recognized ROS 24/1 cells and also showed a low but significant level of binding to baby rat retina. UBIM 13 bound to ROS 24/1 and Rat 2 cells. UBIM 19 recognized, in addition to ROS 17/2.8 cells, the rat insulinoma cell line RIN and the pituitary cell line GH3 and also exhibited a low level of binding to UMR-106 cells. The antigen recognized by UBIM 15 appeared to be restricted to ROS 17/2.8 cells. Group C MAbs (UBIM 9 and 21) bound principally to ROS 1712.8 cells, although UBIM 21 showed low levels of binding to three other cell lines: UMR-106, rat insulinoma, and rat embryonic muscle. MAbs in group D (UBIM 10, 1 1 , 14, and 22) bound to several of the cell lines tested.

Binding of MAbs to marrow cells cultured in vitro Binding of the MAbs to marrow cells was assessed at stages of culture (Table 2). The results are shown as a range of binding ratios (mean cpm with hybridoma supernatant divided by mean cpm of control wells) for each antibody tested in a number of experiments. The level of binding of the MAbs varied considerably between different experiments. This may reflect the heterogeneous nature of these cultures and possibly the selective proliferation of particular cell types from the marrow of different animals. UBIM 1, 2, 3, 12, and 17 (group A) bound to the adherent cell population in both primary and passaged cultures. Group C MAbs (UBIM 9 and 21) bound principally to the nonadherent population, although UBIM 9 also exhibited a low but significant level of binding to passaged adherent cells. MAbs in group B showed varied patterns of cell recognition. UBIM 13 bound to adherent marrow cells, and UBIM 15 also bound to this cell population but at a low level and only after passaging. UBIM 8 recognized both adherent and nonadherent cells. UBIM 19 did not show any significant binding to marrow cells in vitro. UBIM 10, 1 1 , 14, and 22 (group D) recognized both nonadherent and adherent marrow cells at most stages of culture.

Irnrnunohistochernistry of rat tissues The tissue distribution of antigens recognized by MAbs was investigated by immunohistochemical staining of neonatal rat calvaria and long bone and of sections of liver, spleen, kidney, brain, stomach, muscle, and heart of young rats. Group A MAbs (UBIM 1 , 2, 3, 12, and 17) stained OBs in calvarial sections (Fig. lA, B, C, and D) and both OBs and chondrocytes in rat long bone (Fig. lE, F, G, and H). The staining appeared to be localized to the OB cell surface, and the pattern of staining was unchanged in decalcified sections. No staining of the bone matrix was observed. These five MAbs showed no staining of any of the other tissues examined. UBIM 9 and 21 (group C) stained cells of the marrow in long bone (Fig. 11) plus lymphoid cells present in spleen (Fig. 1 J and K). UBIM 10, 11, 14, and 22 (group D) were not restricted to osteogenic cells but stained a number of cell types in calcified tissues, brain, kidney, and spleen and endothelial cells in sections of stomach (Fig. 1L). Group B MAbs UBIM 8, 13, 15, and 19 showed no staining of any of the soft or calcified tissues examined.

competition binding studies Competition binding studies were employed to determine whether any of the MAbs recognized the same or overlapping epitopes (Table 3). Four of the five MAbs in group A, UBIM 1, 2, 3, and 17, appeared to recognize exclusive epitopes. However, some competition appeared to occur between the group A MAb, UBIM 12, and UBIM 15 (group B), although these MAbs showed different patterns of cell recognition both in vivo and in vitro. Competition was observed between the two MAbs in group C, UBIM 9 and UBIM 21, with each reducing the binding of the other by approximately 40%. All the MAbs in group D (UBIM 10, 11, 14, and 22) competed with each other, suggesting recognition of the same or overlapping epitopes. Neither UBIM 8 nor 13 (group B) showed any competition with any other MAbs and thus appeared to recognize distinct epitopes.

FIG. 1. Immunoperoxidase localization of antigens recognized by MAbs in cryostat sections from rat calcified and soft tissues counterstained with Mayer’s hematoxylin. Magnification x 1022. Neonatal rat calvaria stained with MAbs UBIM 1 (A), UBIM 3 (B), UBIM 12 (C), and UBIM 17 (D) showing positive staining of osteoblasts (ob) lining bone surfaces and bone lacunae (L) and in areas of developing bone (db). Bone matrix (m) was not stained. Neonatal rat long bone stained with MAbs UBIM 2 (E), UBIM 3 (F), UBIM 12 (G), and UBIM 17 (H) demonstrating positive staining of osteoblasts (ob) and chondrocytes (c). Cell nuclei stain darkly with hematoxylin, particularly in the periosteum (P). Marrow of neonatal rat long bone stained with UBIM 9 (I) showing positive staining of numerous cells. Spleen of 6-week-old rat, control section (J) and incubated with UBIM 21 (K), showing positive staining of lymphoid cells (lc). Neonatal rat calvaria stained with UBIM 22 (L) showing positive staining of all cell types. Control sections of calvaria (M), long bone (N),and marrow (0)were incubated with RPMI plus 10% FBS.



1 . _2. 3




-Pala FIG. 2. Immunoprecipitation of the antigens recognized by UBIM 1, 2, 3, 14, and 22. (A) Fluorograph of 15% SDSPAGE of immunoprecipitates from L-[35S]methionine-labeledROS 1712.8 cells. MAbs used were as follows: track 1, MRC OX 18; track 2, UBIM 1; track 3, UBIM 2; track 4, UBIM 3. (B) Fluorograph of 12.5% SDS-PAGE of immunoprecipitates from L-[35S]methionine-labeledROS 17/23 cells. MAbs used were as follows: track 1, UBIM 14;track 2, UBIM 22;track 3, MAb to Fucus sperm. MAbs UBIM 10 and 11 also immunoprecipitate a single band of M, 29,000that comigrates with that precipitated by UBIM 14 and 22. Specific bands immunoprecipitated by the MAbs are indicated by arrows.

Molecular size of antigens recognized by MAbs

Figure 2B shows the fluorograph of a 12.5% PAGE of the immunoprecipitation products of UBIM 14 and 22. Track 3 shows the polypeptide bands precipitated using an irrelevant MAb (raised against Fucus sperm) with which the lysate had already been precleared. Several nonspecifically precipitated protein species are seen in this track and in tracks 1 and 2. Both UBIM 14 (track 1) and 22 (track 2) specifically precipitated a protein of M, 29,000. UBIM 10 and 1 1 immunoprecipitated a band of similar M, that comigrated with that of UBIM 14 and 22 (data not shown).

Immunoprecipitation from detergent lysates of ?Mabeled ROS 17/2.8 cells followed by SDS-PAGE under reducing conditions was used to determine the molecular size of the antigens recognized by the MAbs. Figure 2A shows the fluorograph of a l5Olo PAGE of the immunoprecipitation products using MRC OX18 and UBIM 1, 2, and 3. The MHC class 1 antigens recognized by MRC OX18 are clearly seen as two bands of M, 50,000 and 12,000(track 1). The remaining bands in this track are nonspecifically coprecipitated using this technique. These same nonspecific bands are seen using UBIM 1, 2, and 3. However, UBIM 1 can be seen to immunoprecipitate a specific band at M, 225,OOO (track 2). UBIM 2 and 3 both immunopreDISCUSSION cipitated two bands of M, 225,000 and 200,000 (tracks 3 and 4). The relative intensity of these two bands differs, We have raised a panel of 15 MAbs to cell surface antihowever, between UBIM 2 and 3. UBIM 2 exhibits a dense gens of the OB-like cell line ROS 17/2.8.The properties of band at M, 225,000and a fainter one at 200,000,whereas the MAbs are summarized in Table 4. Of these MAbs, 5 the converse is true for UBIM 3. (UBIM 1, 2, 3, 12, and 17, group A) recognize antigens

104 f 4 117 f 7 ND ND 100 f 6 ND

8 13 15 19





89 f 2 ND ND ND

46 f 2 b 98 f 18 96 f 14 ND ND










loof 1 1



108 f 8 95 f 4


104 f 8 44 f 3 b 92 f 3 ND




143 f 110 f 101 f 104 f



12 14 7 12

5 7

91 115


loo f



10 2 11

100 f 13 95 f 7 63 f 13b ND

98 f 113 f 102 f 56 f 106 f


131 f 31 101 f 5 ND ND

105 f 3 99f7 48 f 6b ND ND


Group A

105 f 14 95 f 10 90f6 88 f 13

83 f 12 120 f 15

95 f 9 92 f 5 98 f 6 %f9

120 f 34 99f6 115 f 16 96 f 6 64 f 3 b


ND 93 f 23 ND ND

93f8 ND

68 f lb 107 f 2 ND ND

99f1 95 f 9 116 f 8 ND ND


5 6b 8 2

94f5 91 f 5 115 f 5 89 f 4

94f16 105 f 22

102 f 55 f 116 f 102 f

91f3 110 f 13 134 f 10 98 f 10 102 f 5

9 18 5 2b 2

126 f 115 f 105 f 93 f

13 14 1 3

103 f 9 107 f 9

90f9 105 f 1 53 f 4 b 90f4

100 f 103 f 97 f 71 f 108 f


Group B



15 5 13 7 3

138 f 94 f 99 f 95 f

4 13 17 6

113 f 4 104 f 17

103 f 7 91 f 10 106 f 10 ND

117 f 121 f 105 f 91 f 105 f





D 92 f 85 f 98 f 98 f

4 5 8 8

46 f 2b 60 f 6b


104 f 5 98 f 2 97 f 9






143 f 108 f 109 f 92 f

63 48 7 2 2 5

5b 17b

100 f 5 105 3 100 f 6 %f4

107 f 9 102 f 16 118 f 13 9Of5 125 f 9




7 8 9 6



54 f 56 f 61 f 64 f lob




93 f 8 110 f 18






105 f 91 f 89 f 97 f







11 10 16 8

56 49 55 49

2b 2b f Sb



f lb

93 f 10 94f3





113 f 100 f 129 f 108 f


Group D

6 7 5 12 9



98 f 91 f 41 f 41 f

83 115



15 5 26b

10 8

83 f 10 106 f 5 9Of5

105 f 87 f 95 f 85 f 120


6 6 15 3 1

68 f lob 51 f 2b 47 f 5 b 44 f l l b

111 f 4 119 f 13

111 f 4 106 f 2 112 f 8 100 f 5

128 f 87 f 122 f 98 f 122 f


*ROS 17/2.8 cells were incubated with [%]MAb in the presence or absence of saturating amounts of nonradioactive MAbs for 1 h at 4°C. Results are expressed as a percentage of binding in the absence of unlabeled MAb. Wompetition. CNot determined.

8 13 15 19

6 15 20 15





95 105 106 109




Group C





Group A

SuDdw-f OF THE

Cultured cellsa



Tissue distribution

ROS 17/23, ROS 24/1, Rat 2, adht marrowb


Osteoblasts in 2-dayold calvaria Osteoblasts and chondrocytes in 2- and 6-day-old long bone



Mr 225,000 200,000, 225,000 200,000, 225,000 -

ROS 17/2.8, (BRR)d ROS 17/2.8, ROS 17/2.8, ROS 17/2.8,

24/1, nonadhtc and adht marrow, 24/1, Rat 2, adht marrow (adht marrow) GH3, RIN, (UMR-106)

N o staining



ROS 17/23, nonadht marrow, (adht marrow) ROS 17/23, nonadht marrow, (UMR-106, RIN, REM)

Marrow in 2-and 6day-old long bone, lymphoid tissue of spleen



ROS 17/2.8, Rat 2, GH3, RIN, REM, REB, adht marrow, (nonadht marrow)

Variety of cell types in calcified tissues, brain, kidney, stomach endothelium


ROS 17/28, UMR 106, Rat 2, GH3, RIN, REM, REB, adht marrow, nonadht marrow



C e l l lines derived from rat fibroblasts (Rat 2), baby rat retina (BRR), rat pituitary (GH3), rat insulinoma (RIN), rat embryonic muscle (REM), and rat embryonic brain (REB). bAdherent cell population derived from bone marrow, both primary and passaged cultures. CNonadherent cells (7 days old) from bone marrow. dParentheses indicate low levels of binding. cWithin group D, UBIM 10 and 14 but not UBIM 11 and 22 showed some binding to nonadherent marrow.

that are restricted principally to OBs and chondrocytes. These MAbs stain specifically these cell types in both undecalcified and decalcified sections of rat bone, with no staining of any soft tissues examined. The staining pattern exhibited by these MAbs suggests that these antigens are intimately associated with the OB plasma membrane, with no staining of the bone matrix even after decalcification. Molecular size analysis indicated that the antigens recognized by 3 of the MAbs (UBIM 1, 2, and 3) are of high molecular weight: UBIM 1 immunoprecipitates a protein of M, 225,000, and UBIM 2 and 3 immunoprecipitate two polypeptides of M, 225,000 and 200,000. Competition for binding to ROS 17/2.8 cells was not observed between these MAbs, however, indicating that they probably recognize different antigenic epitopes, although these may reside on the same or related proteins. In addition to binding to ROS 17/2.8 cells, these 5 MAbs recognize antigens expressed on ROS 2411 cells and on the fibroblast-like cell line, Rat 2. All 5 MAbs also recognize antigens present on cells of the stromal population derived from bone marrow.

This stromal cell system contains the precursor cells of a number of cell types, including fibroblasts, reticular cells, adipocytes, chondrocytes, and osteobIasts.(30’This population of cells has been shown to retain the capacity to differentiate to form bone when reimplanted in viva.(") Whether these MAbs recognize antigens restricted to a subpopulation of cells with osteogenic capability remains to be determined. The identity of the antigens recognized by group A MAbs is still unknown. Immunohistochemical data appear to suggest a plasma membrane location of these antigens. However, the highly restricted cell and tissue recognition pattern (particularly the lack of binding to UMR-106 cells) makes it unlikely that any of these antibodies is directed against either alkaline phosphatase or the PTH receptor, both of which are expressed at high levels by UMR-106 osteosarcoma cells. The data presented, however, cannot exclude the possibility that these MAbs are directed against extracellular proteins. OBs are known to synthesize and secrete a number of specific matrix proteins. These include



osteocalcin, osteopontin, osteonectin, bone proteoglycans I and 11, and bone sialoproteins I and II.‘”) The antigens recognized by MAbs UBIM 1, 2, and 3 were of high molecular weight (M, approximately 200,000), similar to that recorded for human bone proteoglycan II.(33’ However, intact bone proteoglycan I1 migrates as a heterogeneous diffuse band on SDS gels, unlike the immunoprecipitation products of UBIM 1, 2, and 3. Definitive identification of the group A antigens awaits purification and further analysis. This study has indicated that it is possible to use ROS 17/2.8 cells as immunogen to generate MAbs to OB- and chondrocyte-restricted antigens. We have also identified antigens that are present on ROS 17/23 cells but are not shared by osteogenic cells in vivo; these are demonstrated by the group C MAbs, UBIM 9 and 21, which bind to cell surface antigens of ROS 17/23 cells but in vivo do not recognize OBs but bind to marrow cells and lymphoid cells in spleen. Four other MAbs (UBIM 8, 13, 15, and 19, group B) fail to stain any cells in cryostat sections of any of the soft or calcified tissues examined. These four MAbs may be recognizing antigens expressed specifically on cells cultured in vitro. The remaining MAbs, UBIM 10, 11, 14, and 22 (group D), bind to a variety of cultured rat cell lines and to a number of calcified and soft rat tissues. Although selected from three different fusions, these MAbs are all directed against the same protein of M, 29,000. Of the 15 MAbs generated, 4 (UBIM 14, 19, 21, and 22) bind to UMR-106 cells, a cell line derived from another rat osteosarcoma, which also exhibits several characteristics of OBS.(’~) It is of interest to note that none of the in vivo OB- and chondrocyte-restricted MAbs (group A) bind to UMR-106 cells. This cell line has also been used extensively as an in vitro model of OBs by many The OB-like cell lines ROS 17/2.8 and UMR-106 may represent OB tumors of different stages of differentiation or may indicate the existence of subsets of mature functional OBs. Four different immunization protocols were used in this study. These different methods were chosen in an attempt to increase the possibility of obtaining MAbs to OB-specific molecules. ROS 17/2.8 cells were treated with dexamethasone, which enhances the differentiated characteristics of OBs, such as alkaline phosphatase activity.(37)and PTH receptors.(”’ ROS 17/23 cells were also coated with polyclonal antisera prior to immunization. This procedure has been used successfully in the selection of MAbs specific to myeloid cellst2*)and Dictostelium. ( 3 9 ) However, in this study, these procedures did not appear to result in any significant increase in the number or specificity of MAbs obtained. At present, few MAbs recognizing osteogenic cells have been reported. Nijweide and Mulder(61have described a MAb raised against chick OBs that recognizes only the terminal differentiated stage of the OB, the osteocyte. Nakano et al.(*) have reported the production of four MAbs to rat OBs. Two of these (AOB-1 and AOB-2), both derived from a common parent hybridoma, recognize specifically the OB phenotype. Biochemical analysis of the antigen(s) identified by these antibodies indicated recogni-

tion of 11 bands, ranging in relative molecular size from 15,000 to 210,000, after Western blotting of SDS gels. The antigenic relationship between these components remains to be determined. We report here the production of a further five MAbs that recognize OB- and chondrocyte-restricted antigens. These MAbs recognize antigens expressed by cells of the stromal system in cultured rat bone marrow, and it is interesting to postulate that these cells may be the osteogenic precursors of bone. The panel of MAbs raised in this study will provide important tools for the study of both the ontogeny and physiology of the normal osteoblast and the value of in vitro cultured osteoblast-like cell lines as model systems for such studies.

ACKNOWLEDGMENTS We wish to thank Dr. A. Howie and J. Gregory for their assistance in immunohistochemistry and Dr. J. Green for help in setting up the techniques for producing monoclonal antibodies. This work has been presented in part at the 20th European Symposium on Calcified Tissues, Sirmione, Italy, October 48,1987, and at the 21st European Symposium on Calcified Tissues, Jerusalem, Israel, March 12- 16, 1989.

REFERENCES 1. Rodan GA, Rodan SB 1984 Expression of the osteoblast phenotype. In: Peck WA (ed) Bone and Mineral Research, Annual 2, Elsevier, Amsterdam, pp. 244-285. 2. Oursler M, Bell I, Cleminger B, Osdoby P 1985 Identifica-

tion of osteoclast-specific monoclonal antibodies. J Cell Biol 100:1592-1600. 3. Horton AM 1988 Osteoclast-specific antigens. In: IS1 Atlas of Science: Immunology, vol. 1, pp. 35-43. 4. Janossy G, Bofill M, Poulter L 1986 Two-colour immunofluorescence: Analysis of the lymphoid system with monoclonal antibodies. In: Polak J, Van Noorden S (eds) Immunocytochemistry; Modern Methods and Applications, 2nd ed., Wright, Bristol, pp. 438-455. 5 . Greaves MF, Delia D, Robinson J, Sutherland R, Newman R 1981 Exploitation of monoclonal antibodies: A “who’s who” of haemopoietic malignancy. Blood Cells 7:257-280. 6. Nijweide PJ, Mulder RJP 1986 Identification of osteocytes

in osteoblast-like cell cultures using a monoclonal antibody specifically directed against osteocytes. Histochemistry 84: 342-347. 7. Weinshank RL, Cain CD, Vasquez NP, Luben RA 1985

Identification of monoclonal antibody which interacts with the parathyroid hormone receptor-adenylate cyclase system in murine bone. Mol Cell Endocrinol 41:237-246. 8. Nakano T, Kimoto S, Tanikawa K, Kim K. Higaki M, Kawase T, Saito S 1989 Identification of osteoblast-specific monoclonal antibodies. Calcif Tissue Int 44:220-227. 9. Majeska RJ, Rodan SB, Rodan GA 1980 Parathyroid hormone-responsive clonal cell lines from rat osteosarcoma. Endocrinology 107:1494-1503. 10. Rodan GA, Majeska RJ, Wiren KM, Rodan SB 1984 Expres-

200 sion of hormonal effects in osteosarcoma osteoblastic cells. In: Cohn DV, Fujita T, Potts JR, Talmage RV (eds) Endocrine Control of Bone and Calcium Metabolism, Elsevier, Amsterdam, pp. 117-124. 1 1 . Rodan GA, Majeska RJ 1983 Phenotypic maturation of osteoblastic osteosarcoma cells in culture. In: Kelley RO, Goetnick P F (eds) Limb Development and Regeneration, Part B. Alan R. Liss, New York, pp. 249-259. 12. Hunter WM, Greenwood FC 1%2 Preparation of iodine-131 labelled human growth hormone of high specific activity. Nature 1W495-496. 13. Partridge NC, Frampton RJ, Eisman JA, Michelangeli VP, Elms E, Bradley TR, Martin TJ 1980 Receptors for 1,25(OH),-vitamin D, enriched in cloned osteoblast-like rat osteogenic sarcoma cells. Febs Lett 115139-142. 14. Topp WC 1981 Normal rat cell lines deficient in nuclear thymidine kinase. Virology 113:408-411. 15. Wiebel FJ, Kiefer F, Murdia US 1984 Phenobarbital induces cytochrome P-450- and cytochrome P-448-dependent monooxygenases in rat hepatoma cells. Chem Biol Interact 52~151162. 16. Tashjian AH, Yasumura Y, Levine L, Sato GH, Parker ML 1968 Establishment of clonal strains of rat pituitary tumour cells that secrete growth hormone. Endocrinology 82342352. 17. Gazdar AF, Chick WL, Oie HK, Sims HL, King DL, Weir GC, Lauris V 1980 Continuous, clonal, insulin and somatostatin secreting cell lines established from a transplantable rat islet cell tumour. Proc Natl Acad Sci USA 77:3519-3523. 18. Ambesi-Impiombato FS, Parks LAM, Coon HG 1980 Culture of hormone dependent functional epithelial cells from rat thyroids. Proc Natl Acad Sci USA 77:3455-3459. 19. Galfre G, Milstein C 1981 Preparation of monoclonal antibodies: Strategies and procedures. Methods Enzymol 73B: 3-46. 20. Fisher A, Bunce CM, Toksoz D, Stone PW, Brown G 1982 Studies of human myeloid antigens using monoclonal antibodies and variant lines from the promyeloid cell line HL60. Clin Exp lmmunol 50:374-381. 21. Galfre G, Howe SC, Milstein C 1977 Antibodies to major histocompatability antigens produced by hybrid cell lines. Nature 266550-552. 22. Suter L, Bruggen J, Sorg C 1980 Use of an enzyme linked immunosorbant assay (ELISA) for screening of hybridoma antibodies against cell surface antigens. J Immunol Methods 39407-41 1 . 23. Fukumoto T, McMaster WR, Williams AJ 1982 Mouse monoclonal antibodies against rat major histocompatability antigens. Two la antigens and expression of la and class 1 antigens in rat thymus. Eur J Immunol 12:237-243. 24. Jones JL, Callow JA, Green JR 1988 Monoclonal antibodies to sperm surface antigens of the brown alga Fucus serrutw exhibit region. gamete, species and genus preferential binding. Planta 176298-306. 25. Thomas ML, Green JR 1983 Molecular nature of the W3/25 and MRC OX-8 marker antigens for rat T lymphocytes: Comparisons with mouse and human antigens. Eur J Immuno1 13:855-858.

PERRY ET AL. 26. Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680684. 27. Laskey RA, Mills AD 1975 Quantitative film detection of 'H and I4C in polyacrylamide gels by fluorography. Eur J Biochem 56:335-341. 28. Mardon HJ, Bee J, Von der Mark K. Owen ME 1987 Development of osteogenic tissue in diffusion chambers from early precursor cells in bone marrow of adult rats. Cell Tissue Res 250:157-165. 29. Heusser GH, Stocker JW, Gister RH 1981 Methods for binding cells to plastic: Application to solid phase immunoassays for cell surface antigens. Methods Enzymol 73:406-418. 30. Owen M 1987 The osteogenic potential of bone marrow. In: Sen A, Thornhill T (eds) Development and Diseases of Cartilage and Bone Matrix. Alan R. Liss, New York, pp. 247-255. 31. Ashton BA, Allen TD, Howlett CR, Eaglesom CC, Hattori A, Owen M 1980 Formation of bone and cartilage by marrow stromal cells in diffusion chambers in vivo. Clin Orthop 151~294-307. 32. Triffitt JT 1987 The special proteins of bone tissue. Clin Sci 72~399-408. 33. Fisher L, Hawkins G, Tuross N, Termine J 1987 Purification and partial characterisation of small proteoglycans I and 11, and osteonectin from the mineral compartment of developing human bone. J Biol Chem 262:9702-9708. 34. Gutierrez GE, Mundy GR, Derynck R, Hewlett EL, Katz MS 1987 Inhibition of parathyroid hormone responsive adenylate cyclase in clonal rat osteoblast-like cells by TGFa and EGF. J Biol Chem 262:15845-15850. 35. Robey PG, Young MF, Flanders KC, Roche NS, Kondaiah P, Reddi AH, Termine JD, Sporn MB, Roberts AB 1987 Osteoblasts synthesise and respond to TGF type fl in vitro. J Cell Biol 105457-463. 36. Gray TK, Flynn TC, Gray KM, Nabell LM 1987 17-&Estradiol acts directly on the clonal osteoblastic cell line UMR 106. Proc Natl Acad Sci USA 84:6267-6271. 37. Majeska RJ, Nair BC, Rodan GA 1985 Glucocorticoid regulation of alkaline phosphatase in the osteoblastic osteosarcoma cell line ROS 17/23. Endocrinology 116178-179. 38. Rodan SB, Fischer MK, Egan JJ, Epstein PM, Rodan GA 1984 The effect of dexamethasone on parathyroid hormone stimulation of adenylate cyclase in ROS 17/2.8 cells. Endocrinology 11995 1-957. 39. Barclay SL, Smith AM 1986 Rapid isolation of monoclonal antibodies specific for cell surface differentiation antigens. Proc Natl Acad Sci USA 83:4336-4340.

Address reprint requests to: Moira Gilligan Department of Medicine Queen Elizabeth Hospital Edgbaston, Birmingham, BI5 2TH, England Received for publication May 12, 1989; in revised form July 19, 1989; accepted July 20, 1989.

2.8 cells recognize antigens, some of which are restricted to osteoblasts and chondrocytes.

We have raised a panel of 15 monoclonal antibodies (MAbs) recognizing cell surface antigens of the rat osteoblast-like cell line ROS 17/2.8. The MAbs ...
3MB Sizes 0 Downloads 0 Views