Murine Cell Surface Glycoproteins CHARACTERIZATION OF A MAJOR COMPONENT OF 80,000 DALTONS AS A POLYMORPHIC DIFFERENTIATION ANTIGEN OF MESENCHYMAL CELLS*

(Received for publication, February 2, 1981)

Edward N. Hughes$, Guadalupe Mengod, and J. Thomas August@ From the Department of Pharmacology and Experimental Therapeutics, The Johns Hopkins University School of Medicine, Baltimore, Mavland 21205

Monoclonal antibodies reactive with NIH/3T3 cell surface antigens were obtained from hybridomas of murine myeloma cells fused to spleen cells of rats im- munized with NIH/3T3 cell plasma membranes. Four of the antibodies, of forty that have been studied, ap- peared to react with allospecific antigenic determi- nants: they bound to NIH/3T3 cells but not to BALB/ 3T3 cells. Each of these four antibodies immunoprecip- itated a glycoprotein of about 80,000 daltons that mi- grated to an isoelectric point of about pH 5.0. Polypep- tides of identical molecular weight and isoelectric points, and yielding the same proteolytic cleavage frag- ments, were present in B&B/3T3 cells, but were not antigenically reactive. The 80,000-dalton glycoprotein was a major constituent of the plasma membrane. It was a predominant lactoperoxidase iodinated compo- nent of intact NIH/3T3 cells, and saturation binding of ‘251-labeled antibody indicated that there were about 10‘ antigenic sites/cell. Studies of the distribution of the immunoreactive glycoprotein among different strains of mice confirmed the polymorphic expression of the determinant: Spleen cells of BALB/c, DBA/l, DBA/2, and CBA mice did not bind anti-80,OOO-dalton glycoprotein monoclonal antibodies, whereas spleen cells of a large number of other strains of mice were positive for antibody-binding. The antigenic reactivity varied markedly among different cell lines and was greatest with the NIH/3T3 mouse embryo fibroblast, G8-1 Swiss Webster myoblast, and IC-21 SV40-trans- formed C57BL/6 mouse peritoneal macrophage. The properties ofthe 80,000-dalton glycoprotein character- ized this molecule as a new cell surface differentiation alloantigen of murine mesenchymal cells.

Alloantigens are properties of a number of cell surface proteins that are specified by polymorphic genes. All of‘ the known proteins of the closely linked major histocompatibility complex genes are polymorphic; these include the H-2 (l), thymus leukemia (TL! (Zj, Qa (31, and immune response (Ia) (4, 5) alloantigens. Several lymphocyte cell surface differen- tiation proteins are alloantigens: the Thy-1 antigen (6), pro- teins of the Lyt series (?, 8), proteins of the Lyb series (9), Pca-1 (lo), the T-200 (11,12), and leucocyte common antigen

* This research was supported by National Institutes of Health Grant RO1 CA19471. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “nduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

the National Institutes of Health. $ Predoctoral fellow supported by Training Grant GM-00184 from

8 To whom correspondence should be addressed.

(13,14), and a recently described glycoprotein of about 100,000 daltons (15). Many murine cells also express on their cell surface the highly polymorphic proteins of endogenous ret- roviruses, particularly the viral glycoproteins of about 70,000 daltons (16).

The importance of these alloantigens is that they identify proteins that have been implicated in a number of cell recog- nition, receptor, and differentiation phenomena. For example, complex genetic polymorphism appears to be a fundamental property of the major histocompatibility system. Other al- loantigens have been found primarily on lymphoid cells. The discovery of the Lyt-1 and Lyt-2 alloantigens, by Boyse and co-workers (Z), pioneered the identification of numerous ather polymorphic antigens that distinguish functional subpopula- tions of hematopoietic cells (17).

We previously have reported a collection of hybridoma monoclonal antibodies prepared by fusion of mouse myeloma cells and spleen cells of rats immunized with NIH/3T3 cells or plasma membranes (18). Hybridoma cell lines that secreted antibodies which bound to intact NfH/3T3 cells were selected and cloned. By use of these antibodies we have identified a number of previously uncharacterized cell surface proteins (18). The present report describes several distinctive features of one of these components, a glycoprotein of about 80,000 daltons. 1) It was a major component of the cell surface with about 10” sites/cell. 2) Four independent monoclonal antibod- ies that specifically precipitated this glycoprotein each reacted with an antigenic determinant that was polymorphic and was present only in cells and tissues of certain strains of mice. 3 ) The glycoprotein showed a unique pattern of cell expression; among a variety of cell lines of diverse strain and tissue origin, it was detected specifically on fibroblast, myoblast, and mac- rophage cell lines. The properties of the 80,000-dalton glyco- protein defined this molecule as a previously undescribed, allospecific, differentiation antigen of the murine cell surface.

EXPERIMENTAL PROCEDURES

Animals-All inbred strains of mice 6 to 10 weeks of age were obtained from Jackson Laboratories (Bar Harbor, ME) except for National Institutes of Health Swiss mice, which were from M. A. Bioproducts (Rockville, MD).

Antibodies-Hybridoma antibodies AMF-8 and AMF-12 were ob- tained from independent clones isolated from the same cell fusion experiment, as previously described (18). AMF-15 and AMF-16 hy- bridoma antibodies were obtained in a similar manner, but from two different cell fusion experiments. The sources of hybridoma antibodies were either the hybrid cell cultwe supernatants concentrated 20-fold by ammonium sulfate precipitation or ascites fluids from BALB/c nu/nu mice bearing hybrid cell tumors. Hybridoma antibody AMF- 12 was purified from ascites fluids by affinity chromatography with Protein-A Sepharose (19). Purified goat IgG’ anti-rat IgG has been

’ The abbreviations used are: IgG, immunoglobulin G; gp80, 80,OOO- dalton glycoprotein; Lgp100, 100,000-dalton glycoprotein.

7023

7024 80,000-Dalton Cell Surface Alloantigen

described (18). Purified rabbit F(ab')* anti-rat IgG was a gift of Dr. Alfonso Colombatti of the Johns Hopkins University School of Med- icine. Antibodies were labeled with lZ5I by the method of Jensenius and WiUiams (20). Class-specific antisera against rat immunoglobulins were purchased from Miles Laboratories (Elkhart, IN).

Antibody- binding Assays-Trace binding assays were performed by incubation of 1 X lo5 to 1 X lo7 freshly prepared spleen cells with 20 ng of purified '251-labeled AMF-12 antibody for 1 h at 0 "C in 100 pJ of Dulbecco's phosphate-buffered saline containing 10 mg/ml of bovine serum albumin (final concentration). The cells were collected by centrifugation, washed three times, and the bound radioactivity was counted.

Saturation binding assays were performed as described by Williams et al. (21). NIH/3T3 cells (2.5 X lo5 glutaraldehyde-fixed cells) in 50 pl were incubated with 50 p1 of concentrated hybrid cell culture supernatant for 1 h at 0 "C and washed three times. The cells were then incubated with various amounts of '2sII-labeled rabbit F(ab):, anti-rat IgG antibody (5 to 8 pCi/pg) in 50 p1 for 1 h at 0 "C. After washing the cells, the radioactivity was counted.

The conditions for the preparation of cell extracts and the solid phase cell extract-binding assay have been described in detail (18).

CeZls"BALB/3T3, NIH/3T3, and NIH/3T3 transformed by Harvey murine sarcoma virus were described (22-24). The SV40- transformed macrophage line IC-21 was obtained through the cour- tesy of Dr. V. Defendi of New York University Medical Center (25). Dr. R. Schnaar of the Johns Hopkins University School of Medicine provided neuroblastoma S-26 (26) and myoblast G8-1 (27). Embryonal carcinoma cells OTTFL were obtained from Dr. K. Huebner of the Wistar Institute of Anatomy and Biology (28). All other cell lines were from the Cell Distribution Center (Salk Institute, San Diego, CA).

Radiolabeling and Immunoprecipitation of Cell Proteins-Intact cells were labeled with "'1 (carrier-free sodium salt, Amersham/ Searle) under conditions similar to those reported by Marchalonis et al. (29) using lactoperoxidase and hydrogen peroxide. Extraction and immunoprecipitation of cell proteins was carried out as described (18).

Partial Proteolytic Peptide Mapping-Cell surface proteins la- beled with "'I and isolated by immunoprecipitation were excised from one-dimensional sodium dodecyl sulfate polyacrylamide gels. Limited proteolytic peptide mapping was carried out as described by Cleve- land et al. (30).

Two-dimensional Gel Electrophoresis-Two-dimensional gel elec- trophoresis was performed as described by O'Farrell(31). Cell extracts or immunoprecipitated cell proteins were suspended in 50 1.11 of OFarrell buffer A (9.5 M urea, 2% (w/v) Nonidet P-40, 2% (w/v) Ampholines, and 5% (v/v) 2-mercaptoethanol). Following isoelectric focusing, the samples were analyzed by discontinuous sodium dodecyl sulfate polyacrylamide gel electrophoresis on 10% gel slabs (32).

RESULTS

Characterization of Monoclonal Antibodies that Bound to NIH/3T3 Fibroblasts but not to BALB/3T3 Fibroblasts- Antibodies that reacted with cell surface antigens of NIH/3T3 cells were generated by immunizing rats with a partially purified plasma membrane fraction from NIH/3T3 cells, as previously described (18). Spleen cells of the immune rats were fused with the murine myeloma cell line P3-NSI/l-Ag4- 1. Hybridomas that produced antibodies which reacted with surface antigens of NIH/3T3 cells were selected and cloned.

Four stable cultures (AMF-8, AMF-12, AMF-15, and AMF- 16), from a total of 40 that have been characterized, produced antibodies which bound to antigens of NIH/3T3 but not BALB/3T3 cells (Fig. 1). Antibody-binding curves with a constant saturating amount of monoclonal antibody and in- creasing amounts of NIH/3T3 cell extracts showed an antigen- dependent binding of each of the four monoclonal antibodies? In contrast, monoclonal antibody binding to BALB/3T3 cell extracts was not observed at any antigen concentration. As a

' One of the monoclonal antibodies, AMF-15, gave greater values in the indirect binding assay than did the other three antibodies. We have not yet determined whether this is due to a higher affinity of the AMF-15 for the antigen or greater binding of the second antibody to the AMF-15 antibody.

- 0 c P C

0

Y

a m

2.0- /+ e!

P

1.0 -

L.*-i,-i , i-ei j 25 50 128

Cell Extract (pg Protein)

FIG. 1. Monoclonal antibody binding to NIH/3T3 and BALB/3T3 cell antigens. Monoclonal antibody binding was meas- ured by the solid phase cell extract assay as described (16), with 50 p.l of hybridoma culture supernatants (AMF-5, A-A; AMF-8,

p1 of '2'I-labeled goat anti-rat IgG antibody solution (0.25 pg/ml; 50 pCi/pg), and different amounts of protein from NIH/3T3 or BALB/ 3T3 cell extracts, as indicated.

X " X ; AMF-12, V-V; AMF-15, M, AMF-16, H), 50

control, another hybridoma antibody which bound to an un- related antigen, a 90,000-dalton glycoprotein, was included in the same experiment; in this case the binding curves were the same with both cell lines. The specificity for NIH/3T3 cells was observed with all antigen preparations, including intact cells, glutaraldehyde-fixed cells, and cell extracts.

Each of these cloned hybridomas secreted a heavy chain of 52,000 daltons and a single light chain, as determined by sodium dodecyl sulfate polyacrylamide gel electrophoresis of reduced immunoglobulins metabolically labeled with [%]me- thionine. In each case, the apparent molecular weight of the light chains was different from that of the K-light chain of the P3-NSI/l-Ag4-1 cell line. Double diffusion analysis, using class-specific antisera, showed that two of the antibodies (AMF-18 and AMF-15) reacted only with anti-rat IgGt, serum; and two (AMF-8 and AMF-16), with anti-rat IgGi serum (data not shown). There was conclusive evidence that each hybrid- oma clone represented a single, unique cell fusion event and was derived from a different antibody-forming cell. The four hybridomas were obtained from three independent experi- ments, 2.e. from three different immunized rats and three different cell fusion experiments (AMF-8 and 12, AMF-15, AMF-16). In addition, although the AMF-8 and AMF-12 hybridomas were derived from the same hybridization exper- iment, the secreted antibodies were of different subclasses.

165= I55

90- 94 -

68 -

43 -

165- 155-

94 - 90- 68-

43-

165, 155

94- 90-

68-

43-

7. I I '

a

b

6;8

80,000-Dalton Cell Surface Alloantigen

6;4 5.4 4.5 1 1 1 1

P

C

FIG. 2. Identification of gp80 in two-dimensional gel auto- radiograms of NIH/3T3 and BALB/3T3 LZ51-labeled cell surface proteins. Cell surface proteins were labeled with ""I, extracted with 0.5% Nonidet P-40 and resolved by two-dimensional gel electropho- resis as described under "Experimental Procedures." A, total "'I- labeled NIH/3T3 cell extract (6.7 X 10" cpm); B, '"I-labeled NIH/ 3T3 cell proteins immunoprecipitated with the AMF-16 antibody (5 X 10' cpm); C, total "'I-labeled BALB/3T3 cell extract (7.5 X 10" cpm). The arrows indicate those polypeptides identified as gp80, glycoproteins with an apparent molecular weight of about 80.000 and isoelectric point of about 5.0. The selected region of the two-dimen- sional gel is shown by appropriate molecular weight and pH markers.

Identification of a n 80,000-dalton Protein Antigen by Im- munoprecipitation-The antigenic target of the monoclonal antibodies was determined by immunoprecipitation and two- dimensional gel electrophoresis of proteins from NIH/3T3 cells labeled externally with '"I. A polypeptide of 80,000 daltons and an isoelectric point of pH 5.0 was a major I2'I- labeled component of the cell surface (Fig. 2 .4 ) . Other studies showed that this molecule biosynthetically incorporated fu- cose.:' Each of the four monoclonal antibodies immunoprecip- itated this glycoprotein (Fig. 2B). The identities of the four immunoprecipitated glycoproteins were fur ther confi ied by partial proteolytic peptide mapping (data not shown).

The two-dimensional gel pattern of labeled proteins from the BALB/3T3 cell surface was identical with that from NIH/ 3T3 cells (Fig. 2C). A major iodinated polypeptide of 80,000 daltons and an isoelectric point of pH 5.0 was present in BALB/3T3 cells, but was not immunoprecipitated by any of the four monoclonal antibodies. Evidence that the BALB/ 3T3 and NIH/3T3 glycoproteins were homologous was ob-

'' G . Mengod, E. N. Hughes, and J . T. August, unpublished obser- vations.

A B

I 0 0

7025

C D E F

FIG. 3. Partial proteolytic peptide maps of 12SI-labeled gp80 from NIH/3T3 and BALB/3T3 cells. "'I-Labeled polypeptides isolated by immunoprecipitation as previously described (18) were excised from one-dimensional gels and analyzed by partial proteolytic peptide mapping (30). The gp80 from NIH/3T3 cells was digested by incubation with A, 1.0 ng; B, 10 ng; or C, 750 ng of papain. The gp80 from BALB/3T3 cells was digested by incubation with D, 1.0 ng; E , 10 ng; or F, 750 ng of papain.

tained by the analysis of the peptide products of proteolytic cleavage. For this experiment, the 80,000-dalton glycoproteins of both NIH/3T3 and BALB/3T3 cells were immunoprecipi- tated by the polyclonal serum of the immunized rat whose spleen cells were used for the cell fusion procedure." The individual immunoprecipitated glycoproteins were purified by polyacrylamide gel electrophoresis and analyzed by partial proteolytic cleavage (Fig. 3). The patterns were identical, confirming the primary structural homology between these two immunologically different glycoproteins. Further charac- terization of the degree of amino acid sequence homology awaits the development of a monospecific antiserum that recognizes the 80,000-dalton glycoprotein expressed on BALB/3T3 cells.

The Number of Antigenic Sites per Cell-The number of antigenic sites of the 80,000-dalton glycoprotein expressed on the surface of NIH/3T3 cells was measured by the indirect binding assay a t saturating amounts of monoclonal antibody and "'I-labeled rabbit F(ab')z anti-rat IgG antibody. Satura- tion binding for the "'I-labeled antibody was obtained (Fig. 4). The binding was specific, as rat IgG of an irrelevant specificity from another hybridoma line (4B4.13) did not bind to NIH/3T3 cells. The number of gp80 antigenic sites per NIH/3T3 cell was calculated to be 0.7 X 10" to 1.4 X lo6, based upon the specific activity of the '"I-labeled F(ab')p antibodies, and a binding ratio of second to first antibody of 4:l to 2:l (33). The results indicated that the 80,000-dalton glycoprotein was a major cell surface constituent of NIH/3T3 cells.

Antigen Expression in Cells of Different Strains of Mice- The polymorphism of the 80,000-dalton glycoprotein was con- f i i e d by measuring the binding of purified, '"I-labeled AMF- 12 antibody to spleen cells of different strains of mice. The results showed an unambiguous positive and negative segre- gation of antigenic determinant expression. Binding of "'I- labeled anti-gp80 monoclonal antibodies to spleen cells of BALB/c, DBA/1, DBA/2, and CBA mice was insignificant as compared with the binding to cells of several other murine strains (Table I). The same pattern of expression of the antigenic determinant in different inbred mice was observed

The serum of the immunized rat was chosen for this experiment since it was expected that this polyclonal, xenogeneic serum would contain antibodies against a wide spectrum of antigenic determinants of the 80,000-dalton protein and would thus precipitate the homolo- gous proteins of BALB/3T3 cells.

7026 80,000-Dalton Cell Surface Alloantigen

I 2 3 '251-Ant~-lg Substrate ( p q )

FIG. 4. Saturation binding of AMF-15 antibody to NIH/3T3 cells. Glutaraldehyde-fixed NIH/3T3 cells (2.5 X 105/sample) were incubated for 1 h at 0 "C with 50 yl of 25-fold concentrated culture supernatants from the AMF-15 hybrid cell line (M) or from the 4B4.13 hybrid cell line ( O " - o ) . After washing, the target cells were then incubated for 1 h at 0 "C with increasing amounts of '251-labeled rabbit F(ab'Iz anti-rat IgG antibody (0.8 yCi/yg), washed, and the bound radioactivity was counted. The results represent averages of duplicate samples.

TABLE I Antibody binding to spleen cells from different murine strains Binding of purified, 1251-labeled AMF-12 antibody to 5 X 10" spleen

cells was measured as described under "Experimental Procedures." Mouse strain Antibody bound

ng Positive

NIH Swiss 1.9 C3H/HeJ 2.2 A/J 1.9 C58/J 1.7 C57BL/6J 1.5 C57BR/cdJ 1.6 AKR/J 1.3 SWR/J 1.7 SJL/J 1.2 129/J 1.5

BALBc/J t o . 1 DBA/lJ <O.l DBA/ZJ 4 . 1 CBA/J <O.l

Negative

-

with each of the four anti-gp80 monoclonal antibodies." Antigen Expression in Different Cell Lines-The effect of

cell differentiation on the expression of the 80,000-dalton glycoprotein was analyzed by measuring antibody binding to extracts of 21 different cultured cell lines. Antigen-dependent binding of the anti-gp80 monoclonal antibodies was observed only with extracts of NIH/3T3 fibroblast cells, NIH/3T3 cells transformed by Harvey sarcoma virus, G8-1 cells (Swiss Webs- ter myoblasts), and IC-21 cells (C57BL/6 SV40-transformed macrophages) (Fig. 5). The antibodies did not react with extracts of several other cell lines derived from mouse strains that were positive for spleen tissue expression of the antigen; these included Lewis lung (a C57BL/6 carcinoma), EL4 (a C57BL/6 T-cell lymphoma), Sarc 180 (a Swiss Webster sar- coma), OTTFl (a 129 embryonal carcinoma), and S-26 (an A/ J neuroblastoma). None of the antibodies bound to extracts prepared from any cell line of DBA/2 or BALB/c origin, such as S49.1 (a BALB/c T-cell lymphoma), D2N (a DBA/2 eryth- roid leukemia), 5774 (a BALB/c macrophage line), P388D1 (a DBA/2 macrophage line), P815 (a DBA/2 mastocytoma), and WEHI-3 (a BALB/c myelomonocytic leukemia) (18).

Cell transformation by oncogenic viruses had no apparent effect in the expression of the antigen. NIH/3T3 cells trans-

, 2.0 /"-- 1

- C

4 vi n N -

OTTFf Sarc 180

50 100 Cell Extracts (pg Protem)

FIG. 5. Monoclonal antibody binding to extracts of cells from various strain and tissue origins. Different amounts of protein from various cell extracts were adsorbed to plastic wells. The extracts were then incubated with 50 yl of 15-fold concentrated culture super- natants of the AMF-12 hybrid cell line. Hybridoma antibody-binding to cell antigens was measured with the solid phase assay as described (18) using 50 p1 of a 12511-labeled goat IgG anti-rat IgG antibody solution (0.25 pg/ml; 25 yCi/pg).

formed by Harvey murine sarcoma virus continued to express the antigen and there was no antibody binding to BALB/3T3 transformed by Kirsten murine sarcoma virus or SV40 (data not shown).

DISCUSSION

A murine cell surface glycoprotein of about 80,000 daltons has been identified and studied by use of monoclonal antibod- ies. This glycoprotein was noteworthy in several respects. 1) It was one of the major constituents of the plasma membrane. 2) The determinants recognized by the monoclonal antibodies were allospecific. 3) This allospecific determinant acted as a predominant immunogen in a xenogenic host. 4) The expres- sion of the glycoprotein was specific to certain types of cells.

Saturation binding of anti-gp80 monoclonal antibodies to the surface of NIH/3T3 cells showed that there were 7 X lo5 to 1.4 X lo6 antigenic sites/ceU. This concentration was com- parable to that of Thy-], the major cell surface glycoprotein of thymocytes, with 6 X lo5 molecules/cell (34). In keeping with this finding, the 80,000-dalton glycoprotein was the most prominent Iz5I-labeled plasma membrane component of NIH/ 3T3 and BALB/3T3 cells, as distinguished by two-dimen- sional polyacrylamide gel electrophoresis. In addition, Bretscher et al. (35) have confirmed the plasma membrane localization of the antigen and have shown by immunoelec- tronmicroscopy with the AMF-12 antibody that the 80,000- dalton glycoprotein, as well as Thy-1, was uniformly distrib- uted throughout the surface of NIH/3T3 cells, except for the coated pits. These data indicate that the 80,000-dalton glyco- protein, with its relatively large mass and high concentration, is one of the major protein components of the plasma mem- brane of the murine fibroblast.

Another remarkable property of the 80,000-dalton glycopro- tein was its polymorphism. The antigenic determinants rec- ognized by the monoclonal antibodies were present in tissues of several strains of mice, but were not detected in tissues of BALB/c, DBA/1, DBA/2, or CBA mice. This difference in antigen reactivity between the strains may be attributed to genetic polymorphism, rather than lack of gene expression, as a major iodinated protein of identical electrophoretic mobility, isoelectric point, and iodinated tryptic peptide composition was present in BALB/3T3 cells. Whether the serologically

80,000-Dalton Cell Surface Alloantigen 7027

detected differences between the 80,000-dalton glycoproteins are the result of variations in the carbohydrate and/or protein portion of the molecule remains to be determined. The 80,000- dalton glycoprotein appears to be different from all other murine cell surface alloantigens identified by polyclonal and monoclonal alloantisera. These include the Ala-1, H-2, Ia, Lyb series, Lyt series, MIS, Mph-1, Pca-1, Qa, and TLa alloantigens (for a review, see Ref. 9). The newly described LgplOO (15), H9/25 (36), and Ly-10.1 (37) alloantigens identified by mon- oclonal antibodies can also be excluded. None of the above other alloantigens showed the same distribution of antigen expression among different strains of mice as did the gp80 and none has been assigned to a molecular species of 80,000 daltons.

The finding of four xenogeneic monoclonal antibodies, each of which reacted with an allospecific determinant of the same protein, was unexpected. Xenogeneic immunizations usually elicit responses to a wide range of antigenic determinants because of the evolutionary changes that occur in many proteins (33). It is for this reason that allospecific antisera have been prepared by selective immunization between dif- ferent inbred mice and therefore were raised against a re- stricted population of antigenic loci (6). Thus, the high fre- quency of response to a polymorphic determinant of the 80,000-dalton glycoprotein was remarkable. The four hybrid- oma cell lines that produced antibodies reactive with the gp80 were present among a total of 40 cloned hybridomas that have been characterized. This frequency was exceeded only by hybridomas secreting antibodies to a 90,000-dalton protein, for which there were six independent isolates. As the four monoclonal antibodies of IgGl and IgG2, subclasses were obtained from splenocytes of three different immunized rats and three separate cell fusion experiments, the multiplicity of these anti-gp80 hybrid clones cannot be attributed to activa- tion and division of a single antibody-forming precursor cell. Moreover, studies in progress suggested that each of the four antibodies acted on the same or proximate allospecific anti- genic determinant.3 This is to be compared with the reported xenogeneic, monoclonal rat antibodies against the Lyt-1, Lyt- 2, Lyt-3, Thy-1, and T-200 molecules, which reacted with framework determinants as well as the allospecific determi- nants of the corresponding polymorphic molecules (38). The glycoprotein is further emphasized by the relatively uncom- mon detection of alloantigens by xenogeneic monoclonal an- tibodies. In continuing studies in this laboratory, among more than 40 monoclonal antibodies that have been tested, alloan- tigenic specificity was observed only with the anti-gp80 anti- bodies. Springer et al. (39) tested 10 rat anti-murine lympho- cyte monoclonal antibodies against a large number of murine strains; no antigenic polymorphism was detected.

The presence of an allospecific immunodominant antigenic site on the 80,000-dalton glycoprotein suggested that this protein might also characterize cellular differentiation. The remarkable finding was that the antigenic determinant reac- tive with the anti-gp80 monoclonal antibodies was present on fibroblast lines from Swiss mice and on a macrophage cell line from C57BL/6 mice, but it was not detected on a C57BL/6 T- cell lymphoma, a C57BL/6 lung carcinoma, an A/J neuro- blastoma, or a 129/J embryonal carcinoma. In other studies, we have also found that the 80,000-dalton glycoprotein was a major differentiation antigen of hematopoietic t i s s~es .~ Ra- dioimmune precipitation and fluorescence-activated cell-sort- ing analyses showed that the glycoprotein antigen character- ized the phagocytic cell lineage, as it was present on macro- phages, granulocytes, monocytes, and many adult bone mar-

"E. N. Hughes, A. Colombatti, and J. T. August, unpublished observations.

row cells but was absent on other lymphoid cells. Thus, among the cells tested, the expression of the glycoprotein was re- stricted to certain types of mesenchymal cells.

Acknowledgments-We are indebted to Drs. Mette Strand and Alfonso Colombatti for helpful discussions and critical reading of this manuscript. We wish to thank Sander Cohen, James Freedy, and Debra Gilbert for their assistance in these experiments and Lydia Cumor for the preparation of the manuscript.

1.

2. 3.

4. 5. 6. 7.

8.

9.

10.

11.

12.

13.

14.

15.

16. 17. 18.

19.

20.

21.

22.

23. 24. 25. 26.

27.

28.

29.

REFERENCES

Nathenson, S. G., and Cullen, S. E. (1974) Biochim. Biophys.

Boyse, E. A., and Old, L. J. (1969) Annu. Rev. Genet. 3,269-290 Stanton, T. H., and Boyse, E. A. (1976) Immunogenetics 3, 525-

Klein, J., and Hauptfeld, V. (1976) Transplant. Reu. 30,83-100 David, C. S. (1976) Transplant. Rev. 30,299-322 Reif, A. E., and Allen, J. M. V. (1964) J. Exp. Med. 120, 413-433 Boyse, E. A,, Miyazawa, M., Aoki, T., and Old, L. J. (1968) Proc.

Boyse, E. A,, Hakura, K., Stockert, L., Iritani, C. A,, and Miura,

McKenzie, I. F. C., and Potter, T . (1979) Adu. Immunol. 27, 179-

Tung, J.-S., Shen, F.-W., Boyse, E. A., and Fleissner, E. (1978)

Trowbridge, I. S., Nilsen-Hamilton, M., Hamilton, R. J., and

Omary, M. B., Trowbridge, I. S., and Scheid, M. P. (1980) J. Exp.

Sunderland, C. A,, McMaster, W. R., and Williams, A. F. (1979)

Carter, P. B., and Sunderland, C. A. (1980) Immunogenetics 10,

Ledbetter, J. A., Goding, J . W., Tsu, T. T., and Herzenberg, L. A.

Strand, M., and August, J . T. (1974) J. Virol. 13, 171-180 Cantor, H., and Boyse, E. A. (1975) J. Exp. Med. 141, 1376-1389 Hughes, E. N., and August, J . T. (1981) J. Biol. Chem. 256,664-

Ey, P. L., Prowse, S. J., and Jenkin, C. R. (1978) Immunochem-

Jensenius, J. C., and Williams, A. F. (1974) Eur. J. Immunol. 4,

Williams, A. F., Galfre, G., and Milstein, C. (1977) Cell 12, 663-

Strand, M., and August, J. T. (1979) Proc. Natl. Acad. Sci. U. S.

Strand, M., and August, J. T. (1978) Cell 13, 399-408 Scolnick, E. M., and Parks, W. P. (1974) J. Virol. 13, 1211-1219 Mauel, J., and Defendi, V. (1971) J. Exp. Med. 134, 335-350 Seeds, N. W., Gilman, A. G., Amano, T., and Nirenberg, M. W.

Christian, C. N., Nelson, P. G., Peacock, J., and Nirenberg, M. W.

Huebner, K., Tsuchida, N., Green, C., and Croce, C. M. (1979) J.

Marchalonis, J. J., Cone, R. E., and Santer, V. (1971) Biochem. J.

Acta 344, 1-25

531

R. SOC. Lond. Ser. B Biol. Sci. 170, 175

M. (1971) Transplantation 11, 351-353

338

Zmrnunogenetics 66,101-105

Bevan, M. J . (1977) Biochem. J. 163,211-217

Med. 151, 1311-1316

Eur. J. Zmmunol. 9, 155-159

583-593

(1979) Immunogenetics 8,347-360

671

istry 15,429-436

91-97

673

A. 74,2729-2733

(1970) Proc. Natl. Acad. Sci. U. S. A. 66, 160-167

(1977) Science (Wash. D. C.) 196, 995-998

Exp. Med. 150, 392-405

124.921-927 30. Cleveland, D. W., Fischer, S. G., Kirschner, M. W., and Laemmli,

U. K. (1977) J. Biol. Chem. 252, 1102-1106 31. O'Farrell, P. H. (1975) J. Biol. Chem. 250,4007-4021 32. Baum, S . G., Honvitz, M. S., and Maizel, J. V. (1972) J. Virol. 10,

33. Williams, A. F. (1977) Contemp. Top. Mol. Zmmunol. 6. 83-116 34. Acton, R. T., Morris, R. J., and Williams, A. F. (1974) Eur. J.

35. Bretscher, M. S., Thomson, J. N., and Pearse, M. F. (1980) Proc.

36. Takei, F., GalfrC, G., Alderson, T., Lennox, E. S., and Milstein, C.

37. Kimura, S., Tada, N., and Hammerling, U. (1980) Zmrnunogenet-

38. Ledbetter, J. A., and Herzenberg, L. A. (1979) Zmmunol. Reu. 47,

39. Springer, T., Galfre, G., Secher, D. S., and Milstein, C. (1978) Eur.

211-219

Zmmunol. 4, 598-602

Natl. Acad. Sci. U. S. A . 77,4156-4159

(1980) Eur. J . Immunol. 10, 241-246

ics 10, 363-372

63-90

J. Zmmunol. 8, 539-551