Proc. Natl. Acad. Sci. USAVol. 81, pp. 4434-4438, July 1984Cell Biology

Transformation-associated proteins in murine B-cell lymphomasthat are distinct from Abelson virus gene products

(monoclonal antibodies/immunoprecipitation/bone marrow cultures)

ERIC PILLEMER, CHERYL WHITLOCK, AND IRVING L. WEISSMANLaboratory of Experimental Oncology, Department of Pathology, Stanford University School of Medicine, Stanford, CA 94305

Communicated by Ray D. Owen, April 5, 1984

ABSTRACT In an effort to identify cellular proteins thatmay be involved in the Abelson murine leukemia virus (A-MuLV) transformation process, we have isolated a hybridomaantibody (6C3) that detects a tumor-associated antigen in allA-MuLV-induced pre-B-cell lymphomas. The 6C3 antibodyimmunoprecipitates two molecules of Mr 160,000 and Mr125,000 from metabolically labeled A-MuLV tumors. The twoproteins recognized by the 6C3 antibody are distinct from theA-MuLV-transforming protein in that they lack viral gag de-terminants and are neither phosphoproteins nor protein ki-nases. The 6C3 proteins can be detected in all A-MuLV pre-B-cell lymphomas and some nonviral B lymphomas but are notdetected on any other tumor or normal cell, including A-MuLV-transformed fibroblast lines. Thus, the 6C3 proteinsmay represent the products of novel cellular genes whoseexpression is induced, stabilized, or amplified in B-cell tumorsof both viral and nonviral origin. Further evidence in supportof this hypothesis is provided by the finding that 6C3 antigenexpression correlates with autonomous cell growth and thetransformed phenotype in both normal bone marrow culturesand those infected with A-MuLV.

Abelson murine leukemia virus (A-MuLV) is a rapidly trans-forming retrovirus that transforms fibroblasts and bone mar-row cells in vitro (1, 2) and induces lymphoid tumors in vivo(3). The product of the A-MuLV genome (v-abl) has beenshown to be a phosphoprotein that possesses an associatedprotein kinase activity capable of phosphorylating tyrosineresidues in vitro (4, 5). Although the expression of the A-MuLV gene product correlates with A-MuLV transforma-tion (6, 7), several recent findings suggest that A-MuLVexpression may not be necessary or sufficient for maintain-ing the transformed state. (i) Early after A-MuLV infectionof bone marrow cultures in vitro, cells require an adherentcell feeder layer for growth even though v-abl expressionand protein kinase activity are similar to those of establishedtumors (8, 9). Only after a period of weeks in culture dothese A-MuLV-infected cell lines acquire the capacity forautonomous growth (9). (ii) Clonal lines of A-MuLV-inducedtumors can lose the A-MuLV proviral genome in vivo whilestill retaining the fully transformed phenotype (10). Some ofthese tumor lines show increased transcription of two cellu-lar oncogenes, c-myb and c-myc (11, 12). (iii) The cellulartransforming sequences from A-MuLV-induced tumorsidentified by transfection into NIH-3T3 cells are distinctfrom the A-MuLV gene product (13). Taken together, theseresults suggest that A-MuLV expression may be the initiat-ing event in a multistep process involving the activation orinduction of other transformation-related proteins of cellularorigin.

In this paper we report the isolation and characterization

of a monoclonal antibody (6C3) that detects transformation-associated proteins in murine B-cell lymphomas that are dis-tinct from the A-MuLV gene product (v-abl) and its normalcellular counterpart (c-abl). The expression of these proteinscorrelates closely with growth independence and the trans-formed phenotype. The 6C3 antibody-detected moleculesmay represent transformation-associated proteins of cellularorigin whose expression is induced, stabilized, or amplifiedby A-MuLV.

MATERIALS AND METHODSMice. C57L mice were obtained from The Jackson Labo-

ratory and were maintained in our animal facilities.Cells. Summaries of all cell lines used in this study are

presented in Tables 1 and 2. All in vitro tumor cell lines weremaintained in culture in our laboratory in RPMI-1640 medi-um (GIBCO) supplemented with 10% fetal calf serum. TheA-MuLV-induced pre-B lymphomas (L1-2, RAW112, 2M3,220-2, 223-18, and 230-23-8) and the A-MuLV-ihduced fibro-sarcoma (ANN-1) were obtained from N. Rosenberg (TuftsUniversity). B-cell lymphomas included BAL-17 (P. Jones,Stanford University), 38C13 (R. Levy, Stanford University),NBL (D. Scott, Duke University), T69 (J. Monaco, StanfordUniversity), CH-1 (G. Haughton, University of North Caro-lina), and BCL-1 (S. Strober, Stanford University). T-celllymphomas included KKT-2 (a spontaneous AKR lympho-ma isolated in our laboratory), UNC-1 (G. Haughton, Uni-versity of North Carolina), MBL-2 (W. Green, FrederickHutchinson Cancer Research Center, Seattle), L691 (H.Kaplan, Stanford University), and S49 (K. Melmon, Stan-ford University). The 8653 non-producer myeloma cell linewas provided by R. Levy (Stanford University). Normalspleen, thymus, and bone marrow were obtained from C57Lanimals and cell suspensions were prepared in cell suspend-ing medium with 5% fetal calf serum. Long-term bone mar-row cultures were established by using the method previous-ly described (14).

Antibodies. The 6C3 antibody was prepared by a modifica-tion of the technique of Galfre et al. (15). Briefly, Fisher ratswere immunized with 2 x 107 cloned cytotoxic T cells and 2X 107 L1-2 stimulator cells intraperitoneally in completeFreund's adjuvant. Fourteen days later, the animals re-ceived a second injection of cells in incomplete Freund's ad-juvant. Three days after this booster injection, spleen cells (2x 108) were fused with 8653 myeloma cells (2 x 107) by using50% polyethylene glycol 1540. Hybrids were grown initiallyin hypoxanthine/aminopterin/thymidine medium (16) in 96-well microtiter plates. Supernatants were tested for bindingto tumor cells by radioimmunoassay. Positive cultures weresubcloned by limiting dilution and grown to larger volume inRPMI-1640 medium with 10% fetal calf serum for furthertesting. Goat antisera specific for whole Moloney MuLV (M-

Abbreviations: A-MuLV, Abelson murine leukemia virus; M-MuLV, Moloney MuLV.

4434

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

July

3, 2

020

Proc. Natl. Acad Sci. USA 81 (1984) 4435

MuLV; lot 1S-166) and Rauscher MuLV p12 (lot 77S-130)were obtained from J. Gruber (National Cancer Institute,National Institutes of Health).

Cell Binding Assays. Cell binding radioimmunoassay wasperformed in 96-well flexible microtiter plates (Dynatech,Alexandria, VA) by using g 125I-labeled rabbit anti-rat Ig sec-ond stage (25,000-50,000 cpm/0.025 ml) as described (17).For immunofluorescence, cells were stained with the 6C3antibody by using as a second stage a fluorescein isothiocya-nate-labeled rabbit antiserum specific for rat Ig (17) and ana-lyzed by fluorescence microscopy or fluorescence-activatedcell sorting.

Cell Labeling. L1-2 cells were metabolically labeled at aconcentration of 107 per ml with [35S]methionine (150 ,uCi/ml; 1 Ci = 37 GBq) or [32P~orthophosphate (250 ,Ci/ml) inmethionine-free or low-phosphate minimal essential medium(Eagle) supplemented with 5% dialyzed fetal calf serum for 2hr at 37°C. Lactoperoxidase-catalyzed iodination was per-formed on washed L1-2 cells at a concentration of 2 x 107per ml in phosphate-buffered saline (18) by using 1251 (500,uCi/ml).

Immunoprecipitation. Labeled cells were lysed into phos-phate lysis buffer (6) supplemented with 5 mM EDTA/5 mMphenylmethylsulfonyl fluoride at a concentration of 2 x 106cells per ml. Cell extracts were clarified by centrifugation for2 hr at 27,000 rpm in an SW 50.1 rotor. One milliliter of ly-sate was allowed to react with either 0.05 ml of monoclonalantibody supernatant or 0.005 ml of goat antiserum overnightat 4°C. Affinity-purified rabbit anti-rat Ig serum in phos-phate-buffered saline (50 ,ug) was added to those samplescontaining rat monoclonal antibodies. Immune complexeswere precipitated with 0.05 ml of a 10% suspension of Staph-ylococcus aureus, washed three times in phosphate lysisbuffer, and eluted into NaDodSO4 sample buffer with orwithout added 2-mercaptoethanol by boiling for 3 min. Forpreclearing A-MuLV proteins, 1 ml of cell lysate was precip-itated three times with 0.005 ml of goat anti-p12 serum and0.05 ml of a 10% solution of S. aureus. Immunoprecipitation

was then performed on the precleared lysates in the usualmanner.

Protein Kinase Assay. Unlabeled L1-2 cells were immuno-precipitated with 6C3 or goat anti-M-MuLV serum in theusual manner. The precipitates were resuspended in 0.05 mlof 50 mM Tris (pH 8.0) with 10 mM MnCl2 and incubatedwith 4 ,uCi of [y-32P]ATP for 5 min at 30°C. The precipitateswere washed once in phosphate lysis buffer and then elutedinto NaDodSO4 sample buffer with 2-mercaptoethanol byboiling for 3 min.NaDodSO4/PAGE. Labeled samples (0.025-0.05 ml) were

run on 8% NaDodSO4/PAGE slab gels for 2 hr at 150 V.Gels were fixed in methanol/acetic acid/water, 50:10:40, for1 hr and then rinsed in distilled water for 1 hr. Gels weretreated with 1 M sodium salicylate for an additional hour,dried, and then exposed to Kodak XAR-5 film for 1-3 days.

RESULTSIsolation of a Monoclonal Antibody Specific for B-Cell Lym-

phomas. As part of a project to isolate T-killer cell-specificantibodies, Fisher rats were immunized with a mixture of A-MuLV tumor-specific T-cell clones (1E4) and their A-MuLVpre-B-cell target (L1-2) in complete Freund's adjuvant. Dur-ing the screening of the resulting hybridomas, one antibodyof the IgG2A subclass was identified (6C3) that showed highlevels of binding to the L1-2 target rather than to the T-killercell clone. This hybridoma was subsequently cloned at limit-ing dilution and grown to larger volume for further testingagainst a panel of tumor cells and normal tissues.The 6C3 antibody binds to all A-MuLV pre-B lymphomas

but does not react with any normal cell population (Table 1).Although the antigen recognized by 6C3 could be a markerfor some rare pre-B cell subset that is only a minor compo-nent of normal lymphoid populations, no 6C3-positive cellswere found out of 10,000 viable cells examined in normalbone marrow by using an indirect immunofluorescence as-say. The 6C3 antigen is present on two of six B-cell lympho-mas of nonviral origin (Table 1) but is not found on any non-

Table 1. 6C3 antibody binding to tumor cell lines and normal lymphoid tissues

cpm bound per

Strain of Mode of 2.5 x 105 cellsCell type Cell line origin induction 6C3 Control

Pre-B lymphoma L1-2 C57L A-MuLV 4801 78RAW112 BALB/c A-MuLV 1377 772M3 BALB/c A-MuLV 1247 80220-2 BALB/c A-MuLV 1191 77223-18 C57L A-MuLV 3412 89230-23-8 C57L A-MuLV 2918 93

B lymphoma NBL NIH/Sw Spontaneous 972 7038C13 C3H Carcinogen 4279 198BAL-17 BALB/c Carcinogen 108 68T69 C57BL/10 Hyperimmune 73 59CH-1 C57BL/10 Hyperimmune 95 81BCL-1 BALB/c Spontaneous 90 70

T lymphoma MBL-2 C57BL/6 M-MuLV 84 71UNC-1 C57BL/10 Hyperimmune 119 102S49 BALB/c Mineral oil 108 111KKT-2 AKR Spontaneous 74 73L691 C57L Radiation 139 113

Fibrosarcoma ANN-1 BALB/c A-MuLV 135 131Mastocytoma P815 DBA/2 Carcinogen 86 81Normal thymus C57L 153 79Normal spleen - C57L 125 94Normal bone marrow C57L 135 75

6C3 expression was determined by radioimmunoassay using 6C3 antibody as a first stage and 12511labeled rabbit anti-rat Ig (absorbed to remove any anti-mouse Ig activity) as a second stage. Back-ground binding was determined using a first stage antibody of irrelevant specificity.

Cell Biology: Pillemer et aL

Dow

nloa

ded

by g

uest

on

July

3, 2

020

4436 Cell Biology: Pillemer et al.

B cell tumors or on lipopolysaccharide-activated normallymphocytes (data not shown). This suggests that the 6C3antigen is not just a marker for proliferating or activatedlymphoid cells but is specific for tumors of the B-cell lin-eage.6C3 Antibody Immunoprecipitates Two Proteins from A-

MuLV-Induced Tumor Cells. 6C3 antibody immunopre-cipitates two proteins of Mr 160,000 (p1606CS) and Mr 125,000(p1256C3) from metabolically labeled L1-2 cells (Fig. 1).These same two, proteins are immunoprecipitated from all6C3 antigen-positive cell lines tested, including the non-A-MuLV-induced B-lymphoma NBL; neither protein is im-munoprecipitated from any 6C3 antigen-negative cell line(data not shown). When L1-2 cells are surface iodinated byusing the lactoperoxidase method, only p1606C3 is precipitat-ed (Fig. 1), indicating that p1606C3 is the serologically de-fined cell surface antigen detected in the binding assays.When the same immunoprecipitations were run under nonre-ducing conditions to preserve disulfide linkages, the 6C3proteins were found to occur predominantly as dimers of Mr320,000 and Mr 250,000 (Fig. 1). It is interesting to note thatthe A-MuLV gene product pl2OabI also exists as a multimericform with L1-2 cells. When L1-2 cells are labeled in the pres-ence of tunicamycin, an inhibitor of glycosylation, a singleprotein of Mr 105,000-110,000 is immunoprecipitated with6C3 (unpublished data). This indicates that the 6C3 antibodydetects a protein determinant on the common precursor ofthe p1256C3 and pl606C3 glycoproteins.6C3 Proteins Are Distinct from the A-MuLV Gene Product

and Its Normal Cellular Counterpart. Although 6C3 proteinsare not precipitated by a broadly reactive anti-M-MuLV se-rum (Fig. 1), the molecules recognized by the 6C3 antibodyare similar in size to the products of the AKMuLV genome:p120abl and p160abl (19). Thus, it was important to establishthe relationship, if any, between the 6C3 proteins and the A-MuLV-transforming gene products. As shown in Fig. 2, pre-clearing of a L1-2 cell extract with anti-gag antibodies re-moves all of the pl20abl from the cell lysate, consistent withthe presence of a portion of the M-MuLV gag at the NH2terminus of pl20abl (6, 7). However, the same method of pre-clearing fails to remove any of the p12563 or p1606C3 fromthe same extract, indicating that these proteins do not con-tain viral gag determinants. In addition, Fig. 3 demonstratesthat in contrast to p120abl, pl256C3 and pl606C3 are not la-

a b c a b c a b c a b c

w IW4op1606C3 * 6

gp7O

-p160eC3pl2- pl56C3

p1 2 abi

a b b a b a

P1606C3pl 256C3

p120abl

Pr 80env

a_

~.-E

a

I j1l ml

FIG. 2. Effect of preclearing with anti-p12 serum on immuno-precipitation of [35S]methionine-labeled proteins from L1-2 cells.L1-2 cells were metabolically labeled with [35S]methionine and celllysates were prepared and precleared with either normal goat serum(lanes a) or goat anti-Rauscher p12 serum (lanes b). Immunoprecip-itation was then performed with goat anti-M-MuLV serum (I), 6C3monoclonal antibody supernatant (II), or medium control (III). Im-munoprecipitated proteins were separated on an 8% NaDodSO4 slabgel and detected by autoradiography.

beled with inorganic phosphate and do not possess in vitroprotein kinase activity. Thus, we conclude that the 6C3 pro-teins represent B-lymphoma-specific proteins distinct fromA-MuLV gene products.Although the 6C3 proteins do not correspond to any

known A-MuLV-encoded protein, they could still be relatedto the normal cellular counterpart of the A-MuLV-trans-forming protein c-abl. The product of the c-abl gene is a pro-tein of Mr 150,000 (NCP150) (20), which is expressed at highlevels within normal lymphoid tissues (21). However, the

a b c a b c a b

po6C3p160

6C

",

p1256C30pl20 abl--

Pr 80 envPr65gag

r

-It

-Pr 80env-Pr 65gag

c

I ]I I R

FIG. 1. Immunoprecipitation of 125I- and [35S]methionine-la-beled proteins from L1-2 cells. L1-2 cells were surface labeled with125I (left-hand lanes) or metabolically labeled with [35S]methionine(right-hand lanes). Cell lysates were prepared and immunoprecipi-tated with goat anti-M-MuLV serum (lanes a), 6C3 monoclonal anti-body supernatant (lanes b), or medium control (lanes c). Immuno-precipitated proteins were separated on an 8% NaDodSO4 slab gelunder reducing conditions in the presence of 2-mercaptoethanol (I)or non-reducing conditions in the absence of 2-mercaptoethanol (II)and detected by autoradiography.

I Ii

FIG. 3. Immunoprecipitation of phosphorylated proteins fromL1-2 cells. L1-2 cells were metabolically labeled with [35S]methio-nine (I) or [32P]orthophosphate (II). Cell lysates were prepared andimmunoprecipitated with goat anti-M-MuLV serum (lanes a), 6C3monoclonal antibody supernatant (lanes b), or medium control(lanes c). For the protein kinase assay, unlabeled L1-2 cells werefirst immunoprecipitated with the same antibodies and then labeledin vitro with [y-32P]ATP (III). Immunoprecipitated proteins wereseparated on an 8% NaDodSO4 slab gel and detected by autoradiog-raphy.

Proc. NatL Acad Sci. USA 81 (1984)

Dow

nloa

ded

by g

uest

on

July

3, 2

020

Proc. NatL. Acad. Sci. USA 81 (1984) 4437

a b c a b c a b c a b c

p1606C3p1256C3P120ab

Pr 80 enV

Pr 65gag

*1I.

L 1-2 Thymus Spleen BoneMarrow

FIG. 4. Immunoprecipitation of [35S]methionine-labeled proteinsfrom normal lymphoid cells. L1-2 cells (2 x 106) or normal lymphoidcells (2 x 107) were labeled metabolically with [35S]methionine. Celllysates were prepared and immunoprecipitated with goat anti-M-MuLV serum (lanes a), 6C3 monoclonal antibody supernatant (lanesb), or medium control (lanes c). Immunoprecipitated proteins were

separated on an 8% NaDodSO4 slab gel and detected by autoradiog-raphy.

6C3 proteins are not detected in thymus or any other normallymphoid tissue either by radioimmunoassay (Table 1) or byimmunoprecipitation, even when using 10-fold excess of nor-mal cells (Fig. 4). The absence of the 6C3 proteins from nor-mal lymphoid tissues argues strongly against any relation-ship with NCP150.6C3 Expression Correlates with a Transformed Phenotype.

Recently, it has become possible to establish long-term cul-tures of normal pre-B and B lymphocytes on adherent bonemarrow feeder layers (14). These lymphocytes can be cul-tured for many months and still retain a strong requirementfor the feeder layer for in vitro growth. Although such cellsare not tumorigenic in vivo, B lineage cells capable of in vivotumor production do arise in a fraction of the cultures aftermany months of passage and these cells no longer require thefeeder layer for in vitro growth (unpublished data). A similar(but more rapid) progression toward autonomous in vitro

growth and high oncogenic potential is observed for bonemarrow cells transformed in vitro with A-MuLV (8, 9).To define the stages in malignant progression when B lin-

eage cells gain expression of 6C3 antigens, a panel of bothnon-virus-infected and A-MuLV-infected pre-B and B-cellpopulations was screened for expression of the 6C3 antigenby immunofluorescence (Table 2). Both the uncloned (C185)and cloned (NBM clone no. 3) non-A-MuLV-infected pre-Band B-cell populations that required feeder layers for growthfailed to express 6C3. Proliferating cells in fresh bone mar-

row cell culture 10 days after infection with p160 A-MuLV(P160BM-EP) still required the feeder layer for growth.These cultures contained a mixture of6C- (91%) and 6C3+(9%) cells. Proliferating cells from cultures 4-6 wk after in-fection with p160 A-MuLV (data not shown) and establishedp160-transformed pre-B-cell lines (C182-E5 and P160-NBMclone no. 3) both showed expression of high levels of 6C3antigen on all cells. Whether the high percentage of 6C3-pos-itive cells in the older cultures resulted from dominance ofthe cultures by the few positive cells present early after in-fection or from conversion of negative cells to positive cellsremains to be determined. Interestingly, a highly oncogenicB-cell line (NBM 10/27) that arose spontaneously in one

bone marrow culture in the absence of virus infection ex-

pressed the 6C3 antigen at a level equivalent to the estab-lished A-MuLV-transformed pre-B-cell line L1-2. Thus, it isevident from these data that 6C3 expression is most closelycorrelated with growth independence from the supportingfeeder layer rather than expression of or infection with eitherp120 or p160 strains of A-MuLV.

DISCUSSIONWe have used a monoclonal antibody (6C3) to isolate trans-formation-associated proteins of cellular origin from B-celllymphomas, and we have demonstrated that these proteinsare distinct from the A-MuLV gene product and its cellularcounterpart. Using a model bone marrow culture system, wehave shown that early after infection with A-MuLV, only a

small subset of cells is 6C3+. Expression of 6C3 proteinsparallels the progression to the fully transformed phenotypein vitro.A number of other proteins have been described that show

increased expression in A-MuLV-transformed cells. An A-MuLV lymphoma antigen has been detected serologically onall A-MuLV-transformed cells that is also a differentiationantigen present on normal bone marrow, spleen, and fetalliver (22). A-MuLV-transformed cells express increasedconcentrations of a phosphoprotein of cellular origin (p53)(23) whose expression is increased in tumors from a variety

Table 2. 6C3 expression in cultured bone marrow cells

6C3 expressionFeeder layer % fluorescein Median

Source of cell line Cell line designation Description of cell line dependence positive fluorescence

Bone marrow culturesNot virus infected C185 Uncloned B cell + 0 0

NBM clone no. 3 Cloned B cell + 0 1NBM 10/27 Spontaneously transformed B cell - 100 100

Virus infected P160 BM-EP Fresh bone marrow transformants early + 9 1after P160 infection

C182-E5 P160-transformed pre-B cell - 100 74P160-NBM clone no. 3 P160-transformed NBM clone no. 3 - 100 57

A-MuLV transformed L1-2 P120-transformed (long-term) pre-B cell - 100 92

6C3 expression was monitored by indirect immunofluorescent staining using 6C3 antibody as a first stage and a fluorescein-conjugated rabbitanti-rat Ig (absorbed to remove any anti-mouse Ig activity) as a second stage. The percent of stained cells was determined by examining 1000cells by fluorescence microscopy. Median fluorescence above background staining was determined by a fluorescence-activated cell sorterstandardized to a fluorescence gain of 1. Background staining for each cell line was determined using as a first stage control a non-bindingantibody of irrelevant specificity.

Cell Biology: Pillemer et aL

Dow

nloa

ded

by g

uest

on

July

3, 2

020

4438 Cell Biology: Pillemer et al.

of sources (24). A Mr 95,000 protein has been isolated that isfound only in L1-2 cells but is not present in other A-MuLV-induced lymphomas or fibroblasts (25). Thus, of the antigenspreviously described for A-MuLV-induced B-cell tumors,the 6C3 proteins are unique in that they are found only onautonomously growing cells of the B-lymphocyte lineage andare not detected on any other tumor or normal cell popula-tion. Recently, a Mr 80,000 phosphoprotein (p80) has beendetected in a variety of murine T-cell lymphomas, includinga number ofA-MuLV-induced tumors that express the T-celldifferentiation antigen Thy-1 (26). This protein is not foundin any normal lymphoid tissues. It is interesting to considerthe possibility that p80 is the murine T-lymphoma equivalentof the B-cell lymphoma-specific 6C3 proteins.Our studies with the bone marrow cultures raise the possi-

bility that the 6C3 proteins represent the product of a cellularoncogene activated by A-MuLV infection. Recently, it hasbeen demonstrated that some A-MuLV-induced tumors thathave lost A-MuLV expression have markedly increasedtranscription of RNA from the cellular homologues of twoviral oncogenes, c-myc and c-myb (11, 12). This situationmay be similar to that of the avian leukosis viruses, in whichit has been shown that the transforming retrovirus integratesnear to and activates the expression of a cellular oncogene(c-myc) but that the virus itself need not maintain continuedexpression in transformed cells (27). Further investigation isneeded to establish the relationship between the 6C3 pro-teins and the products of c-myc and other cellular onco-genes.Although the precise function of the 6C3 proteins in lym-

phoma cells awaits more detailed characterization, the ap-pearance of these proteins in autonomously growing cells inthe absence of added A-MuLV suggests that they may havemore general significance in relation to the control of neo-plastic B-cell proliferation. Since feeder layer dependence invitro seems to be related to the production of soluble factorsproduced by the adherent cell layer (8), it is conceivable thatthe 6C3 proteins on independently growing B lineage cellsmay represent novel B-cell growth factors or receptors forgrowth factors or other molecules. In this regard it is inter-esting to note that A-MuLV-transformed fibroblasts producea tumor growth factor that competes with epidermal growthfactor for binding (28). In any event, further characterizationof the structure and function of the 6C3 proteins should pro-vide new insights into the process of B-lymphocyte transfor-mation.

We thank 0. N. Witte and G. Tidmarsh for helpful discussions.This work was supported by Grant IM-56 from the American CancerSociety and Grant CA-32031 from the National Cancer Institute.

E.P. is a Medical Scientist Trainee under Grant GM-07365 from theU.S. Public Health Service.

1. Sklar, M. D., White, B. J. & Rowe, W. P. (1974) Proc. Natil.Acad. Sci. USA 71, 4077-4081.2. Raschke, W. C., Ralph, P., Watson, J., Sklar, M. & Coon, H.

(1975) J. Natil. Cancer Inst. 54, 1249-1253.3. Abelson, H. T. & Rabstein, L. S. (1970) Cancer 30, 2213-

2222.4. Witte, 0. N., Dasgupta, A. & Baltimore, D. (1980) Nature

(London) 283, 826-831.5. van de Ven, W. J. M., Reynolds, F. H., Jr., & Stephenson,

J. R. (1980) Virology 101, 185-197.6. Witte, 0. N., Rosenberg, N., Paskind, M., Shields, A. & Balti-

more, D. (1978) Proc. Natil. Acad. Sci. USA 75, 2488-2492.7. Reynolds, F. H., Jr., Sacks, T. L., Deobaghar, D. N. & Ste-

phenson, J. R. (1978) Proc. Natil. Acad. Sci. USA 75, 3974-3978.8. Whitlock, C. A. & Witte, 0. N. (1981) J. Virol. 40, 577-584.9. Whitlock, C. A., Ziegler, S. F. & Witte, 0. N. (1983) Mol.

Cell. Biol. 3, 596-604.10. Grundwald, D. J., Dale, B., Dudley, J., Lamph, W., Sugden,

B., Ozanne, B. & Risser, R. (1982) J. Virol. 43, 92-103.11. Mushinski, J. F., Bauer, S. R., Potter, M. & Reddy, E. P.

(1983) Proc. NatIl. Acad. Sci. USA 80, 1073-1077.12. Mushinski, J. F., Potter, M., Bauer, S. R. & Reddy, E. P.

(1983) Science 220, 795-798.13. Lane, M. A., Neary, D. & Cooper, G. M. (1982) Nature (Lon-

don) 300, 659-661.14. Whitlock, C. A. & Witte, 0. N. (1982) Proc. Natil. Acad. Sci.

USA 79, 3608-3612.15. Galfre, G., Howe, S. C., Milstein, C., Butcher, G. W. & How-

ard, J. C. (1977) Nature (London) 266, 550-552.16. Littlefield, J. W. (1964) Science 145, 709-710.17. Pillemer, E. & Weissman, I. L. (1981) J. Exp. Med. 153, 1068-

1079.18. Witte, 0. N. & Weissman, I. L. (1976) Virology 69, 464-473.19. Rosenberg, N. & Witte, 0. N. (1980) J. Virol. 33, 340-348.20. Witte, 0. N., Rosenberg, N. & Baltimore, D. (1979) Nature

(London) 281, 396-398.21. Wang, J. Y. & Baltimore, D. (1983) Mol. Cell. Biol. 3, 773-

779.22. Risser, R., Stockert, E. & Old, L. J. (1978) Proc. NatIl. Acad.

Sci. USA 75, 3918-3922.23. Rotter, V., Witte, 0. N., Coffman, R. & Baltimore, D. (1980)

J. Virol. 36, 547-555.24. DeLeo, A. B., Jay, G., Appella, E., Dubois, G. C., Law,

L. W. & Old, L. J. (1979) Proc. NatIl. Acad. Sci. USA 76,2420-2424.25. Machida, C. A. & Kabat, D. (1983) Cancer Res. 43, 1275-

1281.26. Grundwald, D. J., Miller, K. & Risser, R. (1982) J. Exp. Med.

156, 1195-1209.27. Payne, G. S., Bishop, J. M. & Varmus, H. E. (1982) Nature

(London) 295, 209-214.28. Twardzik, D. R., Todaro, G. J., Marguardt, H., Reynolds,F. H., Jr., & Stephenson, J. R. (1982) Science 216, 894-8%.

Proc. NatL Acad ScL USA 81 (1984)

Dow

nloa

ded

by g

uest

on

July

3, 2

020