immunological phenotype of lymphomas induced by avian

MOLECULAR AND CELLULAR BIOLOGY, June 1983, p. 1077-10850270-7306/83/061077-O9S02.00/0Copyright C) 1983, American Society for Microbiology

Vol. 3, No. 6

Immunological Phenotype of Lymphomas Induced by AvianLeukosis Viruses

LING-CHUN CHEN,* SARA A. COURTNEIDGE,t AND J. MICHAEL BISHOPDepartment ofMicrobiology and Immunology, University of California, San Francisco, California 94143

Received 28 January 1983/Accepted 29 March 1983

The production of immunoglobulin by six cell lines derived from bursal tumorsinduced by avian leukosis virus follows two general patterns: (i) three cell linesthat have been extensively passaged in culture synthesize and secrete light chainsonly; (ii) three cell lines that are recently isolated produce and secrete monomericimmunoglobulin M in addition to free light chains. All six cell lines synthesize andsecrete both glycosylated and unglycosylated forms of light chain. We concludethat the cell lines established from lymphomas induced by avian leukosis virusrepresent relatively mature, but possibly abnormal, stages in the development ofchicken B-lymphocytes. The immunoglobulin M produced by the cell lines failedto form detectable immune complexes with avian leukosis virus. It thereforeappears that the imunoglobulin M is not directed against viral antigens and thatautogenous antigenic stimulus cannot account for the sustained growth of theneoplastic B-lymphocytes.

Avian leukosis virus (ALV) induces lymphoidleukosis in newly hatched chickens of suscepti-ble strains after a long period of latency. Theearliest neoplastic changes are observed in thebursa 4 to 8 weeks after infection, but deathfrom disseminated lymphomas occurs only 5 to 9months later (31). Since ALV-induced lymphoidleukosis can be prevented by bursectomy, thehematopoietic target cells for ALV are thoughtto reside in the bursa of Fabricius (30). Althoughprevious work demonstrated the production ofimmunoglobulin M (IgM) by the tumors (6),suggesting that the lymphoma is a malignancy ofB-lymphocytes, the precise immunological phe-notype of the tumor cells has not been reported.

It has been postulated that hemopoietic tu-mors induced by leukemia viruses are cellsarrested at a particular stage during their differ-entiation. In this view, the tumor cells continueto express "normal" differentiation markers (2).However, it is also possible that tumorigenesisperturbs the differentiation program and that thetumor cells no longer represent a normal stage indevelopment (7, 8). To ascertain whether leuke-mogenesis perturbs differentiation or merelyleads to differentiation arrest, it is necessary todetermine how precisely the tumor cells repre-sent their normal counterparts. If tumors in-duced by leukemia viruses are caricatures ofnormal cells, study of the tumor cells may lead

t Present address: Division of Biochemistry, National Insti-tute for Medical Research, Mill Hill, London, NW7 1AA,United Kingdom.

to the discovery of previously unknown stagesin development or may otherwise enhance ourknowledge of the differentiation pathway inquestion.The process of immunoglobulin gene expres-

sion during B-cell development in mice can bedivided into four stages (14): (i) ,t heavy chain,but not light chain, is expressed in the cyto-plasm; (ii) light chain is synthesized and theassembled IgM monomer is displayed on the cellsurface; (iii) IgM is secreted in the form ofpentamer, and (iv) other classes of immunoglob-ulin are synthesized and secreted. Assumingthat a similar sequence of events occurs duringavian B-lymphocyte development, we analyzedthe expression of immunoglobulin genes in celllines derived from lymphomas induced by ALV.Our results indicate that the avian lymphomasrepresent a relatively advanced stage in B-celldevelopment and that their immunological phe-notypes are not entirely normal.

Presently there are two models proposed toexplain leukemogenesis by ALV. According tothe first model, neoplastic transformation byALV results from the activation of a cellularoncogene(s). This conclusion is based mainly ontwo findings: (i) most of the lymphomas displayenhanced expression of a previously recognizedcellular oncogene (c-myc), apparently conse-quent to the integration of ALV DNA (12, 28);and (ii) the DNA of lymphomas contains anactive transforming gene that can be detected bytransfection into NIH 3T3 mouse cells and is notrelated to c-myc (5). The second model suggests

1077 Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 14

Feb

ruar

y 20

22 b

y 12

0.71

.5.1

18.

1078 CHEN, COURTNEIDGE, AND BISHOP

that mitogenic stimulation of lymphocytes leadsto neoplastic growth of the cells. Lymphoidtumors induced by murine leukemia virus(MuLV) are hypothesized to represent clonescarrying receptors for viral antigens. Continuingmitogenic stimulation of the cells by antigencould account for (or at least contribute to)neoplastic growth (16, 20). Since at least some

B-cells derived from ALV-induced lymphomasproduce virus and IgM, both antigen and poten-tial receptor are available for analysis. We testedthe role of antigen-induced mitogenesis by ask-ing whether the tumor cells produce IgM direct-ed against the virus. Our results suggest that thismay not be the case. Furthermore, since some ofthe cell lines we have studied do not makecomplete antibody, we conclude that mitogenicstimulation mediated by antigen binding to cell-surface antibody is not required for maintainingthe neoplastic growth of B-cells derived fromALV-induced lymphomas.

MATERIALS AND METHODS

Tumor cell lines. Table 1 summarizes the origins andpassage histories of the six tumor cell lines underinvestigation. The isolation and maintenance of thesecell lines have been described previously (13, 25, 29).

Biosynthetic labeling and immunoprecipitation. Inmost cases, 2 x 106 cells were washed with phosphate-buffered saline (0.01 M phosphate and 0.15 M NaCl,pH 7.2) and suspended in 1 ml of Dulbecco modifiedEagle medium deficient in methionine or leucine andsupplemented with [35S]methionine or [3H]leucine,respectively. Carbohydrate labeling was accomplishedby incubating cells in Dulbecco modified Eagle medi-um containing 1 ,ug of glucose per ml for 4 h before theaddition of [3H]mannose. Unglycosylated immuno-globulin was also labeled with [35S]methionine after a

4-h treatment with 1 ,ug of tunicamycin per ml (Calbio-chem). The viability of cells was determined to begreater than 90% by trypan blue exclusion at the end ofthe labeling period.

After the labeling period, cell lysates were preparedand clarified as previously described (26). The labeledproteins were precipitated with antisera and Formalin-fixed Staphylococcus aureus according to the proce-

dures described by Oppermann et al. (26). Secretedproteins were recovered directly from culture superna-

tant of labeled cells by immunoprecipitation. Antiseraused for immunoprecipitation of the labeled proteinsincluded normal rabbit serum, rabbit anti-chicken totalimmunoglobulin serum (GIBCO Laboratories), goatanti-chicken light chains serum (Miles Laboratories),goat anti-chicken IgM (,. chain specific) serum (MilesLaboratories, Cappel Laboratories, Pel-Freez Bio-chemicals), and rabbit anti-Rous sarcoma virus virionproteins (17, 27).

Immunoprecipitated proteins were resolved by so-

dium dodecyl sulfate (SDS)-polyacrylamide gel elec-trophoresis according to Laemmli (15). The labeledproteins were detected by autoradiography or fluorog-raphy (3).

Protease mapping. Partial proteolysis with S. aureus

V8 protease in the presence of SDS was done by themethod of Cleveland et al. (4) as modified by Opper-mann et al. (26).Sepharose 4B gel filtration. Viral particles and free

IgM molecules from culture supernatants were sepa-

rated on a Sepharose 4B column (30 by 1.5 cm) underconditions to allow recovery of infectious retroviruses(19). Cells were labeled with [35S]methionine for 6 h,and the culture supernatant was clarified and loadedon the column. Fractions of 1 ml were collected andadjusted to 1% Nonidet P-40. Alternate fractions wereimmunoprecipitated with antiserum directed againstchicken immunoglobulin or antiserum directed againstviral proteins. The immune complexes were analyzedby SDS-polyacrylamide gel electrophoresis as de-scribed above.

RESULTS

Biosynthesis of immunoglobulins in ALV-in-duced lymphoma cells. The production ofimmunoglobulin was analyzed by immunopre-cipitation with specific antisera. The pattern ofimmunoglobulin production varied among differ-ent cell lines, but took two general forms. Figure1 shows the immunoglobulin synthesized by the1104X-5 cell line, representative of one generalpattern. When antiserum against all classes ofchicken immunoglobulin was used, three poly-peptides with molecular weights of 30,000,28,000, and 27,000 were specifically precipitated

TABLE 1. Passage histories of cell lines derived from ALV-induced lymphomasChicken Subgroup Date of Source of tumor Immunoglobulin

Cell line line of ALV infection cells Passage history production

1104B-1 15 A 1974 Bursa Established by Hihara et al. Light chains onlywhen lymphoma developed

1104X-5 15 A 1974 Bursa Same as 1104B-1 Light chains onlyBK4484A BK A 1974 Bursa Same as 1104B-1 Light chains onlyR2B 15 x 7 B 1979 Bursa Established by Courtneidge in IgM

1980 from tumor passagedpectorally by Okazaki

SC-2B SC A 1981 Bursa Established by Courtneidge in IgM1981 from tumor passagedpectorally by Wright

SC-2L SC A 1981 Liver metastasis Same as SC-2B IgM

MOL. CELL. BIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 14

Feb

ruar

y 20

22 b

y 12

0.71

.5.1

18.

IMMUNOLOGICAL PHENOTYPE OF ALV LYMPHOMAS 1079

Intra-cellular

Extra -

cellular

a b c d e f

30K_28K--27K-

"------ 30.5K- --27K

FIG. 1. Electrophoretic analysis of immunoglob-ulins synthesized by 1104X-5 cells. After the cellswere labeled for 4 h with [35S]methionine, both cyto-plasmic extract and culture media were immunopre-cipitated with normal rabbit serum (lanes a and d),anti-chicken immunoglobulin serum (lanes b and e), or

anti-chicken L-chain serum (lanes c and f). The precip-itated immunoglobulins were resolved in their reducedforms by SDS-polyacrylamide gel electrophoresis andautoradiography.

from the cell lysate (Fig. 1, lane b). By contrast,only two polypeptides with molecular weights of30,500 and 27,000 were specifically precipitatedfrom the culture medium (Fig. 1, lane e). Whenantiserum directed against chicken light chainswas used, only the proteins with molecularweights of 30,500, 30,000, and 27,000 were re-covered (Fig. 1, lanes c and f). Since these threeproteins could be specifically precipitated byseveral antisera directed against chicken lightchains, we have designated them as light chains.

The intracellular protein with a molecularweight of 28,000 was precipitated by only one lotof antiserum, directed against total immunoglob-ulin. We therefore suspect that the 28,000 pro-tein is not immunoglobulin but a protein precip-itated by contaminating antibody in only onebatch of serum. We conclude that the 1104X-5cell line synthesizes and secretes two forms oflight chains without any detectable heavy chain.Two other cell lines, 1104B-1 and BK4484A,have the same pattern of immunoglobulin pro-duction, although the unidentified 28,000-molec-ular-weight protein was not detected with anyantiserum (Table 2).

Figure 2 illustrates the second general patternof immunoglobulin production, as obtained withthe cell line SC-2L. When the immune complex-es precipitated by anti-chicken immunoglobulinserum were analyzed without reduction in a 6 to9% SDS-polyacrylamide gel, a molecule whichmigrated with an apparent molecular weight of210,000 was recovered from both the cellularextract and culture medium (Fig. 2, lanes a andd). The intensity of the bands seen in the middleof lane a varied in different immunoprecipita-tions; we have not identified these proteins. Thelower two bands in lanes a and d are free lightchains which will be discussed below.When the samples were analyzed under re-

ducing conditions, three classes of polypeptideswith molecular weights of 77,000, 32,000 to32,500, and 29,000 were recovered (Fig. 2, lanesb and e). It therefore appears that the 210,000 Mrprotein precipitated without denaturation repre-sents immunoglobulin composed of one dimer ofheavy chain (Mr 77,000) and one dimer of lightchain (Mr 32,000 to 32,500 or Mr 29,000). Anti-serum directed against chicken R heavy chainprecipitated the same set of proteins (Fig. 2,lanes c and f). We therefore conclude that thecells produce and secrete IgM. We recognizethat the antisera used in this identification maynot be monospecific for the IgM isotype, but ourconclusion is nevertheless in accord with previ-ous reports (6, 25).The pattern of light chains produced in cell

TABLE 2. Biosynthesis of immunoglobulin by ALV-transformed cellsIntracellular" (mol wt) Secretiona (mol wt)

Type CelllineHLHLH L H L

I BK4484A 32K, 29K 32.5K, 29K1104B-1 32K, 29K 32.5K, 29K1104X-5 30K, 27K 30.5K, 27K

II R2B 77K (,L) 28K, 25K 77K (AL) 28.5K, 25KSC-2B 77K (,u) 32K, 29K 77K (,u) 32.5K, 29KSC-2L 77K (p.) 32K, 29K 77K (pL) 32.5K, 29K

a H, Heavy chain; L, light chain; ,u, ,u heavy chain. K, X103.

VOL. 3, 1983

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 14

Feb

ruar

y 20

22 b

y 12

0.71

.5.1

18.


Intracellular

210K -19M)

21OK-

77K-_ _

32K-_29K-w-

a b C

FIG. 2. Immunoglobulins synthecell line. After cells were labeled forcine, both cellular extract and culimmunoprecipitated with anti-chiciulin serum (lanes a, b, d, and e) (

serum (lanes c and f). The immuneanalyzed under nonreducing conditiSDS gel (lanes a and d) or in their re9%o SDS gel (lanes b, c, e, and f).

Extracellular cence using antibody against ,u chain, whereasthe first did not (data not shown).

Glycosylation of immunoglobulin. All of thecell lines from ALV-induced lymphoma that weexamined produced and secreted two forms of

77K-im_ light chains. This could be due to synthesis oftwo different proteins or different modificationsof the same protein. We have examined thepossibility of different degrees of glycosylationby treating cells with tunicamycin. Tunicamycin

32.5K-4.- is known to block the metabolic pathway used to29K-w glycosylate immunoglobulins (9, 34).

When 1104X-5 cells were treated with tunica-mycin, the appearance of the more slowly mi-grating light chains was prevented (Fig. 3, lanes

d e t a through d). It was also observed that in thepresence of tunicamycin the electrophoretic mo-

:sized by SC-2L bility of the more rapidly migrating light chainsr 4 h with [3H]leu- was unchanged and that secretion of the proteinsIture media were continued. When R2B cells were treated withken immunoglob- tunicamycin, a similar effect on light-chain pro-or anti-chickeneL duction was observed (Fig. 3, lanes e through h).complexes wereions on a 6 to Wo In addition, instead of the Mr 77,000 p. heavy-duced forms on a chain, two polypeptides with molecular weights

of 76,000 and 68,000 were recovered (Fig. 3,lanes g and h). It seems reasonable to assumethat the proteins with molecular weights of

line SC-2L was the same as in the 1104X-5 cellsdescribed above, although the molecularweights of the light chains were different (Table2). In addition, the 28,000 Mr protein whichcould be precipitated from 1104X-5 cells bycontaminating antibody was not present in cellline SC-2L.To detect the presence of pentameric IgM, we

also analyzed unreduced samples on a 3% poly-acrylamide gel. The results indicated that thecells secreted monomeric IgM (Mr 210,000, asdescribed above) instead of pentameric IgM(data not shown). We also observed that, where-as equal amounts of p. heavy chain were precip-itated by antiserum against total chickenimmunoglobulin and antiserum specific for p.heavy chain, at least fivefold more light chainswere precipitated by antiserum against all class-es of chicken immunoglobulin (compare lanes band c and lanes e and f in Fig. 2). This indicatedthat the cells have a large pool of free lightchains and that free light chains were secreted.Our data show that the pattern of immuno-

globulin production by ALV-transformed cellscan be classified into two types (Table 2). Threecell lines (1104X-5, 1104B-1, and BK4484A)synthesize and secrete only light chains. Threeother cell lines (R2B, SC-2L, and SC-2B) syn-thesize and secrete monomeric IgM and freelight chains. The second type displayed 1gM ontheir surface when examined by immunofloures-

a b c d e f g h

,

L =

- + + _ -- 4.1. +

FIG. 3. Production of immunoglobulin in the pres-ence of tunicamycin. Cells were pretreated with medi-um containing 1 ,ug of tunicamycin per ml or withnormal medium for 4 h, then labeled with [35S]methi-onine in the presence of drug (+) or in control medium(-) for 4 h. Both cellular extract (lanes a, c, e, and g)and culture media (lanes b, d, f, and h) were immuno-precipitated with anti-chicken immunoglobulin serum.The immune complexes were analyzed on a 9o gel andautoradiographed. Lanes a through d, Cell line 1104X-5; lanes e through h, cell line R2B.

MOL. CELL. BIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 14

Feb

ruar

y 20

22 b

y 12

0.71

.5.1

18.


76,000 and 68,000 represent the membrane-bound and secreted form, respectively, of ,uchain.These findings suggest that the slowly migrat-

ing light chain in each cell line is glycosylated,but that the rapidly migrating light chain is notdetectably glycosylated. To confirm this obser-vation, we labeled 1104X-5 cells with [3H]man-nose and analyzed the glycosylated polypeptidesby immunoprecipitation and polyacrylamide gelelectrophoresis. Figure 4 illustrates that only theslowly migrating light chains were glycosylated;this result is consistent with the results obtainedwith tunicamycin.The above data indicate that the two forms of

light chains are probably related; one form isglycosylated, the other is not. We have also usedpartial hydrolysis with the V8 protease to com-pare light chains produced and secreted by the1104X-5 cell line in the absence and presence oftumicamycin. Several points emerged. First, thesame fragments were obtained by hydrolysis ofthe intracellular and extracellular forms of thelower-molecular-weight light chain (Fig. 5, lanesa and d). Second, the single light chain remain-ing in the presence of tunicamycin and thelower-molecular-weight form observed in theabsence of tunicamycin yielded identical pat-terns offragments (Fig. 5, compare lanes a and band lanes d and e). It is therefore unlikely thatthe higher-molecular-weight light chain repre-sents a second protein that in its unglycosylatedform happens to comigrate with the lower-mo-

a b

30K- -- -30.5K

FIG. 4. Production of glycosylated L chains by1104X-5 cell line. Cells were incubated with mediumcontaining 1 p.g of glucose per ml for 4 h before beinglabeled with [3H]mannose for 8 h. The glycosylated Lchains from cytoplasmic extract (lane a) and culturemedium (lane b) were analyzed by immunoprecipita-tion with anti-L serum.

Intracellular Sec reted

a b c d e f

21.5K-I*.-21K

4w -17K 17.5K- _ _4w _ -15K 4- __ _-12K ---

FIG. 5. Peptide maps of L chains from 1104X-5cells in the presence or absence of tunicamycin.[35S]methionine-labeled L chains were immunoprecip-itated from cellular extract or culture medium withanti-L serum and subjected to electrophoresis on a 9%oSDS-polyacrylamide gel. Bands corresponding to Lchains were excised from this unfixed gel and subject-ed to electrophoresis on a 14% SDS-polyacrylamidegel in the presence of 500 ng of S. aureus V8 protease.Lanes a, b, d, and e, Unglycosylated L chains; lanes cand f, glycosylated L chains; lanes b and e, L chainsprecipitated in the presence of tunicamycin (+).

lecular-weight light chain. It has been shownthat ALV-induced bursal lymphomas are clonalpopulations oftumor cells (22, 23); thus each cellline we examined is probably producing a singletype of light chain, both unglycosylated andglycosylated. Third, the pattern of fragmentsobtained by hydrolysis with V8 protease re-vealed glycosylated domains within the lightchains. The electrophoretic mobility of one frag-ment from the intracellular glycosylated lightchain was greatly retarded as compared to theunglycosylated form (Fig. 5, lanes a and c). Weattribute the retardation of mobility to glycosyla-tion (34). The same peptide was also glycosylat-ed in the secreted form of the light chain, as wasan additional peptide (Fig. 5, lanes d and f). Thesecond glycosylation in the extracellular lightchain presumably accounts for the fact that theintracellular and exracellular forms of glycosy-lated light chain consistently differed by a smallamount in their apparent molecular weights (Ta-ble 2).Does IgM produced by B-cells from ALV-

induced lymphomas bind to ALV? The antigen-specific cell surface receptors of B-lymphocytesare known to be immunoglobulins (35). Sincesome B-cell lines from ALV-induced lympho-mas produces both IgM and virus, we havetested the receptor-mediated leukemogenesishypothesis by asking whether the IgM produced

VOL. 3, 1983

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 14

Feb

ruar

y 20

22 b

y 12

0.71

.5.1

18.


void 700K 160K

WIA

Pr76-

p27

B9 a -1gM

-110K

L cha ins

elut ion

FIG. 6. Separation of virus and IgM produced bySC-2L cells. SC-2L cells were labeled with [35S]methi-onine for 6 h. The culture fluid was harvested andsubjected to filtration through Sepharose 4B. (A) Pat-tern of viral proteins specifically precipitated fromeach column fraction by antivirus serum. (B) Patternof IgM precipitated by anti-chicken immunoglobulinserum. The immune complexes were analyzed byelectrophoresis through a 6% polyacrylamide gel un-der nonreducing conditions.

by the cells is directed against the viruses theyproduce. This was done by labeling cells with[35S]methionine for 6 h and analyzing the culturemedium, which contained both labeled virusesand IgM, by gel filtration on a Sepharose 4Bcolumn. Viral particles and the free IgM elutedfrom this column in separate fractions. Figure6A shows the pattern of viral proteins immuno-precipitated from each fraction. The majority ofviral proteins eluted in the void volume. Thep27rar observed in the later fractions was proba-bly the result of cell lysis during the labelingperiod. Figure 6B gives the pattern of proteinsspecifically precipitated by anti-chickenimmunoglobulin serum. All of the IgM waseluted in the included fractions, well separatedfrom intact virions. The IgM-virus complexshould be eluted from this column in the voidfractons.

Since the immune complexes shown in Fig.6B were analyzed without reduction, this indi-cated that the monomeric IgM secreted by thecells remained intact during column chromatog-raphy. When the samples were reduced, the ,uheavy chain and two forms of light chain wererecovered (data not shown). Elution of lightchains in the later fractions, separated fromIgM, was the result of excess production andsecretion of free light chain by this cell line asdescribed before. We do not know the origin ofthe 110,000-molecular-weight protein that ap-peared in the fractions with IgM.

In the presence of antigen excess, we shouldsee IgM in the fractions where viruses eluted ifIgM bound to virus. By labeling with either[3HJleucine or [35S]methionine, we were able toshow that the cells under examination producemuch more viral protein than IgM (data notshown). The absence of antigen-antibody com-plexes indicated either that IgM made by thecells is not directed against virus, or that theaffinity between the IgM and virus is so low thatimmune complexes do not survive under ourexperimental conditions.

DISCUSSIONProduction of immunoglobulin by ALV-in-

duced lymphomas. Our results confirm and ex-tend previous reports that cell lines derived fromALV-induced bursal tumors represent the B-lymphocyte lineage. All of the cell lines weexamined synthesize and secrete immunoglob-ulins.As shown in Table 1, there is a correlation

between the length of passages in culture and thetype of immunoglobulin produced. The newlyisolated cell lines produce both heavy and lightchains, but the extensively passaged cell linesmake light chains only. In the murine system,immunoglobulin heavy-chain gene rearrange-ment appears to precede light-chain gene rear-rangement. Moreover, it was reported thatheavy-chain gene rearrangement continues incultured cells transformed by Abelson MuLV(1). Unless B-lymphocyte differentiation inchickens follows a scheme that is different fromthe one in mice, it is likely that the L heavy-chain gene in the extensively passaged chickenlymphoma cells is either deleted or rearranged ina nonproductive configuration. To confirm thisprediction, it will be necessary to study thearrangement of heavy-chain genes in these celllines. For the moment, it remains possible thatthe abnormality is at the level of transcription,mRNA stability, or translation.

Unstimulated, immunocompetent B-lympho-cytes synthesize membrane-bound monomericIgM (P,2L2), which serves as antigen receptors.

MOL. CELL. BIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 14

Feb

ruar

y 20

22 b

y 12

0.71

.5.1

18.


After antigenic stimulation, these cells differen-tiate into blast cells which secrete pentamericIgM (R2L2)5. However, all of the IgM-producingcells we examined secrete monomeric IgM. Wepresume that this is an abnormality due to theabsence of J chain or other deficiencies. Itremains possible, however, that secretion ofmonomeric IgM represents a discrete stage inavian B-lymphocyte development.

Expression of two forms of light chain. Most ofthe B-lymphocytes transformed by AbelsonMuLV synthesize IL heavy chain but not lightchain (32). By contrast, all of the ALV-inducedtumor cells we studied synthesize and secretelight chains. For want of suitable reagents, wehave not identified the light chains more specifi-cally, but previous reports suggest that at least90%o of the light chains produced by chickencells are of the lambda variety (11).The tumor cells studied here produce and

secrete both unglycosylated and glycosylatedforms of single light chains. It appears that as theglycosylated light chain is secreted, an addition-al glycosylation takes place. It has been reportedthat secretion of mouse K-type light chain isaccompanied by additional glycosylation (21).We presume, however, that glycosylation is notrequired for secretion because the unglycosylat-ed forms of the light chains are also secreted inabundance. Secretion of unglycosylated lightchains has been observed before (33), but this isto our knowledge the first report of the simulta-neous secretion of almost equal quantities ofunglycosylated and glycosylated light chains, ineither the absence or presence of heavy chains.

Antigenic specificity of IgM produced by ALV-induced B lymphomas. The genesis of thymiclymphomas by infection with MuLV has beenattributed to mitogenic stimulus arising from thebinding of virus to antigen receptors on thesurface of the neoplastic cells (16, 20). A deci-sive test of this hypothesis has not beenachieved, primarily because the antigen recep-tors of T-cells have themselves remained elu-sive. The production of B-cell tumors by ALVand MuLV provides easier access to the poten-tial role of antigen receptors in tumorigenesis. Ifbinding of viral antigen is an essential compo-nent in the genesis of these tumors, then theneoplastic cells should possess antigen receptorsthat recognize viral proteins (most likely, theglycoproteins exposed on the surface of viralparticles), and antibody produced by the tumorcells should in turn be directed against viralantigen. We have explored these possibilitieswith tumors induced by ALV and have failed tofind any evidence that the tumors produce anti-body against ALV. We sought the antibody inthe form of immune complexes. As a result, ourfindings are not definitive: there is an inestima-

ble chance that the affinity between antibodyand viral antigen is low and, as a consequence,immune complexes do not survive our experi-mental conditions.Some of the cell lines we have characterized

do not produce complete immunoglobulin. It istherefore impossible that these cells still rely onmitogenic stimulus by antigen for sustenance ofgrowth. These same lines have been in culturefor extended periods of time, however, and theymay have undergone secondary changes thathave circumvented the need for antigenic stimu-lus.Phenotype of ALV-induced lymphoma cells.

Transformation of hematopoietic cells by leuke-mia viruses may cause arrest of the tumor cellsat a particular stage in differentiation (2). Ourstudy of the phenotype of B-lymphocytes de-rived from ALV-induced lymphomas revealedseveral characteristics which cannot presentlybe assigned to any known stage of the pathwayfor B-lymphocyte differentiation. These includethe following: (i) secretion of monomeric IgM;(ii) production of both glycosylated and nongly-cosylated light chains from a clonal population;and (iii) the synthesis and secretion of light chainwithout detectable heavy chain. It seems likelyto us that either neoplastic transformation orpropagation in culture disturbed the normal pro-gram of development so that the tumor cells nowhave no normal counterparts. However, in theabsence of a thorough study of normal chickenB-lymphocyte development, the possibility thatB-cells from ALV-induced lymphomas displaya normal immunological phenotype cannot beeliminated.

In summary, the ALV-induced B lymphomaseems to be more mature than the B-lympho-cytes transformed by either reticuloendothelio-sis virus-T (18) or Abelson MuLV (32). It hasbecome apparent, however, that the phenotypeof tumor cells does not necessarily reveal thenature of the cell in which tumorigenesis began.For example, B-cells transformed by AbelsonMuLV continue to differentiate during propaga-tion in culture (1), and the erythroleukemiasinduced by avian erythroblastosis virus are com-posed of cells more mature than the cells that arevulnerable to initial attack by the virus (10). It istherefore possible that tumorigenesis by ALVbegins early in lymphoid development, perhapsbeyond the confines of the bursa, but becomesevident only as B-cells differentiate under theinfluence of bursal microenvironment.We have good reason to view tumorigenesis

by ALV as a protracted affair. First, tumorsemerge only many months after viral infection.Second, many of the early ("preneoplastic")lesions that appear in the bursa regress withtime; only a few survive to progress to frank

VOL. 3, 1983

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 14

Feb

ruar

y 20

22 b

y 12

0.71

.5.1

18.


malignancy (24). Third, at least two geneticabnormalities apparently participate in tumori-genesis induced by ALV: the enhanced expres-sion of c-myc (12, 28), elicited by integration ofviral DNA and thus an early event in tumorigen-esis; and the activation or mutation of a secondcellular oncogene, detectable by transfectioninto NIH 3T3 mouse cells (5). Either or both ofthese abnormalities may influence the course ofB-cell development.

ACKNOWLEDGMENTS

We thank H. Oppermann, H. E. Varmus, G. K. Lewis,G. Evans, C. Nottenberg, and E. Matthews for assistance andadvice, and J. Marinos for preparing the manuscript.The work reported here was supported by grants from the

National Cancer Institute and the American Cancer Society.

LITERATURE CITED

1. Alt, F. W., N. E. Rosenberg, S. Lewis, E. Thomas, andD. Baltimore. 1981. Organization and reorganization ofimmunoglobulin genes in Abelson murine leukemia virus-transformed cells: rearrangement of heavy but not lightchain genes. Cell 27:381-390.

2. Beug, H., M. J. Hayman, and T. Graf. 1982. Leukaemia asa disease of differentiation: retroviruses causing acuteleukaemias in chickens. Cancer Surv. 1:205-231.

3. Bonner, W. M., and R. A. Laskey. 1974. A film detectionmethod for tritium-labelled proteins and nucleic acid inpolyacrylamide gels. Eur. J. Biochem. 46:83-88.

4. Cleveland, D., S. J. Fisher, M. W. Kirschner, and U. K.Laemmli. 1978. Peptide mapping by limited proteolysis insodium dodecyl sulfate and analysis by gel electrophore-sis. J. Biol. Chem. 252:1102-1106.

5. Cooper, G. M., and P. E. Neiman. 1981. Two distinctcandidate transforming genes of lymphoid leukosis virus-induced neoplasms. Nature (London) 292:857-858.

6. Cooper, M. D., H. G. Purchase, D. E. Bockman, and W. E.Gathings. 1974. Study on the nature of the abnormality ofB cell differentiation in avian lymphoid leukosis: produc-tion of heterogeneous IgM by tumor cells. J. Immunol.113:1210-1222.

7. Durban, E. M., and D. Boettiger. 1981. Replicating differ-entiated macrophages can serve as in vitro targets fortransformation by avian myeloblastosis virus. J. Virol.37:488-492.

8. Durban, E. M., and D. Boettiger. 1981. Differential effectsof transforming avian RNA tumor viruses on avian macro-

phages. Proc. Natl. Acad. Sci. U.S.A. 78:3600-3604.9. Elbein, A. D. 1981. The tunicamycins-useful tools for

studies on glycoproteins. Trends Biol. Sci. 6:219-221.10. Gazzolo, L., J. Samarut, M. Bouabdelli, and J. P. Blan-

chet. 1980. Early precursors in the erythroid lineage are

the specific target cells of avian erythroblastosis virus invitro. Cell 22:683-691.

11. Grant, J. A., B. Sanders, and L. Hood. 1971. Partial aminoacid sequences of chicken and turkey immunoglobulinlight chains. Homology with mammalian chains. Bio-chemistry 10:3123-3132.

12. Hayward, W. S., B. G. Neel, and S. M. Astrin. 1981.Activation of a cellular onc gene by promoter insertion inALV-induced lymphoid leukosis. Nature (London) 290:475-480.

13. Hihara, H., T. Shimizu, and H. Yamamoto. 1974. Estab-lishment of tumor cell line cultures from chickens withavian lymphoid leukosis. Natl. Inst. Anim. Health Q.14:163-173.

14. Hood, L., J. Griffin, S. Crews, H. Huang, M. Kronenberg,and S. Kim. 1981. Antibody and MHC genes, p. 805-824.In C. Janeway, E. E. Sercarz, and H. Wigzell (ed.),Immunoglobulins and idiotypes. Academic Press, Inc.,

New York.15. Laemmli, U. K. 1970. Cleavage of structural proteins

during the assembly of the head of bacteriophage T4.Nature (London) 227:680-684.

16. Lee, J. C., and J. N. Ihie. 1981. Chronic immune stimula-tion is required for Moloney leukemia virus-induced lym-phoma. Nature (London) 289:407-409.

17. Levinson, A. D., H. Oppermann, L. Levintow, H. E.Varmus, and J. M. Bishop. 1978. Evidence that transform-ing gene for avian sarcoma virus encode a protein kinaseassociated with a phosphoprotein. Cell 15:561-572.

18. Lewis, R. B., J. McClure, B. Rup, D. W. Niesel, R. F.Garry, J. D. Hoelzer, K. Nazerian, and H. R. Bose. 1981.Avian reticuloendotheliosis virus: identification of thehematopoietic target cell for transformation. Cell 25:421-431.

19. McGrath, M., 0. Witte, T. Pincus, and 1. L. Weissnan.1978. Retrovirus purification: method that conserves en-velope glycoprotein and maximizes infectivity. J. Virol.25:923-927.

20. McGrath, M. S. and I. L. Weissman. 1978. AKR leukemo-genesis: identification and biological significance of thy-mic lymphoma receptors for AKR retroviruses. Cell17:65-76.

21. Melchers, F. 1970. Biosynthesis of the carbohydrate por-tion of immunoglobulins. Biochem. J. 119:765-772.

22. Neel, B. G., W. S. Hayward, H. L. Robinson, J. Fang, andS. M. Astrin. 1981. Avian leukosis virus-induced tumorshave common proviral integration sites and synthesizediscrete new RNAs: oncogenesis by promoter insertion.Cell 23:323-334.

23. Neiman, P., L. N. Payne, and R. A. Weiss. 1980. ViralDNA in bursal lymphomas induced by avian leukosisviruses. J. Virol. 34:178-186.

24. Neiman, P. E., L. Jordan, R. A. Weiss, and L. N. Payne.1980. Malignant lymphoma of the bursa of Fabricius:analysis of early transformation, p. 519-528. In M. Essex,G. Todaro, and H. Hausen (ed.), Viruses in naturallyoccurring cancers. Cold Spring Harbor Conferences onCell Proliferation, vol. 7. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

25. Okazaki, W., R. L. Witter, C. Romero, K. Nazerian, J. M.Sharma, A. Fadly, and D. Ewert. 1980. Induction oflymphoid leukosis transplantable tumors and the estab-lishment of lymphoblastoid cell lines. Avian Pathol.9:311-329.

26. Oppermann, H., A. D. Levinson, and H. E. Varmus. 1981.The structure and protein kinase activity of proteinsencoded by nonconditional mutants and back mutants inthe src genes of avian sarcoma virus. Virology 108:47-70.

27. Opperman, H., A. D. Levinson, H. E. Varmus, L. Levin-tow, and J. M. Bishop. 1979. Uninfected vertebrate cellscontain a protein that is closely related to the product ofthe avian sarcoma virus transforming gene (src). Proc.Natl. Acad. Sci. U.S.A. 76:1804-1808.

28. Payne, G. S., J. M. Bishop, and H. E. Varmus. 1982.Multiple arrangements of viral DNA and an activated hostoncogene in bursal lymphomas. Nature (London) 295:209-217.

29. Payne, G. S., S. A. Courtneige, L. B. Crittenden, A. M.Fadly, J. M. Bishop, and H. E. Varmus. 1981. Analysis ofavian leukosis virus DNA and RNA in bursal tumors: viralgene expression is not required for maintenance of thetumor state. Cell 23:311-322.

30. Purchase, H. G., and D. G. Glhmour. 1975. Lymphoidleukosis in chicken chemically bursectomized and subse-quently inoculated with bursa cells. J. Nati. Cancer Inst.55:851-855.

31. Purchase, H. G., and B. R. Burmester. 1978. Neoplasticdiseases: leukosis/sarcoma group, p. 418-468. In M. S.Hofad, B. W. Calnestk, D. F. Helmboit, W. M. Reid, andH. M. Yoder (ed.), Disease of poultry, Iowa State Univer-sity Press, Ames.

32. Siden, E. J., D. Baltimore, D. Clark, and N. E. Rosenberg.1979. Immunoglobulin synthesis by lymphoid cells trans-

MOL. CELL. BIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 14

Feb

ruar

y 20

22 b

y 12

0.71

.5.1

18.


formed in vitro by Abelson murine leukemia virus. Cell16:389-396.

33. Tartakoff, A., and P. Vassalli. 1979. Plasma cell immuno-globulin M molecules: their biosynthesis, assembly andintracellular transport. J. Cell Biol. 83:284-299.

34. Vassalli, P., R. Tedghl, B. Lisowska-Bernstein, A. Tarta-

koff, and J.-C. Jaton. 1979. Evidence for hydrophobicregion within heavy chains of mouse B lymphocyte mem-brane-bound IgM. Proc. Nati. Acad. Sci. U.S.A. 76:5515-5519.

35. Vitetta, E. S., and J. W. Uhr. 1975. Immunoglobulin-receptors revisited. Science 189:964-969.

VOL. 3, 1983

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 14

Feb

ruar

y 20

22 b

y 12

0.71

.5.1

18.

immunological phenotype of lymphomas induced by avian

Documents