Proc. Nati. Acad. Sci. USAVol. 80, pp. 6376-6380, October 1983Medical Sciences

Characterization of human hybridomas secreting antibody totetanus toxoid

(monoclonal antibodies/cell fusion/cell sorter)

JAMES W. LARRICK, KENNETH E. TRUITT, ANDREW A. RAUBITSCHEK, GEORGE SENYK, ANDJANET C. N. WANGCetus Immune Research Laboratories, 3400 West Bayshore Road, Palo Alto, CA 94303

Communicated by Hugh 0. McDevitt, July 5, 1983

ABSTRACT We have selected a thioguanine-resistant lym-phoblastoid cell line (LTR228) that forms human-human hybridswith high efficiency. Fusions with peripheral B cells consistentlyyield one colony per 105 cells plated. To produce antitetanusmonoclonal antibodies, we withdrew blood from persons who hadrecently received booster injections of tetanus toxoid. T cells wereseparated from peripheral mononuclear cells by 2-aminoethyliso-thiouronium bromide-induced rosette formation, given 1,500 rads(1 rad = 0.01 gray), and cultured in a 1:1 ratio with nonrosettingcells. After 3 days of pokeweed mitogen stimulation, heterokary-ons were produced by a plate-fusion technique and cultured inIscove's Dulbecco's minimal essential medium for 24 hr prior tohybrid selection. Colonies appeared after 10-14 days in hypoxan-thine/azaserine supplemented medium. A direct binding enzyme-linked immunosorbent assay with specific tetanus toxoid inhibitionidentified positive wells. The hybridomas were cloned twice in softagarose and by limiting dilution. The subcloned hybridomas dou-ble every 26 hr (vs. every 16 hr for LTR228) and produce 1-5 ,ugof specific IgG,K antibody per 106 cells per ml per 24 hr. All sub-clones (almost 200) continue to secrete antibody after 11 monthsof continuous culture. Twelve representative subclones have neartetraploid amounts of DNA. From hybridomas grown in 5-literspinner flasks, milligram quantities of the IgG,K antibody werepurified by staphylococcus protein A affinity chromatography.Specific antibody from hybridoma cultures protected mice in-jected with 1,000 times the LD50 of tetanus toxin. Our cell line andassociated techniques should permit the production of therapeu-tically important human monoclonal antibodies.

Many attempts to produce human monoclonal antibodies havebeen made since the early mouse hybridoma work of Kohlerand Milstein (1). Initial reports of successful fusions of humanB cells and the U266 myeloma line (2) have not been confirmedin other laboratories. Problems with mycoplasma contamina-tion of parent lines, suitable pre-fusion preparation of B cells,hybrid stability, and quantity of secreted monoclonal antibodyhave hampered most efforts with other human myeloma lines(3). Other approaches have also met with limited success. Thesehave included mouse plasmacytoma by human B-cell hybrid-ization (4, 5), direct Epstein-Barr virus (EBV) transformationof human B cells (6, 7), and fusion with transformed lympho-blastoid cell lines (8, 9), or combinations of the above (10, 11).We have developed an EBV-positive splenic lymphoblastoid

cell line, LTR228, that forms stable human-human hybridomaswith high efficiency. Here we describe the production andcharacterization of antitetanus human monoclonal antibodies.

MATERIALS AND METHODSCell Line. LTR228 is a lymphoblastoid cell line that origi-

nated in spleen cell culture (12). We selected a spontaneous 6-

thioguanine (20 ,g/ml)-resistant mutant for its high fusing ca-pacity. Our line is mycoplasma-free and doubles every 16 hrwhen grown in Iscove's Dulbecco's minimal essential medium(DME medium) (GIBCO) with 15% fetal calf serum. LTR228has a typical lymphoblastoid nuclear EBV nuclear antigen stain-ing pattern.

Preparation of B Cells. Volunteers were vaccinated intra-muscularly with tetanus toxoid (Wyeth, Philadelphia). Periph-eral blood (80 ml) was drawn in heparin 9 days after vaccina-tion. Lymphocytes were separated by Ficoll-Hypaque densitygradient centrifugation. The lymphocytes were washed twicein Hanks' balanced salt solution and the T cells formed rosetteswith 2-aminoethylisothiouronium bromide-treated sheep eryth-rocytes as described (13). The rosetted T cells (6 X 106) wereexposed to 1,500 rads (1 rad = 0.01 gray) in a Gammacell 40gamma-irradiator (Atomic Energy of Canada) and subsequentlymixed with 6 x 106 nonrosetting B cells. The mixed cells werecultured at a density of 106 per ml in Iscove's DME medium/15% fetal calf serum/1% pokeweed mitogen (GIBCO). After 3days the cells were again separated by Ficoll-Hypaque to obtaina population of primarily B-cell blasts. Preliminary experi-ments showed that B-cell blasts fused better than unstimulatedlymphocytes. The irradiated rosette-forming T cells gave muchless suppression than unfractionated lymphocytes.

Fusion Protocol. Cells were combined and fused accordingto the protocol of Brahe and Serra (14) with slight modification.Cells were fused with 40% polyethylene glycol 4,000 (BDH)/10%0 dimethylsulfoxide/poly(L-Arg) (5 ,ug/ml) (Sigma) in Hanks'balanced salt solution without calcium and with magnesium. Byusing this technique, we consistently observed low cell deathand formation of 3-8% heterokaryons. Cells were cultured 24hr after fusion prior to addition of hypoxanthine (100 ,uM)/aza-serine (8 ,g/ml) in Iscove's DME medium with 15% fetal calfserum. Fusions were plated at a density of 105 cells per mi-crotiter well for hybrid selection. Cultures were subsequentlyfed every 3 days with selective medium. Hybrid colonies nor-mally appeared after 10-14 days.

Enzyme-Linked Immunosorbent Assays (ELISA). Culturesupernatants were tested for anti-tetanus toxoid antibody andtotal Ig by the ELISA technique (15).

Tetanus ELISA. Microtiter plates (Immulon II, Dynatech,Alexandria, VA) were coated with purified tetanus toxoid (10/,g/ml) (Massachusetts State Laboratories, Boston) in sodiumcarbonate buffer (50 mM, pH 9.5) at 4°C overnight. Wells wereblocked with 1% bovine serum albumin/0. 1% gelatin in phos-phate-buffered saline (Pi/NaCI) for 2 hr at 37C. Culture su-pernatants and high-titer antitetanus human serum standardswere diluted in Pi/NaCl and incubated for 1 hr at 37°C. Tetanustoxoid (10 ,g/ml) was added to duplicate wells to demonstrate

Abbreviations: EBV, Epstein-Barr virus; ELISA, enzyme-linked im-munosorbent assay; Pi/NaCl, phosphate-buffered saline.

6376

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

M:Proc. Natl. Acad. Sci. USA 80 (1983) 6377

specific inhibition. Wells were washed with Pi/NaCl and de-veloped with peroxidase-conjugated rabbit anti-human y-, ,u-,K-, or A-chain-specific antibodies (DAKO Labs, AccurateChemicals, Westburg, NY) diluted 1/1,000 in P;/NaCl/1% bo-vine serum albumin. After a final wash, peroxidase substrate2,2'-azino-di(3-ethylbenzthiazoline sulfonic acid) (ZYMED, SouthSan Francisco, CA) was added and OD415 was read with a Ti-tertek ELISA plate reader (Flow Laboratories).

Ig ELISA. Ig levels of culture supernatants were measuredby a chain-specific inhibition ELISA. Plates were coated withpurified human Ig (10 pg/ml)/carbonate buffer, pH 9.5, at 40Covernight. Wells were blocked as described above. Superna-tants or standards and affinity-purified peroxidase-labeled anti-K, A, y, ,u, or a were mixed and added to the wells. Plates wereprocessed in the same manner as the tetanus ELISA.

Biosynthetic Labeling of Igs. Cells (20 X 106) were culturedovernight in 4 ml of methionine-free medium/1% dialyzed fe-tal calf serum/0.8 mCi of [3S]methionine (1 Ci = 3.7 X 10"'Bq; New England Nuclear). Cells were pelleted and Igs fromthe tissue culture supernatant were precipitated with chain-specific rabbit anti-human antibodies attached to Staphylococ-cus aureus (Cowan I strain). Immunoprecipitates were elec-trophoresed on a 5-15% polyacrylamide gradient gel in Na-DodSO4 according to the procedures of Weber and Osborn(16). Gels were stained with Coomassie blue, soaked for 1 hrin EN3HANCE (New England Nuclear), dried, and fluoro-graphed on Dupont Cronex film.

Purification of' Monoclonal Antibody. Monoclonal antiteta-nus antibody was purified by staphylococcus protein A-Seph-arose affinity chromatography (Pharmacia). Antibody was elutedwith 100 mM acetate/150 mM NaCl, pH 2.3.

Electron Microscopy. Logarithmic phase cells were washedtwice in Pi/NaCl and fixed with 1% paraformaldehyde in Pi/NaCl. Cells were attached by poly(L-Lys) to microscope slides,dehydrated, critical-point dried, and gold-coated prior to ex-amination in an ISI-40 scanning electron microscope.

Cell Sorter Methods. DNA histograms. Preparation of cellsfor DNA analysis was performed as described (17) with a slightmodification. Cells (105) were sedimented in U-bottom micro-titer wells, medium was aspirated, and cold 50% ethanol/Hanks'balanced salt solution was added. Cells were incubated for 30min on ice, pelleted, and washed. The pellet was resuspendedin 0.1 M Tris.HCI/0.1 M NaCl, pH 7/Hoechst-33342 at 0.4;g/ml (Calbiochem) and incubated at 22°C for 15 min.

Cytofluorography was carried out on an EPICS V cell sorter(Coulter) equipped with a 5-W argon laser (Coherent Innova-90). The laser output was adjusted to 100 mW UV, at 350-364nm. A 408-nM-long pass filter shielded the detector. At least5,000 cells were accumulated for each histogram.

Cell sorter cloning. The EPICS V was sterilized with 0.05%NaOCI, and laser.output was adjusted to 488 nM with 1,000mW of power.. Scatter and fluorescence windows were set toensure sorting of viable cells. Hybrids were cloned at one cellper well into U-bottom microtiter wells with Iscove's DME me-.dium/20%' fetal calf serum.

Surface antigen analysis. Cells were stained with fluores-cent antibodies by the method of Loken and Stall (18). Fluo-rescein isothiocyanate-anti-Leu-10 and anti-DR, anti-y, .anti-,reagents were purchased from Becton Dickinson. Anti-Bi(Coulter) was developed with goat F(ab')2 from TAGO (Burlin-game, CA) and ,u-chain specific goat anti-mouse F(ab')2 frag-ments (TAGO).

Cloning and Reverse-Plaque Techniques. Limiting dilution(0.3 cells per well) cloning was performed in 96-well U-bottomplates (Costar, Cambridge, MA) in Iscove's DME medium/20%fetal calf serum.

For soft agar cloning, 1,000 cells in 1 ml of 0.33% SeaPlaque

agarose (FMC,. Rockland, ME), made in Iscove's DME me-dium/20% fetal calf serum, were placed over a bed of 4 ml ofSeaKem agarose (04%) in 60-mm culture dishes (Falcon).To select for a nonproducer parent cell line or for high-pro-

ducer hybrids a reverse-plaque technique was used. Protein A-coated sheep erythrocytes (1.0%) were added to the upper layerof soft agar according to the method of Gronowicz et al. (19).Clones secreting more antitetanus Ig had the largest plaques.A subclone of LTR228 secreting no Ig made no plaques.

RESULTSTable 1 summarizes our preparation of antitetanus humanmonoclonal antibodies; several points are noteworthy. (i) Pe-ripheral blood from an immunized subject provided a suitablesource of B cells. (ii) The eventual appearance of colonies inevery well indicated that the fusion process produced at leastone viable hybrid for every 120,000 cells plated. This efficiencyof 1 in 5 x 104 LTR228 cells plated approaches that routinelyachieved in the mouse system. (iii) Almost one-half of the wellswith growing cells produced IgG or IgM (or both); 5% of thewells showed specific antitetanus activity.We initially cloned the antitetanus cells by limiting dilution

and in soft agar. Well El produced the greatest number of an-titetanus-positive clones, and we selected these cells for furtherstudy. Sublines were recloned by limiting dilution, cell sorter,and soft agar methods. All subsequent clones, about 200, havecontinued to produce antitetanus antibody.

Morphological Characteristics of, Human Hybridomas. Bylight microscopy-the hybrids appeared to be larger than LTR228.However, scanning electron microscopy studies (Fig. 1) did notconfirm this difference.

Although the hybridomas secrete up to 5 A.g of specific an-tibody per ml, transmission electron microscopy shows bothhybrids and the parent LTR228 to have very little rough en-

Table 1. Antitetanus human monoclonal antibody cellfision protocol

Day Status ProcedureImmunization

0 Subject immunized intramuscularly withtetanus toxoid

Pre-fusion preparation9 After immuni- Withdraw blood; PWM stimulation

zation

Fusion3 After PWM Fusion of 6 x i05 B-cell blasts and 6 x 106

LTR2281 After. fusion Plate 120,000 per well; add hypoxanthine/

azaserine; feed every 3 days14 Colonies appear19 Tet ELISA 5+ wells, 2± wells21 47 wells IgG', 39 wells IgM'29 Growth in 100% of wells

First cloning0 A4, El, Gi, G3 200 wells each by lim. dil.; 2 x 60-mm

dishes each by soft agar16 After cloning Screen 200 soft agar and 800 lim. dil. clones:

.11/48 El soft agar clones antitet positive;3/100 El lim. dil. clones antitet positiveSecond cloning

16 After cloning All soft agar clones antitet positive; all lim.dil. clones antitet positive; 12/12 clonestetraploid by cell sorter

PWM, pokeweed mitogen; tet, tetanus; lim. dil., limiting dilution.

Medical Sciences: Larrick et al.

6378 Medical Sciences: Larrick et al.

FIG. 1. Scanning electron microscope photographs-of LTR228 (A)and antitetanus hybridomas (B).

doplasmic reticulum (data not shown). Other workers have alsonoted that the presence of rough endoplasmic reticulum maynot correlate with the capacity of B-cell lineage cell lines to se-crete Ig (20).-Growth Characteristics of Human Hybridomas. The hy-

bridomas double every 25-28 hr as compared to every 16 hr forthe parent (Fig. 2). They can easily be, adapted to growth inserum-free Iscove's medium supplemented with bovine serumalbumin (700 ,ug/ml)/transferrin (10 ,g/ml)/insulin (0.2 unit/ml). Under such conditions, the time required for hybrid cul-tures to reach confluence is 30% longer than in Iscove's DMEmedium/15% fetal calf serum.

Stability of Human Hybridomas. After two rounds of clon-ing, all sublines continue to secrete antitetanus antibody. Eightmonths in culture and expansion of cells in 4-liter vessels hasnot altered the secretion of specific antibody. Furthermore,cell sorter histograms showed 12 representative clones all tohave nearly tetraploid DNA levels (Fig. 3). Chromosome stud-ies yield similar conclusions (Fig. 4). LTR228 has a median of48 chromosomes; hybrids have 91.

Quantity of Monoclonal Antibody Secreted. Fig. 5 showsthe levels of antitetanus Ig produced by 12 subclones. Spenthybridoma culture medium (24 hr) of 106 cells contains 1-2.5tkg of specific antibody per ml. A reverse-plaque technique wasused to select for high-producer clones. With this method, wehave been able to. derive sublines that secrete twice as muchIg (data not shown).

Biochemical Characterization of Monoclonal AntitetanusAntibodies. Fig. 6 shows the results of chain-specific immu-noprecipitation carried out on [35S]methionine-labeled culture

V-E54 / 'I' -e

X 0

1',0 50 100

Time, hr

FIG. 2. Comparison of-growth rates of parent cell line (closed sym-bols) and four antitetanus hybridomas (open symbols). Doubling timeof parent is 16 hr; doubling time of hybrids is 26 hr.

supernatants. LTR228 produces an IgM, K antibody; the hybridsubclones secrete both an antitetanus IgG, K and the parent Ig.The IgG antibody binds to staphylococcus protein A with high

affinity. We have used staphylococcus protein A affinity chro-matography to purify milligram quantities of monoclonal an-titetanus Ig. We have purified antibody from both serum-sup-plemented and serum-free cultures.

Surface- Phenotype of Parent and Hybridoma Cells. B-celllineage surface markers were measured on hybrid and LTR228cells by cell sorter methods. It was hoped that one or more ofthe B-cell lineage surface markers might correlate with in-creased secretion of Ig. Fig. 7 shows plots of forward-angle lightscattering (cell size) vs. logarithm fluorescence for surface Igs,DR, Leu-10 (a D-region-associated antigen) and B1 (a B-celldifferentiation antigen) (21). When compared with LTR228, thehybrids express less BL, more DR, more Leu-10, and approx-imately equivalent amounts of surface Ig. The parent LTR228also stained with several anti-y reagents, despite the fact thatIgM was the only Ig internally labeled and immunoprecipitated(see Fig. 6). The plots also show the cloned hybrids to be moreheterogeneous in size than the parent cell lines.-Hybrids- cloned by the cell sorter for increased or decreased

expression of B1, DR, or Leu-10 antigens failed to secrete alevel of Ig different from the parent (data not shown).

Ia)

-0E

C2)

logarithm of fluorescence

FIG. 3. DNA histograms of parent LTR228 (left peak) to 12 anti-tetanus-producing hybridomas (right peaks). Note that all 12 are neartetraploid.

Proc. Natl. Acad. Sci. USA 80 (1983)

Proc. Natl. Acad. Sci. USA 80 (1983) 6379

3.0

bgj

"g 2.0a~a0

a3& 1.0-6a

1 2 3 4 5 6 7 8 9 101112

B

I+

. .A..l;@lL:.K

FIG. 4. Representative metaphase chromosome spreads of our an-titetanus hybridoma (A) and.parent cell line LTR228 (B).

Monoclonal Antibody Protects Mice from Lethal TetanusToxin Injection. Outbred Swiss-Webster mice (20 g) were in-jected intraperitoneally (0.5 ml) with various -doses of tetanustoxin mixed with serum-free medium conditioned by either hy-bridoma or LTR228 cells. Doses of toxin were 1-1,000 timesthe LD' . Results in Table2 show that the monoclonal antibodyis protective in vivo against the lethal effects of tetanus toxin.

DISCUSSIONSeveral workers have shown that EBV-infected B-cell lines canbe used to produce human monoclonal antibodies (8, 9, 11, 22).Although human myeloma lines are available,- they have notbeen as useful in producing hybridomas as EBV-infected B-celllines. Even under optimal conditions, human myeloma cell linesfuse with- lower frequency, grow more slowly, and are unstable.producers of Ig (unpublished data).We have developed an EBV-infected B-cell line that greatly

facilitates production of human hybridomas. By using-this linewe have produced and purified-a specific human monoclonalantibody, IgG,K, that protects mice against tetanus toxin. Thecell line we developed produces hybrids with a high frequency(1 per 5 x 104 parent cells) and has been used to produce hu-man monoclonal antibodies to other antigens (unpublished re-

CloneFIG. 5. Production of specific antitetanus monoclonal antibodies

by 12 representative clones. Values are for 106 cells at 24 hr of spentculture medium.

suits). Using this technique, we have shown that.multiple clonesare stable producers of antibody over a period of 11 months.We have been able to boost antibody production from less than1 jig/ml to almost 10 pg per 106 cells per 24 hr. It is noteworthythat the parent line LTR228 and all antitetanus hybridomas clonewith atieast 10% efficiency in soft agar and by limiting dilution.This characteristic allowed us to clone these cells frequentlyand to select for high producers by the reverse-plaque tech-nique. Because the cells grow rapidly in soft agar, it is oftenpossible to pluck positive clones after 3 days. We have also usedthe reverse-plaque technique to select for nonproducer cell lines,and initial experiments suggest these lines produce hybrids asefficiently as LTR228. We have constructed a ouabain/6-thio-guanine-resistant derivative of LTR228 and have fused this lineto EBV-transformed B cells secreting anti-blood group A an-tibody (unpublished results). These hybrids and cells from threesuccessive subclones produce an IgM anti-blood group A Igsimilar to the parent lymphoblastoid line. Kozbor et al. (11) havealso used this technique to immortalize rare B-cell clones.The reasons why EBV-transformed lines (8, 9, 11, 22) such

as LTR228 have to date proven superior to myeloma cell lines

1 2 1 2 1 2 1 2

Al- a k _ _

r -0

Light _chains

dm

,j I_ _J I-I

y K

I I

A

FIG. 6. Chain-specific immunoprecipitation ofparent LTR228 (lanes2) and antitetanus hybridoma (lanes 1). Reduced samples have-beenresolved by NaDodS04/polyacrylamide gel electrophoresis. Anti-yprecipitates the antitetanus Ig. Anti-,u precipitates parental IgM in su-pernatants from hybrid and LTR228. Anti-K precipitates both.anti-bodies from hybrid supernatant but only the parental Ig in the.case ofLTR228. Anti-A fails to react with either hybrid or parent supernatant.

Medical Sciences: Larrick et al.

V 4. -,`.iol- i.4

El.. /.i- .:.,M .... - maL'. -Ei I is'

iI%..filIV--. le-.IW--( ..ik

_%N+4000.JOU "

6380 Medical Sciences: Larrick et al

DR

Leu-10

Bi

FIG. 7. Cell sorter plots of forward light scattering (vertical) vs.

logarithm offluorescence (horizontal) for several B-cell lineage surfacemarkers. (Left) LTR228. (Right) Hybrid.

for generation of human B-cell hybridomas are unclear. A rapidgrowth rate and a high cloning efficiency both in soft agar andby limiting dilution are important characteristics of LTR228.Selection for subclones that efficiently form heterokaryons ("fu-sibility") also improves the chances of successful hybridomaoutgrowth. Tetraploid human hybrids are not inherently un-

stable, but certain differentiated functions including produc-tion of Igs are often lost after fusion (23, 24). The available hu-man myeloma lines may be in a state of differentiation unableto form antibody-producing hybrids with more immature B-cellblasts. Although a myeloma-plasmacytoma parent might be ex-

pected to produce more Ig, our efforts to correlate surface Bcell differentiation markers and Ig secretion rates have to datebeen unsuccessful (Fig. 7). However, the ease with which hy-brids of LTR228 clone and form reverse plaques has permittedus to select for high-secreting clones approaching the produc-tion of mouse hybridomas.One of the antitetanus hybrids has been adapted to growth

Table 2. Protection of mice against tetanus toxin

Toxin, Mediumtunits* Parent cell line Antitetanus hybrid

0 3/3 3/31 1/2 3/310 0/3 3/3

100 0/3 3/31,000 0/3 2/3

Results indicate surviving mice (at day 4)/total mice injected.* 1 unit = LD50..to.5 ml of-spent medium was mixed with tetanus toxin and injectedintraperitoneally.

in nude mice (unpublished results). Inoculation of 107 cells sub-cutaneously produces tumors within 3-4 weeks; After removalfrom the mouse and subsequent in vitro culture, this clone haspermitted the production of ascites with a high titer of humanmonoclonal antibody. When nude mice were injected withadapted cells (2 x 107), ascites containing 0.5 mg of specificantibody per ml formed within 2 weeks.

Production of purified antigen-specific monoclonal antibodyin milligram quantities will make it possible to use human an-tibodies when murine products are either ineffective or im-munogenic. Passively administered human antibodies can beused to treat hemolytic disease of the newborn (Rh), tetanus,rabies, hepatitis B, Gram-negative sepsis, and snakebite. Fu-ture uses may include radioimaging with tracer-labeled anti-bodies, toxin-antibody conjugates to modulate immunologicaldiseases, and antibodies to eliminate toxins or drugs. Humanproteins should be less antigenic and, in some cases, more spe-cific, and they should have longer in vivo half-lives because ofthe correct species-specific carbohydrate side chains and Fc se-quences. Human Fc sequences may also augment the bioac-tivity of antibodies binding to macrophages or lymphocytes invivo.We gratefully acknowledge the expert technical assistance of Hanna

Hutchins and Howard Weintraub. Bradley Dyer and David Buck pro-vided helpful discussions. We are indebted to Dr. Wally Laird for hisin vivo testing of our antibodies and to Dr. Jon Kosek for the scanningelectron microscope photographs. Joan Murphy and Dianne Jacobs gavesuperb secretarial help. Tim Culp prepared the graphics.

1. Kohler, G. & Milstein, C. (1975) Nature (London) 256, 495-497.2. Olsson, L. & Kaplan, H. S. (1980) Proc. Nati. Acad. Sci. USA 77,

5429-5431.3. Sikora, K., Alderson, T., Phillips, J. & Watson, J. V. (1982) Lan-

cet i, 11-13.4. Lane, H. C., Shelhamer, J. H., Mostowski, H. S. & Fauci, A. S.

(1982) J. Exp. Med. 155, 333-338.5. Butler, J. L., Lane, H. C. & Fauci, A. S. (1983)J. Immunol. 130,

165-168.6. Steinitz, M., Klein, G., Koskimies, S. & Makela, 0. (1977) Na-

ture (London) 269, 420-422.7. Kozbor, D. & Roder, J. C. (1981) J. Immunol. 127, 1275-1280.8. Croce, C. M., Linnenback, A., Hall, W., Steplewski, Z. & Ko-

prowski, H. (1980) Nature (London) 288, 488-489.9. Chiorazzi, N., Wasserman, R. L. & Kunkel, H. G. (1983)J. Exp.

Med. 156, 930-935.10. Kozbor, D., Roder, J. C., Chang, T. H., Steplewski; Z. & Ko-

prowski, H. (1982) Hybridoma 1, 323-328.11. Kozbor, D., Lagarde, A. E. & Roder, J. C. (1982) Proc. Natl. Acad.

Sci. USA 79, 6651-6655.12. Levy, J. A., Viroloinen, V. & Defendi, V. (1968) Cancer 22, 517-

524.13. Madsen, M. & Johnson, H. E. (1979)J. Immunol. Methods 27, 61-

74.14. Brahe, C. & Serra, A. (1981) Somatic Cell Genet. 7, 109-115.15. Engvall, E. & Perlmann, P. (1971) Immunochemistry 8, 871-880.16. Weber, K. & Osborn, M. (1969)J. Biol. Chem. 244, 4406-4411.17. Taylor, I. W. & Milthorpe, B. K. (1980)J. Histochem. Cytochem.

28, 1224-1232.18. Loken, M. R. & Stall, A. M. (1982)J. Immunol. Methods 50, R85-

R112.19. Gronowicz, E., Coutinho, A. & Melchers, F. (1976) Eur. J. Im-

munol. 6, 588-590.20. Kozbor, D., Dexter, D. & Roder, J. C. (1983) Hybridoma 2, 7-

16.21. Stashenko, P., Nadler, L. M., Haroy, R. & Schlossman, S. F. (1980)

J. Immunol. 125, 1678-1685.22. Shoenfeld, Y., Rauch, J., Massicotte, H., Datta, S. K., Andre-

Schwarts, J., Stollar, B. D. & Schwartz, R. S. (1983) N. Engl. J.Med. 308, 414-420.

23. Bengtsson, B. D., Nabholz, M., Kennett, R. H. & Bodmer, W.F. (1975) Somatic Cell Genet. 1, 41-64.

24. Stanbridge, E. J.,, Der, C. J., Roerson, C. J., Nishimi, R. Y., Peehl,D. M., Weissman, B. E. & Wilkinson, J. E. (1982) Science 215,252-259.

Proc. Natl. Acad. Sci. USA 80 (1983)