Tn~ JOURNAL OF IMMUNOLOGY Vol, 114, No. 2, Par t 2, February 1975 Copyright © 1975 by The Will iams & Wilklns Co. Printed in U.S.A.

R E T I C U L U M CELL S A R C O M A : A N E F F E C T O R CELL IN A N T I B O D Y - D E P E N D E N T C E L L - M E D I A T E D I M M U N I T Y 1

PETER RALPH, JOHN PRICHARD, AND MELVIN COHN

From the Salk Institute for Biological Studies, San Diego, California 92112

A transplantable, murine reticulum cell sarcoma is described which exhibits the cytologic, adherence, and phagocytic properties of macrophages. It forms specific rosettes with erythrocytes in the presence of the corre- sponding anti-serum. The ascites cells mediate antibody- dependent cellular immunity as assayed by release of radioactivity from 51Cr-labeled erythrocytes. The contribu- tion of contaminating host cells in the cytotoxic reaction was ruled out by growing the tumor in F1 mice and removing the host cells by anti-H2 serum and complement. The tumor cells have receptors for IgG2a and IgG2b immunoglobtdins. The availability of a pure population of effector cells in the immune system allows study of the biochemical processes pursuant to lysis of foreign cells.

Cell types other than lymphocytes are involved in a number of reactions in the immune system. These range from part ici- pat ion in the induction of humoral and cel l -mediated immune responses to effector functions of phagocytosis and lysis. Immunogens are rapidly collected at the dendri t ic processes of ret icular cells in the follicles of lymphoid organs (1). Antigen bound in this way abuts on most of the follicular lymphocytes and is postulated to increase the efficiency of induction of B cells ~ to ant ibody production (2). In addi t ion certain antigens which have been ingested by macrophages show increased immunogenici ty (3). The adherent populat ion which includes maerophages is required in vitro fbr the induction of humoral immuni ty (4) and T cell cytotoxicity (5). Such adherent cells are reported to possess receptors for T cell monomeric IgM which allow them to "present" antigen (6) or inductive signals (7) to precursors of ant ibody-fbrming cells.

Macrophages and other phagocytic ceils (8) also play a key role in effector mechanisms of the immune system. Cells in the adherent populat ion media te cytotoxicity dependent on hu- moral ant ibody (9, 10). The effector cells showing act ivi ty comparable to adul t spleen cells are already present in fetal liver and newborn spleen organs well before immune function appears (9, 11). Macrophages are also reported to exert cytostat ic and cytotoxic effects on syngeneic and allogeneic tumor cells in the presence of specific T lymphocyte factors (12, 13). Another type of cell which mediates ant ibody-depend- ent cytotoxici ty has been described as non-phagocytic (11, 14)

Submitted for publication July 30, 1974. 1This work was supported by the Ford Foundation, Rockefeller

Foundation, National Science Foundation Grant GB-37869 to Peter Ralph, National Institute of Allergy and Infectious Diseases Grant AI-05875, and AI-00430 to Melvin Cohn.

2Abbreviations used in this paper: B cell, thymus-independent lymphocyte; T cell, thymus-derived lymphocyte; SRBC, sheep red blood cells; BRBC, burro red blood cells; PBS, phosphate-buffered saline.

or nonadherent (15), with some characterist ics of m o n o c y t e s (16) and of lymphocytes (11). This killing mechanism is effective against microorganisms (17), xenogeneic cells (9-11, 14), and allogeneic (13, 18) and syngeneic tumor cells (19).

In this paper we characterize a ret iculum cell sarcoma with the morphologic, adherent, and phagocytic properties of mac- rophages. This ascites tumor line bears receptors for IgG2a and IgG2b immunoglobulin and mediates an t ibody-dependent cy- totoxicity. In view of the heterogeneity which exists in lymph- oid and peri toneal cell populations, the avai labi l i ty of such tumors which can be used to replace a normal cell type will be invaluable in analyzing immune processes.

METHODS AND MATERIALS

Mouse tumor lines. J774 tumor arose in a female BALB/c/ NIH mouse in 1968 during a p lasmacytoma induction program (20). The init ial neoplasm was ascitic, with solid abdominal tumors also found. The tumor was passaged subcutaneously in which case it metas tas ized to the liver, or as ascites. Experi- ments with J774A ascites were performed with cells from passages 7-24, showing greater than 90% viabi l i ty by t rypan blue exclusion.

Staining and microscopy. Direct smears of ascitic fluid were fixed in methanol and s tained with Giemsa. Tissues removed at necropsy were fixed in neutral buffered 10% formalin, imbedded in paraffin, sectioned and stained with hematoxylin and eosin (courtesy of Monique Lacorbiere). For electron microscopy, cells were fixed in osmium tetroxide, embedded in Vostopol-W, sectioned and s ta ined with uranyl acetate and lead citrate (courtesy of' Marlene Bajak).

Antisera and antigens. Four-month-old BALB/c females were immunized with 0.2 ml of a 1% suspension of either sheep (SRBC) or burro (BRBC) erythrocytes (Colorado Serum Co., Denver, Colo.) incorporated into complete Freund 's adjuvant . Immunizat ions were repeated after 4 weeks in incomplete Freund 's adjuvant . After an addi t ional 4 weeks a single i.v. injection of 0.2 ml of 1% RBC in saline was given. Sera were prepared from mice 10 days after the last injection. Hemag- glutination titers of BALB ant i -SRBC and BALB anti-BRBC were 4000 to 8000 against the homologous antigen, and less than 4 against the heterologous antigen. BALB/c ant i-H2 b sera were prepared by 8 weekly i.p. injections of 107 live EL4 ascites cells, a C57BL leukemia (21). The cytotoxic t i ter of this ant iserum to EL4 cells with rabbi t complement was 1:1000.

Erythrocyte antigen-binding assay. Ascites cells were har- vested into cold Eagle 's medium without serum, containing 5 I.U. sodium heparin per milli l i ter. Cells were washed by low speed centrifugation three t imes with cold Eagle 's medium to remove red cells and debris. Next 1 × 105 viable cells were plated in 2-ml tissue culture dishes (Falcon 3001) in Eagle's medium minus serum. Cells were allowed to incubate over- night in order to adhere to the culture dishes. Adherent cells

898

at Em

ory University H

ealth Sciences Library on D

ecember 9, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

1975 ]

were washed twice with phosphate-buffered saline (PBS). One mill i l i ter of immune or normal sera, di luted in PBS, was added to each dish followed by the addit ion of 1.0 mI of a 0.25% RBC suspension. The dishes were then incubated for 20 min in a humidif ied CO2 incubator at 37°C. After incubation the dishes were washed four times with 2 ml volumes of PBS. Final ly 1 ml H~O was added to lyse the remaining adherent erythrocytes, and the supernatant was removed after agitation. Superna- tants were either diluted in H20 or read directly at 415 nm in a Zeiss spectrophotometer to quant i ta te erythrocyte binding to J774.

Antibody directed cytotoxicity mediated by cells. The chro- mium release assay was used as described for the assay of' cellular immuni ty in vitro (9). 10 ~ ~lCr-labeled RBC were used as targets. Cells tested for killer function in the presence of' antisera were 107 spleen cells from 2-month-old female BALB/c mice, or 2 × 106 cells from J774A ascites, MOPC 70A ascites (a BALB/c myeloma secreting IgG1) or P388 ascites (a DBA/2 leukemia). Cells were cultured at 37°C in a total volume of 1 ml of Dulbecco modified Eagle 's medium plus 5% fetal calf" serum in Falcon 3001 tissue culture dishes with rocking. Specific ant iserum or control serum was added in a 10-ttl volume. After 6 hr, the contents of' dishes were suspended by scraping, centrifuged, and the percentage of radioactive chromium released to the supernatant was determined. Specific chro- mium release was calculated by correcting for background release in the absence of ant iserum or test effector cells (9). Assays were performed in duplicate and s tandard errors were _<5% of the means.

Removal of host cells from J774A ascites preparations. J774A was grown as ascites in (C57BL/6 × BALB/c)F1 mice. We suspended 4 × 107 cells in siliconized tubes in balanced salt solution containing absorbed rabbi t complement (I:10 final concentration) and anti-H2 b serum (1:i00), and incubated them for 45 min at 37°C. Less than 10% of the cells were kil led by this t rea tment , mostly the small host cells present in the preparat ion. Control incubations used complement and nor- mal BALB/c serum at 1:100. F~ spleen cells were s imilar ly t reated; incubation with anti-H2 serum caused over 80% lysis of nucleated cells, whereas incubation with normal serum and complement killed less than 20% of cells. After incubation cells were washed twice with PBS and tested for effector cell cytotoxicity.

Aggregated myeloma proteins. The following purified pro- teins were used: IgM, MOPC104E; IgG1, MOPC21; IgG2a, ADJPC5; IgG2b, GPC5; IgG3, J606; and IgA, S107. Samples were aggregated with bis-diazotized benzidine (23), dialyzed, and centr ifuged 1 min at 10,000 × G. The supernatants containing moderate ly aggregated protein which showed ab- sorbance of 0.1 to 0.5 at 400 nm for a protein concentration of 1 mgN/ml were used (see Table V in Reference 22). Samples were iodinated by the chloramine-T method (23) with 12~I to a specific act ivi ty of 104 to 108 cpm/t~g.

RESULTS

Characterization of tumors. At autopsy, mice carrying J774A tumors were found to have metastases of cells to liver, lymph nodes, uterus, ovaries, and lungs. If injected subcutaneously, tumors did not develop at the injection site, but rather metastas ized to the liver. Neoplastic cells were not found in the peripheral blood.

Microscopic examinat ion showed metastasis to all abdomi- nal organs and lungs (Figs. 1, and 2a). Silver impregnat ion of

RETICULUM CELL SARCOMA--MODEL IN CELLULAR IMMUNITY 899

tissue sections revealed abundant reticulin fibers within tumor loci (Fig. 2b). The ascites cells are about 30 t*m in diameter , adhere firmly to plastic dishes, and contain ample lightly- staining cytoplasm with numerous vacuoles (Fig. 3). Most of the cells develop long processes after 24-hr incubation. The cells rapidly take up neutral red stain and carbonyl iron and form rosettes with glutaraldehyde-f ixed RBC, as do macro- phages (24). By gos s and microscopic morphology, J744A is identical to the type A ret iculum cell sarcoma previously described by Dunn (25), and belongs to the general category of histiocytic lymphomas.

Binding of antigen-antibody complexes by J774A. Retention of erythrocytes by J774A cells in the presence of' anti-eryth- rocyte serum is shown in Figures 4 and 5. At the higher ant i-serum concentrations, each tumor cell had bound at least 10 erythrocytes, as seen in the microscope after extensive washing. Fibroblas ts failed to b ind erythrocytes under the same conditions. If the adherent J774A cells were washed after incubation with immune sera, no antigen was seen to be bound after a second incubation with antigen alone (Fig. 5a). Reten- tion of erythrocytes was specific in tha t incubation of J774A ceils with BALB/c ant i -burro erythrocytes failed to induce binding of sheep erythrocytes (Fig. 5b), and vice versa (not shown).

Antibody-directed cell lysis mediated by J774A ascites cells. Since J774A cells bind antigens in the presence of the appropr ia te antisera, we tested fbr killing of' the target cells by the chromium release assay. Cells harvested from J774A ascites fluid were highly efficient in the lysis of ~Cr- labeled erythrocytes, releasing 50 to 77% of the radioactive chromium by 6 hr of incubation together with either rabbi t or mouse ant i -SRBC (Table I). The magni tude of lysis by 2 × 108 J774A cells was similar to that observed with 107 normal spleen cells. In control cultures incubated with normal rabbi t or mouse serum (Table I), or with ret iculum ascites cells replaced by myeloma or lymphoma ascites (Table II), only negligible release of radioact ivi ty over background values was observed. Occasionally the J774A cells showed a slight lysis of erythro- cytes in the absence of antiserum, usually less than 10% of the opt imal lysis in the presence of specific ant iserum. As with killing by normal spleen cells, the ret iculum cell sarcoma was effective at very low concentrations of antisera, up to 10 8 dilution. The cytotoxicity does not appear to require comple- ment since the same lysis of SRBC occurs in medium with or without fetal calf serum, as shown for spleen effector cells (9).

The J774A ascites preparat ions include various numbers (1 to 10%) of small, nucleated cells, presumably normal perito- neal cells. The an t ibody-dependent cytotoxicity demonst ra ted here appears to be media ted by the large tumor cells ra ther than contaminat ing normal peri toneal ceils since killing did not correlate with the degree of contaminat ion, and ascites forms of other tumors were not effective (Table II). Pre l iminary experiments with normal peri toneal cells and peritoneal exu- dates indicate tha t the number of host cells would have to approximate the numbers of tumor cells in ascites preparat ions to give the cytotoxicity seen here.

In order to prove more directly tha t the J774A cells were the effector cells, the BALB/c tumor was passaged in (C57BL/6 > BALB/c)F1 mice and the host F1 cells in the ascites populat ion killed by anti-H2 b serum plus complement. This t rea tment abolished the ant ibody-dependent cytotoxicity of normal F1 spleen cells, while leaving in tac t the function of ascites tumor cells to lyse SRBC in the presence of ant i -SRBC (Table III). The ascites fluid contained mostly H2 d tumor cells resis tant to

at Em

ory University H

ealth Sciences Library on D

ecember 9, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

9OO PETER RALPH, JOHN PRICHARD, AND MELVIN COHN [VOL. 114

anti-H2 b killing, whereas there was a massive lysis of the F1 spleen cells which might block the function of surviving F~ effector cells. Therefore a mixing experiment was performed, which showed that surviving J774 effector cells could still function after incubation in the presence of antibody lysis of spleen cells (Table lII).

Specificity of immunoglobuIin receptor on J774A cells. Myeloma proteins with known antigen-binding activity were tested tbr their ability to mediate rosette formation with erythrocytes coupled with the corresponding antigen. The systems tested were IgM: MOPC104E with dextran; IgA: J558 and dextran; IgG3:J606 and levan (26). Despite hemagglutina- tion titers of 1000 to 8000 against the homologous antigen, no rosette tbrmation occurred (data not shown). Myeloma pro- teins were then tested for inhibition of the binding of antibody- coated erythrocytes to J774A. Amounts of IgM, IgG1, IgG2a, IgG2b, IgG3, and IgA myeloma proteins up to 1 mg/ml did not

diminish adherence of erythrocytes when preincubated 10 to 30 min with tumor cells before addition of RBC.

Since myeloma proteins did not prevent antigen-antibody complex binding by competition, a positive experiment was performed to determine the specificity of the immunoglobulin receptor on J774 cells. Myeloma proteins and normal immuno- globulins of the proper class frequently gain biologic activity nonspecifically by aggregation, comparable to the correspond- ing antigen-antibody complex. Thus when aggregated, IgG induces increased capillary permeability and consequent skin reaction and fixes complement; IgM fixes complement only; IgE gives the skin reaction only; and IgA and IgD are inactive (27). Since antibody directing binding and lysis of foreign targets has a much higher affinity for the effector cell when complexed with antigen (Fig. 5a; References 9, 10), aggregated myeloma proteins were used to mimic this situation. Conse- quently we tested 12~I-labeled, aggregated myeloma proteins

Figure 1. Metastatic lesion of the reticulum cell sarcoma in the lung of a BALB/c mouse. Tumor cells are quite compact, having deeply basophilic nuclei. Note cellular pleomorphism within an isolated nodule (H & E, × 320).

at Em

ory University H

ealth Sciences Library on D

ecember 9, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

1975] RETICULUM CELL SARCOMA--MODEL IN CELLULAR IMMUNITY 901

Figure 2. A, section of intra-abdominal tumor which was adherent to omentum, peritoneum and bowl (H & E, × 160); B, same area as 2A (Gomori's silver stain, x 160).

TABLE I

Antibody-dependent lysis of SRBC by reticulum cell sarcoma

Antiserum Dilution % Specific Lysis

Spleen cells J774A

NRS 10 4 1 < 1 RaSRBC 10 4 56 53

NMS 10 a < 1 < 1 NMS 10 " < 1 2

MaSRBC 10 3 82 66 MaSRBC 10-' 79 77 MaSRBC 10- 5 32 34 MaSRBC 10 -8 7 13

Each incubation contained 105 5~Cr-labeled SRBC and either I0 ~ nucleated spleen cells from 2-month-old normal unimmunized female BALB/c mice or 2 × 108 J774A cells, as described in Materials and Methods, Rabbit anti-SRBC (RaSRBC, Pentex) or normal rabbit serum (NRS) was added (10 ~1 of 1:100) to a final dilution of 1:10000 (10 4), as indicated. To other incubations, 10 ul of dilutions of normal BALB/c mouse serum (NMS) or BALB/c anti-SRBC (MaSRBC) were added to obtain the final dilutions indicated. Radioactivity released to the supernatant was determined after 6 hr. Results are corrected for background release of ~Cr with spleen cells (11%) o~ J774A (12%) in the absence of rabbit or mouse serum.

for b ind ing to J774A. Aggregat ion was pe r fo rmed with bis-di- azot ized benz id ine unt i l the tu rb id i ty was op t ima l for comple- m e n t f ixat ion (22). As shown in Figure 6, J774A cells b o u n d IgG2a and IgG2b, and to a lesser ex ten t IgG1, bu t no t IgM, IgG3, or IgA aggrega ted immunog lobu l in . At t he s ame cell

concen t r a t i on sp leen cells b o u n d only smal l a m o u n t s of* IgG2a and IgG2b. A 0 + leukemia , EL4, showed no s igni f icant b ind ing

of immunog lobu l in .

DISCUSSION

The r e t i cu lum cell s a r coma line descr ibed here has a n u m b e r of proper t ies charac te r i s t ic of macrophages . The J774A cells

are readi ly adhe ren t , phagocyt ic , and able to b i n d specif ical ly a n t i b o d y - c o a t e d eD~throcytes. The sa rcoma cells func t ion very well as effector ceils in a n t i b o d y - d e p e n d e n t cytotoxic i ty . J774A

cells are as act ive as 5 t imes as m a n y spleen cells in lysing

e ry throcy tes in the p resence of the app rop r i a t e a n t i s e r u m

(Tables I, II). S ince the J774A ceils are not a d a p t e d to growth in cul ture , all e x p e r i m e n t s were pe r fb rmed wi th asc i tes p repa- ra t ions . The smal l n u m b e r s of hos t cells in the asci tes f luid

could not accoun t for the kill ing react ion, s ince when the t u m o r

TABLE II

J774A as the effector cell in ascites preparations

Antiserum Dilution Cytotoxicity of Effector Cells °

Spleen J774A P388 MOPC70A

<1 4 <1 <1 MaSRBC 1 0 ~ 64 77 9 < 1 NMS 10 -4 <1 6 4 <1

a Lysis of 5~Cr-labeled SRBC was determined as in Table t, using 107 BALB/c spleen cells or 2 × 10 e J774A, P388, or MOPC70A ascites cells. Results were corrected for background release of ~lCr from SRBC (10%) in the presence of spleen cells but no antiserum.

at Em

ory University H

ealth Sciences Library on D

ecember 9, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

902 PETER RALPH, JOHN PRICHARD, AND MELVIN COHN [VOL. I14

was grown in F~ hosts, t rea tment with complement and ant iserum to F1 cells blocked all cytotoxicity by host spleen cells but left the ascites tumor act ivi ty intact (Table IID.

In considering the signaling of an effector cell by ant ibody- antigen complexes, the interaction of free immunoglobulin with cell Fc receptors must be considered. If the binding constant for free immunoglobulin (or a cytophilic subclass) approximates its serum concentration, there may be competi- tion for binding of ant ibody-ant igen complexes. In the case of the basophil or mas t cell which mediates anaphylac to id reactions via h is tamine release, the affinity of the cell for free immunoglobul in of the IgE class is high. However, its dissocia- t ion constant (~10 8 M) is equal to or higher than the serum level of IgE which is ~10 9 M (28). Consequently, these cell receptors are sa tu ra ted with IgE only in unusual cases (28, 29), and normal ly IgE cannot compete effectively with an antigen- ant ibody complex.

In the present case of phagocytosis or cel l -mediated lysis, the affinity of monomeric (non-aggregated) immunoglobulin for the adherent cell is low. The normal level of' the IgG classes

which arm these cells is of the order of 10 .5 M (approximately 1 mg/ml). The dissociation constant is not measurable and might be >_ 10 -4 M, so that these Fc receptor sites are also not normally sa tura ted. In both the macrophage and mast cell, the effector function is triggered only if the cell~bound ant ibody is aggregated on the surface, normal ly by interact ion with a mul t ide te rminant antigen. For complexes of several antibodies bound to a mul t iva lent antigen, it is likely tha t the complex has a much higher binding constant than free ant ibody.

In par t these considerations explain the difficulty in estab- lishing the class of ant ibody directing non-T cell cellular immunity . The use of competi t ion experiments with myeloma proteins to define the class of ant ibody normally media t ing this type of cellular immuni ty can be misleading. Fi rs t the concen- trat ion of the competing monomeric (non-aggregated) immu- noglobulin must be at least 10-fbld above the binding constant (~10 4 M) to see an inhibit ing effect. Second, nonspecific interactions between the effector cell and the ant igen-ant ibody complex, e.g., charge, could s tabi l ize the interaction irreversi- bly once the complex is fixed to the cell via the ant ibody Fc

: i i ̧ ¸¸ ~ i ! : / } : : :

i!iiiiii!

Figure 3. Direct Smear of J774A ascites (Giemsa stain, × 1200).

at Em

ory University H

ealth Sciences Library on D

ecember 9, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

1975] RETICULUM CELL SARCOMA--MODEL IN CELLULAR IMMUNITY 903

iiiiiiiiiiiiiii!iiiiiiiiiiiiiii%,!i~/~!

i~iiii!iii!i!i!iiiiiiiiii~!~ili!iii~i!i~i

iiiiiiiiiiiiiiiii!!ii!ii!iiiiiiii!ii!ii!!iil

~i: ~: :~!~:i~ ::~i: ~:~! :ii: •:~: : i/¸ ~F :~: ::: ~:: !/)•: ~i ~ :::~: :+::i:iii: ,~

i ...... i~ii/ ~!:i (i i I •

.......

F ~ +iiiii~i~ ¸/ : ; L ̧ : / : )///:T~ ̧ )

/ i ii~ I~ i ii

i (:~!iii I i~ ~

Figure 4. Binding of SRBC to J774A cells in the presence of anti-SRBC. An erythrocyte has apparently been phagocytosed, although phagocytic vacuoles were not observed (electron micrograph, × 7500).

E ,~ 8

I.tJ C.)

Z

m n -

O co 4 r n

I ' 1 ' I' I I " 1

anti BRBC anti BRBC

BRBC

washed ~0 SRBC

gso Vzo V,o ,/400 V,oo VsO ,/40 V2oo

SERUM DILUTION

Figure 5. a, Binding of BRBC to the reticulum cell sarcoma in the presence of antiserum. J774A was incubated with anti-BRBC 20 min at 37°C before addition of antigen (O), as described in Methods and Materials. In another series of dishes, after incubation with anti- BRBC, the adherent J774A cells were washed three times with saline prior to the addition of BRBC (@). Absorbed erythrocytes, as measured by hemoglobin absorption at 415 nm, is plotted against dilutions of anti-BRBC serum. Each point is the average of triplicate incubations with the range indicated by bars. Absorbance at 415 nm with antiserum replaced by PBS = 0.04. b, Specificity of antigen binding.

portion. Third, if an effector cell has independen t b inding sites for different classes of immunog lobu l in it will be possible to block the binding of an an t igen-an t ibody complex only by a mix ture of mye loma proteins, not any one. These considera- t ions may explain our inabi l i ty to block rosette format ion with J774A by using individual mye loma proteins and for a s imilar failure to show class specificity in the case of a myelomonocyt ic l eukemia (30).

In spite of these difficulties, mouse IgG2b appears to be most active in blocking the killing by nonadhe ren t effector cells (16). We have shown tha t the adheren t J774A effector cells have receptors for both IgG2 classes (Fig. 3). The par t ic ipa t ion of IgG2a and IgG2b ant ibody in cytotoxici ty is curious. Both classes fix complemen t (31). Al though external complemen t components are not required in the reaction (9, 32), it is possible t ha t the effector cell itself synthesizes one or more c o m p l e m e n t components which are secreted and act ivated lo- cally to effect cytotoxicity. Synthesis of c o m p l e m e n t compo- nents by macrophages has been repor ted (33). J774A cells do not bind mouse IgG3, and apparen t ly macrophages do not have receptors for this class (34). Thus a l though IgG3 immunog lobu- lin is t r a n s m i t t e d to the fetus more efficiently t han other IgG

BRBC (O) or SRBC (@) were incubated with J774A in the presence of anti-BRBC, and the retained erythrocytes measured as in a. Absorb- ance for J774A + BRBC in PBS = 0.01" for J774A + SRBC in PBS = 0.03.

at Em

ory University H

ealth Sciences Library on D

ecember 9, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

904 PETER RALPH, JOHN PRICHARD, AND MELVIN COHN [VOL. I14

TABLE HI RBC lysis by J774A after removal of contaminating host cells

Cytotoxicity of Effector Cells ~ Trea tment

Spleen J774 Spleen + J774

NMS + C 45 (<1) 51 (5) 43 (3) Anti-H2 b + C <1 (<1) 54 (4) 41 (3)

a (C57BL/6 × BALB/c)F, spleen cells, J774A ascites cells grown in F1 mice, or a mixture of these two populations were incubated with complement and either normal mouse serum (NMS + C) or anti-H2 b serum (anti-H2 b ~ C) to kill F~ cells, as described in Materials and Methods. 107 control or treated spleen cells and 2 × l0 s control or treated J774, based on cell numbers before serum plus complement treatment, were tested for cytotoxicity with 10 -4 dilution of BALB/c anti-SRBC as in Table It. Numbers in parentheses indicate RBC lysis using normal mouse serum in the assay.

Q Z

0

z _q

S 0

20

I 0

IO

I0

J 7 7 4 A

SPLEEN

[ 7 n

EL4

F3 - -

r - i r - i I"-1 r-3 - - M GI G2 a G2 b G3 'A"

Figure 6. J774A ascites, BALB/c spleen or EL4 ascites cells were washed four times in balanced salt solution (BSS). 107 cells (open bars) or 2 × 10 ~ cells (hatched bars) were incubated with 1 ~g aggregated, 125I-labeled myeloma proteins {described in Methods and Materials) in BSS in siliconized tubes. After 30 min incubation on ice, the cells were washed four times, transferring to new tubes after the second wash. Radioactivity remaining with the cells was determined and is shown as the percentage of initial protein of IgM, IgG~, IgG2~, IgG2b, IgGs, and IgA classes.

classes (34), it is not responsible ibr the ant ibody-dependent cellular immuni ty which we have shown to be derived from the maternal circulation (9).

There are contradictions in the literature which need expla- nation. For example, mouse macrophages have receptors demonstrat ing predominately IgG2a specificity, however, some but not all IgG1 myeloma proteins block macrophage rosette formation as efficiently as IgG2 molecules (35). Either there are two subclasses of IgG1 or some of the IgG1 preparations were denatured. All IgG immunoglobulin classes both bind to mast cells and block passive cutaneous anaphylaxis. Yet only IgGI and IgE molecules will mediate this reaction (36).

Since J774A cells resemble macrophages, we considered the possibility that these cells would exhibit the function of

adherent cells in the induction of immune responses in vitro. When tested at 102 to l0 s J774A cells/1 ml cultures containing 107 spleen cells depleted of adherent cells (24), there was no restoration of the induction of antibody formation (unpub- lished results of P. R., J. Watson, and E. Trenkner). In fact, when 10 s to 108 J774A cells were added to normal spleen cultures, the normal induction of antibody-forming cells was prevented. The tumor cells do not grow in culture. However, they may exert an inhibitory effect on cultures which could override the manifestation of an adherent cell cooperating function. Activated macrophages (37-40) and leukocytes (41) are generally toxic to many other ceil types. On the other hand we have presented evidence elsewhere (24) that pure popula- tions of macrophage-like cells derived from bone marrow cultures in the presence of colony-stimulating factor also do not replace adherent cells in immune responses in vitro. A mlmber of adherent or phagocytic cell types in lymphoid organs have been distinguished (42, 43). A virus-transformed macrophage cell line has been described which restores the immune response of nonadherent spleen cells (44). It is therefore likely that there are several classes of macrophages or adherent cells with different functions.

Acknowledgmen t s . We thank Dr. R. Hyman for advice and encouragement, Drs. J. Watson and J. Andersson for helpful discussion, Drs. M. Weigert, R. Riblet, and B. Slade for purified myeloma proteins, and Ms. I. Nakoinz for expert assistance.

Note added in proof: Walker and Demus have recently de- scribed a virally transformed macrophage cell line which ex- hibits extensive phagocytosis and moderate lysis of chicken erythrocytes in the presence of specific antisera (J. Immunol., this issue, p. 765). We have now adapted J774A ascites to growth in culture. The cell line shows mainly phagocytosis, in contrast to the original ascites line or ascites-passed culture line in which lysis predominates.

REFERENCES

1. Nossal, G. J. V., Abbot, A., Mitchell, J., and Lummus, Z., J. Exp. Med, 127: 277, 1968.

2. Guttman, G. and Weissman, I. L., Immunology, 23: 465, 1972. 3. Pierce, C. W., Kapp, J. A., Wood, D. D., and BenacerraL B., J.

Immunol., 112: 1181, 1974. 4. Hartmann, K. U., Dutton, R. W., McCarthy, M. M., and Mishell,

R. I., Cell. Immunol., 1: 182, 1970. 5. Wagner, H., Feldmann, M., Boyle, W., and Schrader, J. W., J.

Exp. Med., 136: 331, 1972. 6. Feldmann, M., Cone, R. E., and Marchalonis, J. J., Cell. Im-

munol., 9: 1, 1973. 7. Bretscher, P. A. and Cohn, M., Science, 169: 1042, 1970. 8. Simmons, S. R. and Karnovsky, M. L., J. Exp. Med., 138: 44, 1973. 9. Ralph, P., Nakoinz, I., and Cohn, M., Nature (New Biol.), 245:

157, 1973. 10. Weissman, I., Lannin, D., Jerabek, L., and Barclay, T., Cell.

Immunol, 7: 222, 1973. 11. Britton, S., Perlmann, H., and Perlmann, P., Cell. Immunol., 8:

420, 1973. 12. Lohmann-Matthes, M. L., Ziegler, F. G., and Fisher, H., Eur. J.

Immunol., 3: 56, 1973. 13. Grant, C. K., Currie, G. A., and Alexander, P., J. Exp. Med., 135:

150, 1972. 14. Greenberg, A. H., Hudson, L., Shen, L., and Roitt, I. M., Nature

(New Biol.), 242: 111, 1973. 15. Wisloff, F. and Froland, S. S., Scand. J. Immunol., 2: 151, 1973. 16. Greenberg, A. H., Shen, L., and Roitt, I. M., Clin. Exp. Immunol.,

15: 251, 1973.

at Em

ory University H

ealth Sciences Library on D

ecember 9, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from

1975] RETICULUM CELL SARCOMA--MODEL IN CELLULAR IMMUNITY 905

17. Diamond, R. D., Nature, 247: 148, 1974. 18. Kiessling, R. and Klein, E., J. Exp. Med., 137: 527, 1973. 19. Pollack, S. and Nelson, K., J. Immunol., 110: 1440, 1973. 20. Hirst, J., Jones, G., and Cohn, M., J. Immunol., 107: 926, 1971. 21. Ralph, P., J. Immunol., 110: 1470, 1973. 22. Ishizaka, K. and Ishizaka, T., J. Immunol., 85: 163, 1960. 23. Williams, C. A. and Chase, M. W. (Editors), Methods in Immuno l -

ogy and I m m u n o c h e m i s t r y , Vol. 1, p. 391, Academic Press, New York, 1967.

24. Watson, J., Thoman, M., Ralph, P., and Trenkner, E., J. Immunol., 112: 1873, 1974.

25. Dunn, T., J. Natl. Cancer Inst., 14: 1281, 1954. 26. Lundblad, A., Steller, R., Kabat, E., Hirst, J., Weigert, M., and

Cohn, M., Immunochemistry, 9: 535, 1972. 27. Ishizaka, T., Ishizaka, K., Bennich, H., and Johansson, S., J.

Immunol., 104: 854, 1970. 28. Ishizaka, T , Soto, C. S., and Ishizaka, K., J. Immunol., 111: 500,

1973. 29. Bazaral, M., Orgel, H. A., and Hamburger, R. N., Clin. Exp.

Immunol., 14: 117, 1973. 30. Cline, M. J. and Metcalf, D., Blood, 39: 771, 1972.

31. Warner, N. L., Vaz, N. M., and Ovary, Z., Immunology, 14: 725, 1968.

32. Van Boxel, J. A., Paul, W. E., Green, I., and Frank, M. M., J. Immunol., I12: 398, 1974.

33. Colten, H. R. and Boros, T , J. Immunol., 112: 1107, I974. 34. Grey, H. M., Hirst, J. W., and Cohn, M., J. Exp. Med., I33: 289,

1971. 35. Cline, M. J. and Warner, N. L., J. Immunol., 108: 339, 1972. 36. Tigelaar, R. E., Vaz, N. M., and Ovary, Z. J., J. Immunol., 106:

661, t971. 37. Keller, R., J. Exp. Med., 138: 625, 1973. 38. Evans, R. and Alexander, P., Nature, 236: 168, 1972. 39. Hibbs, J. B., Jr., Lambert, L. H., Jr., and Remington, J. S.,

Science, 177." 998, 1972. 40. Scott, M. T., Cell. Immunol., 5: 459, 1972. 41. Holtermann, 0. A., Klein, E., and Casale, G. P., Cell. Immunol.,

9: 339, 1973. 42. Steinman, R. M. and Cohn, Z. A., J. Exp. Med., 137: 1142, 1973. 43. Walker, W. S., Immunology, 26: 1025, 1974. 44. Mocarelli, P., Palmer, J., and Defendi, V., Immunol. Commun., 2:

441, 1973.

at Em

ory University H

ealth Sciences Library on D

ecember 9, 2016

http://ww

w.jim

munol.org/

Dow

nloaded from