Influenza viral glycoproteins induce cell-mediated cytotoxicity by an interferon-independent mechanism

CELLULAR

97, 102- 109 ( 1986)

IMMUNOLOGY

Influenza Viral Glycoproteins Induce Cell-Mediated Cytotoxicity an Interferon-Independent Mechanism

by

D. JIT S. ARORA AND DOMINIC M. JUSTEWICZ Institut Armand-Frappier,

University of Quebec, Lava/-des-Rapides,

Lava/, Quebec H7N 423, Canada

ReceivedJuly IS, 1985;acceptedSeptember13. 1985 Human natural killer (NK) cells exposed to the influenza surface antigen neuraminidase (NA) show high cytotoxic activity, as evaluated using chromium-labeled K562 target cells in a standard overnight cytotoxicity assay. The role of interferon (IFN) in the stimulation of NK cells was examined by using three separateapproaches.(i) The use of appropriate antibodies to check IFNand NA-specific cell-mediated cytotoxicity (CMC) stimulation showed that antibodies to IFN (-a, -y) did not alter NA-induced CMC, and vice versa. (ii) The treatment of NK cells with actinomycin D, before or after stimulation with IFN and NA revealed that only EN-induced CMC was inhibited (50 to 100%).However, NK cells that were stimulated with NA before their exposure to actinomycin D became susceptibleto stimulation by IFN. (iii) The interaction kinetics between IFN and NA demonstrated the presenceof two mechanisms of CMC stimulation. Taken together, the resultsclearly show that stimulation of CMC by a viral component is effectedthrough an EN-independent pathway, and that this mechanism is probably followed by IFN under certain conditions. 0 1986 Academic Res, Inc.

INTRODUCTION

A nonspecific resistance system to virus infection NK activity (or CMC), in addition to known specific immune responses[for review seeRef. (l)], has been implicated in the initial protection (2- 13). NK cells may impede the progress of the infection by lysing virus-infected cells or by producing IFN. Recently, we (14, 15) and others (2-6, 10, 13, 16-21) have reported that CMC can be enhanced by whole virions as well as by specific viral proteins. It has been shown that CMC activity due to whole virions is associated with IFN release (22-25), although NK cytotoxicity and IFN production are not temporally correlated (2,4-6, 10, 12, 16- 18). Viral antigens induce CMC without the concomitant releaseof IFN (2, 3, 5, 16, 17). Although the findings with the influenza virus glycoproteins, hemagglutinin and NA, implied that they could account for a large part of the CMC induced by whole vii-ions, the mechanism of enhancement has not been clearly established (14). We now present evidence that influenza viral antigen-induced CMC activity is an IFN-independent process, which does not require de novo RNA synthesis, and which can be clearly segregatedfrom the IFN-dependent route. MATERIALS

AND METHODS

Isolation and purification of NA. Influenza virus A/Port Chalmers/3/73 (H3N2), purified as described previously (14), was treated with SDS. The NA was separated 102 0008-8749/86 $3.00 Copyight Q 1986 by Academic F’res, Inc. AlI rights of reproduction in any form rcsrved.

VIRAL

ANTIGEN-INDUCED

CMC

IS IFN INDEPENDENT

103

by electrophoresis on cellulose acetate paper, and was further purified by affinity chromatography with appropriate adsorbents as already described (26). Protein determination. Protein concentration was determined by the Lowry et al. technique (27). Preparation ofhuman lymphocytes. Peripheral blood lymphocytes (PBL) were sep arated from heparinized venous blood by density gradient centrifugation on FicollPaque (Pharmacia Fine Chemicals, Uppsala, Sweden) as outlined (14). Washed PBL were adjusted to a concentration of 2 X lo6 cells/ml with RPMI-1640 (Flow Laboratories, McLean, Va.) supplemented with 25 mM N-2-hydroxyethylpiperazine-N’-2ethanesulfonic acid, 4 mM L-glutamine, 10%heat-inactivated fetal bovine serum, and antibiotics (referred to as medium). PBL were stored at 37°C overnight prior to use. Target cells. The human erythroleukemia cell line K562 was used as targets for NK activity and were maintained in suspensionculture in medium. K562 cells were labeled with Na25’CrQ4(sp act, 2mCi/mg; Frosst, Kirkland, Quebec, Canada) at 37°C for 45 min, washed twice, and adjusted to a concentration of 1 X lo5 cells/ml (14). Cytotoxicity assay. NK activity was measured by a direct overnight “Cr release assayas described (14). Briefly, 1 X lo5 chromium-labeled target cells in medium in triplicate cultures of 50 and 100 ~1, respectively. Assays were conducted in 96-well, round-bottom microtiter plates (Linbro, New Haven, Conn.) to give an effecter/target cell (E/T) ratio of 10:1. After centrifugation for 2 min at 2OOg,the plates were incubated at 37°C in a humid 5% CO*, 95% air atmosphere for 16 to 18 hr unless otherwise specified. At the completion of the incubation, the plates were centrifuged for 10 min at 2008, and the supernatant was harvested by the Skatron Supematant Collection System (Skatron, Sterling, Va.) and counted on a gamma counter. The percentage cytotoxicity (referred to as CMC or NK activity) was calculated according to the equation: % specific “Cr release= [cpm (test) - cpm (spontaneous)]/[cpm (maximum) - cpm (spontaneous)] X 100 where cpm is counts per minute. The cpm (spontaneous) was determined by counting the radioactivity of the supematant of the centrifuged labeled target cell suspension, whereas the cpm (maximum) was the radioactivity of resuspended labeled target cells. The results presented are the means of three independent observations (+ standard deviation). Various E/T ratios (20:1, lO:l, 5:1, and 2.5:1) were tested in our prior reports ( 14, 15), and the results of CMC stimulation by viral antigens and IFN-(u using the E/T ratios of 10:1 were consistently reproducible among donors. AX-induced CMC The stimulation of CMC by NA was effected by the following two types of assays:(i) aliquots of purified NA at a concentration of 12 X 10-l bg/ml in medium were preincubated with 2 X lo6 effector cells/ml at 37°C for 1 hr before the addition of the target cells; or (ii) NA was added directly to the test system as described earlier ( 14). An E/T ratio of 10:1 was used in all experiments. We previously tested various dosesof NA ( 11 X 10m7to 11 X 10-l pg/ml) in a dose-responseanalysis for their potential to enhance CMC (at E/T 10:l), and showed that 11 X 10-l kg/ml of NA provided optimal stimulation of CMC ( 14). The effector cells responsible for viral antigen-induced CMC (and IFN-induced CMC) were previously shown to be NK cells by a number of criteria: (i) the effector cells shared the cell surface phenotype of NK cells since they were depleted of activity by treatment with the anti-NK antibody Leu-7 [HNK-I; Refs. (28, 29)] plus complement [Ref. ( 15); Justewicz et al., unpublished observations]; (ii) the effector cells were found in Percoll gradient (30) fractions enriched for large granular lymphocytes (Jus-

104

ARORA AND JUSTEWICZ

tewicz et al., unpublished observations); (iii) the effector cells showed cytotoxicity against the NK-sensitive K562 target cell line (14, 15). We also found that the NK activity of the effector cells could be boosted by pretreatment with IFN-a (this report), another criterion of NK cells (8, 3 1, 32). Inhibition of NA-induced CMC by monospec$c anti-NA antibodies. The effect of monospecific rabbit antiserum [preparation and characterization described in Ref. (26)] on NA-induced CMC was determined by the preincubation of the antigen suspension at 37°C for 1 hr before the addition of effector and target cells. Samples of 10 ~1 of a IO-’ dilution of antiserum were used as described ( 14). Since the presence of the Fc portion of the immunoglobulin molecule might nonspecifically modulate the stimulation of NK activity under the experimental conditions used, the validity of the results reported with monospecific antiserum was previously examined by using the F(ab’)z fragment isolated from the antiserum (14). Similar levels of inhibition were obtained, which confirmed the results with the monospecific antiserum. ZFN. Human leucocyte IFN-CYwas a generous gift from our colleague Dr. J.-M. Dupuy (the original source was Dr. J. L. Virelizier, Hopital des Enfants Malades, Paris). Preliminary experiments were performed to determine the optimal concentration of IFN-a that would maximally stimulate CMC. Among the concentrations of IFN tested, ranging from 0.05 to 50, 100 and 1000 III/ml, the latter concentration was found to be the optimal dose. Neutralization oflFN. Neutralization of IFN activity was carried out by using sheep antisera specific to human IFN-a (Enzo Biochemical, New York, N.Y.) and IFN-7 (Interferon Sciences, New Brunswick, N.J.) as follows. (i) The neutralization of any endogenous IFN releasedduring the stimulation of CMC was achieved by the addition of antisera (sufficient to neutralize 10,000 III/ml IFN) to the test system containing the incubation mixture of effector and target cells. (ii) The neutralization of exogenous IFN-cu was effected by the preincubation of the antigen with its homologous antisera at 37°C for 1 hr before the addition of effector and target cells. Pretreatmentof efictor cells with exogenousIFN. NK effector cells were incubated at a concentration of 2 X 106/ml in medium containing 1000 II-J/ml of IFN-(I! at 37°C for 4 hr. Cells were washed twice before use. This procedure was reported previously ( 17, 3 1) to give maximal enhancement of NK activity by IFN-a, and was confirmed in our preliminary observations using two concentrations of IFN-a (100 and 1000 IU/ml). Pretreatmentof efictor cells with actinomycin D. NK effector cells were incubated at a concentration of 2 X 106/ml in medium containing 0.5 pg/ml actinomycin D (an RNA synthesis inhibitor; Sigma Chemical Co., St. Louis, MO.) at 37°C for 1 hr. Cells were washed three times before use. Earlier reports (2, 10, 31) used different concentrations of actinomycin D tanging from 0.0 1 to 1.O&ml. Our preliminary observations showed that 0.5 &ml was the optimal dose to inhibit CMC. RESULTS AND DISCUSSION

Effect of anti-IFN and anti-NA antibodies to checkspec$c CMC stimulation. Previously, we reported that the stimulation of CMC by the influenza virus glycoproteins accounts for a large part of the CMC induced by the whole virion ( 14). Moreover, the CMC stimulation due to each glycoprotein was independent of the presence of the

VIRAL

ANTIGEN-INDUCED

CMC

IS IFN INDEPENDENT

105

other, as no additive effect was observed when a l/ 1 mixture was used ( 14). Purified influenza NA was chosen in this study as the viral antigen because proteins from several N 1 and N2 subtypes have already been shown to stimulate NK activity (30 to 50% above normal activity) and becauseits neutralization by monospecific and their F(ab’), fragments ablates the enhanced CMC activity [Ref. (14); D. J. S. Arora and D. M. Justewicz, Abstr. Annu. Meet. Amer. Sot. Microbial. 1985. E46, p. 821. In examining the mechanism of CMC enhancement by influenza viral antigens, it was argued that if the presence of IFN was a prerequisite for mediating the stimulation, addition of the appropriate anti-IFN antibody should abrogate viral antigen-induced CMC, as suggestedearlier (24, 25, 31). Results showed that whereas NA stimulated CMC in a manner comparable to the performance of IFN-a (1000 IU/ml), the enhancement was specific because the addition of anti-IFN (-a, -y) antibodies had no suppressive effect on NA-induced CMC, but it was totally neutralized with its homologous anti-NA antibody (Table 1). Similarly, IFN-induced CMC was neutralized only by anti-IFN antibody while anti-NA had no effect. These results suggestedthat extracellular IFN release was not necessaryfor the viral antigen to exert its effect(s) on NK cells, and provided evidence that the two stimulators, IFN and NA, were capable of inducing specific CMCs through two independent mechanisms. However, results did not exclude the possibility that the intracellular generation of IFN may have been adequate for these NA-mediated effect(s) to occur. This latter aspect was examined next. Effect of actinomycin D on specijic CMC stimulation. Since CMC stimulation by IFN is related to RNA synthesis (3 1,32), it was considered appropriate to test whether or not RNA synthesis underlies the two mechanisms. PBL were treated with actinomycin D, known to inhibit irreversibly de novo RNA synthesis without decreasing normal NK activity (3 1, 32), and then exposed to IFN or NA in a standard CMC assay system. Since it has been reported (33) that the prolonged incubation of NK effector cells at 37°C reduced their capacity to generate IFN without altering their cytotoxic activity, the PBL used in the present study were kept at 37’C overnight before actinomycin D treatment. Results demonstrated that IFN-induced CMC was inhibited by 50%, whereas the inhibitor had little effect on NA-induced CMC (Table 2). Experiments were repeated by using antigen-prestimulated PBL in an actinomycin D-treated system (PADS) with either IFN or NA as antigen.Results showed that CMC was effectively blocked in IFN-PADS, confirming the earlier finding (31). On the other hand, NA-induced CMC was not affected by treatment with the inhibitor, suggesting that it is unlikely that RNA synthesis is involved in the early stages of NA-induced CMC. This observation provided further evidence that NAinduced CMC was different from IFNdependent CMC. In a parallel set of experiments, the effect of an additional dose of IFN or NA on IFN- or NA-PADS was tested with the following results. No increase in CMC was observed when IFN was added to IFNPADS. NA enhanced CMC by 40% when the antigen was added to NA-PADS. Further enhancement of CMC by NA could occur if the initial concentration of NA used did not saturate the CMC assaysystem. It was interesting to note that an increase of 33% in CMC was observed when IFN was added to NA-PADS. These results demonstrated that IFN was able to boost CMC when the IFNdependent mechanism was blocked. A probable explanation for this could be that the NA-receptor interaction provided a favorable condition for IFN to follow the pathway that did not necessitate de novo RNA synthesis (IFN-independent pathway). The addition of NA to IFN-PADS pro-

ARORA AND JUSTEWICZ TABLE 1 Effects of Anti-IFN and Anti-NA Antibodies on Stimulated CMC

Donor treatment RB Control” IFN-a’ + anti-IFN-cud + anti-IFN-r

5%Specific r’Cr release (E/T 1O:l)

18 + 0.3” 31 f 0.9 11 * 0.7 31 + 0.8

DM Control + anti-IFN-cY + anti-RN-y

44 f 0.4 41 f 1.0 40 f 1.1

IFN-(u + anti-FIG + anti-NAe

49 + 0.6 35 * 0.5 48 + 0.7

NAI + anti-IPN-a + anti-IPN-y

50 * 1.0 53 f 0.5 48 + 0.7

JB Control NA + anti-NA

37 f 1.9 47 + 1.6 35 f 2.6

a Normal NK activity of PBL. In these experiments PBL were obtained from three different donors, from a larger sample of eight healthy male donors (20 to 45 years of age) [Refs. (14, 15)]. Their PBL, tested from 8 to 13 times, provide consistent and reproducible levels of CMC stimulation. b Mean value of triplicate samples f standard deviation. ’ Human leucocyte IFN-a was used at 1000 II-l/ml. d Sheep antisera to IPN-ol, and -y, sufficient to neutralize 10,000 II-l/ml IPN, were incubated with IFN or NA for 1 hr at 37°C before effector and target cells were added. ’ Rabbit monospecific anti-NA antibody was added to NA or IPN as in d. fPurified NA was used at 12 X 10-l &ml.

duced a twofold increase in CMC activity compared to IFN-induced CMC. This suggested that the effect(s) of NA antigen in the initial stagesof stimulation followed an IFN-independent mechanism and RNA synthesis was not required. This possibility is compatible with the results obtained from the interaction experiments described next. Interaction kinetics betweenNA and IFN. The logic behind these experiments was that if indeed there were two independent mechanisms present and they were operating independently to stimulate CMC, then the addition of NA to the IPN-initiated CMC assaysystem should result in an increase in stimulated CMC, and vice versa. On the other hand, no tiuther increase in CMC would suggestsome interaction between the two stimulants. The following experiments were performed: PBL were incubated for 4 hr with a saturating dose of IFN (1000 IU/ml, as described in Table 2) and NA

VIRAL ANTIGEN-INDUCED

CMC IS IFN INDEPENDENT

107

TABLE 2 Effects of Actinomycin D on Stimulated CMC Treatment

% Specific 5’Cr release

ControY +lFNb +NA=

23 f 1.3 33 f 1.2 33 f 1.0

Control-PADSd +1FN +NA

23 + 0.9 26 + 1.1 32 + 0.8

IFN-PADS +1FN +NA

19 + 1.6 20 + 1.5 36 + 0.03

NA-PADS/ fIFN +NA

31 kO.5 37 + 1.7 39 f 2.4

,JNormal NK activity of PBL after a 16-hr incubation period (E/T IO:I). b IFN-(Y (1000 IU/ml) was added at the beginning of the CMC assay. ’ NA (12 X 10-l &ml) was added as in b. d PBL were incubated with medium for 4 hr, and treated with actinomycin D (0.5 &ml) for 1 hr at 37°C before they were tested in the CMC assaysystem (PADS). ePBL were treated as in d, but with IFN-LU(1000 W/ml) (IFN-PADS). fSame as in d, but with NA (12 X 10-l &ml) (NA-PADS).

separately. To these stimulator initiated CMC systems (ICMCS) an additional dose of either IFN or NA was added, and the assayswere allowed to run to completion as described in Table 3. Results showed that the kinetics of IFN-ICMCS were not affected by an additional dose of IFN, indicating that the assaywas IFN saturated. However, the addition of NA to IFN-ICMCS produced a small but significant increase. This observation is similar to that shown in Tables 1 and 2. The NA-ICMCS was further enhanced with the addition of NA. The 16% increase in CMC again indicated that the system was probably not saturated with NA. When IFN was added to NA-ICMCS, an enhancement of 23% was observed. These results provided further evidence that the NA-primed CMC system allowed IFN to stimulate CMC. However, whether this was IFN dependent or IFN independent was not established by the present results. Taken together, the results obtained provide evidence for two distinct and separate mechanisms of CMC stimulation. The data further show that (i) enhancement by one stimulant is unaffected by the presence of heterologous antibodies, (ii)de novo RNA synthesis is related to only the IFN-dependent pathway, (iii) the viral NA antigen can follow a different mechanism to stimulate CMC under the experimental conditions tested, and (iv) once a given mechanism of CMC is initiated by one stimulant, an alternate mechanism of CMC remains available to the other stimulant. Considering that immediate immune responsesare important to the cell in virus infection, this alternate mechanism may serve as an additional security measure. The mechanism of enhancement of CMC activity can occur in at least two different ways (34): (i) by increasing the proportion of active NK cells, known as effector cell recruitment; or (ii) by accelerating the kinetics of lysis of active NK cells, thereby

108

ARORA AND JUSTEWICZ TABLE 3 Interaction Kinetics of CMC Stimulation between IFN and NA 70Specific 51Crrelease Treatment Control* +lFNc +NA

2 hr” 5 k 0.2 -

4 hr -

8 hr

16 hr

I1 kO.1 23 + 0.5 21 f 0.2

44 + 0.4 51 + 1.4 50 +- 0.3

IFN-ICMCSd +1FNe +NA

10 + 0.6 -

14 f 0.2 -

23 + 1.0 21 + 0.2 27 If: 1.1

49 f 0.6 49 + 0.7 52 + 0.7

NA-ICMCS’ +1BN* +NA

10 + 0.1 -

16 f 1.0 -

23 zk 0.3 23 zk 0.6 26 z!z0.8

50 zk 1.0 60 AZ1.8 55 xk 0.5

’ Incubation time of the assaysystem. ’ Normal NK activity of PBL (E/T 10:1). ’ To a 8- and 16-hr CMC assaysystem, IFN-(Y ( 1000 III/ml) or NA ( 12 X 10-l j&ml) was added after 4 hr of incubation. d IFN-(u (1000 IU/ml) was added at the beginning of the CMC assay(IFN-ICMCS). ‘Same as in c, but to IFN-ICMCS. /Same as in d, but with NA (12 X 10-l &ml) (NA-ICMCS). *Same as in c, but to NA-ICMCS.

facilitating the recycling of active NK cells. These alternative but often coexisting mechanisms of CMC enhancement are responsible for the stimulation by IFN (32, 34). The type of target cell, used in the CMC assay,can play an important role in differentiating NK cell proliferation from NK cell activation. For example, Timonen et al. (34) reported that the mechanism of NK cell activity stimulation by IFN against the suspension-grown target, the erythroleukemia cell line K562, was restricted to effector cell activation, whereasCMC enhancement toward anchorage-dependenttarget cells was effected by both mechanisms described above. Since influenza NA could directly stimulate CMC, using K562 target cells, it is probable that the viral glycoprotein acted upon a preexisting population of NK cells. This and other possibilities remain to be tested to elucidate the manner by which NA effects CMC stimulation. It is known that IFN mediates the stimulation of NK cells through interaction with the cell surface [for review seeRef. (35)]. Briefly, the IFN-receptor interaction induces a series of physiologic responses inside the cell (35), followed by the synthesis of a requisite protein(s) (3 1, 34) which is probably expressedon the cell surface, resulting in the activation of NK cells (36-39). The early stagesof this mechanism require de novo RNA synthesis. However, our results clearly show that NA-induced CMC bypassesthis requirement but probably shares the remaining pathway described above. Furthermore, IFN does show the capacity to induce CMC without requiring the participation of de novo RNA synthesis. We speculate that IFN follows an alternate pathway provided the NK cell is initiated by NA or a similar molecule. This shunt mechanism probably provides an alternate route of stimulation of the nonspecific response in viral infection, as mediated by the NK cell.

VIRAL

ANTIGEN-INDUCED

CMC

IS IFN INDEPENDENT

109

ACKNOWLEDGMENT This investigation was supported by the Natural Sciencesand Engineering Research Council of Canada, Grant A8371.

REFERENCES 1. Couch, R. B., and Kasel, J. A., Annu. Rev. Microbial. 37, 529, 1983. 2. Bishop, G. A., Glorioso, J. C., and Schwartz, S. A., J. Immunol. 131, 1849, 1983. 3. Bishop, G. A., Marlin, S. D., Schwartz, S. A., and Glorioso, J. C., J. Immunol. 133,2206, 1984. 4. Bukowski, J. F., Woda, B. A., Habu, S., Okumura, K., and Welsh, R. M., J. Viral. 52, 119, 1984. 5. Casali, P., Sissons,J. G. P., Buchmeir, M. J., and Oldstone, M. B. A., J. Exp. Med. 154, 840, 1981. 6. Ching, C., and Lopez, C., Infect. Immun. 26,48, 1979. 7. Ennis, F. A., Meager, A., Beane, A. S., Yi-Hua, Q., Riley, D., Schwartz, G., Schild, G. C., and Rook, A. H., Lancet ii, 891, 1981. 8. Fitzgerald, P. A., Mendelsohn, M., and Lopez, C., J. Immunol. 134,2666, 1985. 9. Herberman, R. B., and Ortaldo, J. R., Science (Washington, DC.) 214,24, 1981. 10. Kurane, I., Hebblewaite, D., Brandt, W. E., and Ennis, F. A., J. Viral. 52,223, 1984. 11. Lopez, C., In “The Immunobiology of Herpes Simplex Virus Infections” (C. Lopez and B. T. Rouse, Eds.), pp. 45-69, CRC Press,Inc., Boca Raton, Fla., 1984. 12. Minato, N., Bloom, B. R., Jones, C., Holland, J., and Reid, L. M., J. Exp. Med. 149, 1117, 1979. 13. Welsh, R. M., Curr. Top. Microbial. Immunol. 92, 83, 1981. 14. Arora, D. J. S., Houde, M., Justewicz, D. M., and Mandeville, R., J. Viral. 52, 839, 1984. 15. Justewicz, D. M., Lecomte, J., and Mandeville, R., J. Infect. Dis. 150, 348, 1984. 16. Fitzgerald, P. A., Evans, R., Kirkpatrick, D., and Lopez, C., J. Immunol. 130, 1663, 1983. 17. Fitzgerald, P. A., von Wussow, P., and Lopez, C., J. Immunol. 129, 818, 1982. 18. Habu, S., Akamatsu, K. I., Tamaoki, N., and Okumura, K., J. Immunol. 133, 2743, 1984. 19. Yasukawa, M., and Zarling, M. J., J. Immunol. 131, 2011, 1983. 20. Alsheikhly, A., Orvell, C., Harfast, B., Andersson, T., Perlmann, P., and Norrby, E., Stand. J. Immunol. 17, 129, 1983.

2 1. Harfast, B., Orvell, C., Alsheikhly, A., Andersson, T., Perlmann, P., and Nonby, E., Scund. J. Immunol. 11, 391, 1980.

22. Djeu, J. Y., Stocks, N., Zoon, K., Stanton, G. J., Timonen, T., and Herberman, R. B., J. Exp. Med. 156, 1222, 1982.

23. 24. 25. 26. 27. 28. 29.

Ennis, F. A., and Meager, A., J. Exp. Med. 154, 1279, 1981. Santoli, D., and Koprowski, H., Immunol. Rev. 44, 125, 1979. Trinchieri, G., Santoli, D., Dee, R. R., and Knowles, B. B., J. Exp. Med. 147, 1299, 1978. Arora, D. J. S., Hill-Schubert, J., and Vincent, L., Cunad. J. Microbial. 26, 243, 1980. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J., J. Biol. Chem. 193,265, 1951. Abo, T., Cooper, M. D., and Balch, C. M., J. Immunol. 129, 1752, 1982. Lanier, L. L., Le, A. M., Phillips, J. H., Warner, N. L., and Babcock, G. F., J. Immunol. 131, 1789, 1983.

30. Pertoft, H., Hirstenstein, M., and Kagedal, L., In “Cell Populations, Methodological Surveys (B) Biochemistry, Ix” (E. Reid, Eds.), pp. 67-80, Ellis Horwood Ltd., Chichester, West Sussex,U.K., 1979. 31. Ortaldo, J. R., Phillips, W., Wasserman, K., and Herberman, R. B., J. Immunol. 125, 1839, 1980. 32. Bishop, G. A., and Schwartz, S. A., Clin. Immunol. Immunopathol. 25, 374, 1982. 33. Copeland, C. S., Koren, H. S., and Jensen, P. J., Cell. Immunol. 62, 220, 1981. 34. Timonen, T., Ortaldo, J. R., and Herberman, R. B., J. Immunol. 128,2514, 1982. 35. Raziuddin, A., Sarkar, F. H., Dutkowski, R., Shulman, L., Ruddle, F. H., and Gupta, S. L., Proc. N&l. Acad. Sci. USA 81,5504,

36. 37. 38. 39.

1984,

Hiserodt, J. C., B&van, L. J., and Targan, S. R., J. Immunol. Hiserodt, J. C., Britvan, L. J., and Targan, S. R., J. Immunol. Hiserodt, J. C., B&van, L. J., and Targan, S. R., J. Immunol. Newman, R. A., Warner, J. F., and Dennert, G., J. Immunol.

131,2705, 131,2710, 129, 1782, 133, 1841,

1983a. 1983b. 1982. 1984.