CELLULAR
IMMUNOLOGY
113,1-9 (1988)
Enhianced Interferon Production and Lymphokine-Activated Cytotoxicity of Human Placental Cells’ TERRYW.CHIN,~BONNIE J. ANK,SHEILAR. STROM,AND E. RICHARDSTIEHM~ Department of Pediatrics, UCLA School of Medicine, Los Angeles, California
90024
Received December 22, 1986; accepted November 24, 1987
The immunologic competence of human placental mononuclear cells was compared to that ofaduli. and cord blood mononuclear cells. Mononuclear cells were isolated from fresh placentas by digestion with collagenase and DNase, followed by Ficoll-Hypaque and discontinuous Perco11separation. Placental cells incubated with phytohemagglutinin (PHA) synthesized significantly more interferon-y (IFN-7) at 2 days (29 f 5.5 IU/ml) and 5 days (46 + 8.5 W/ml) than PHA-amctivatedcord cells (3.6 f 0.6 IU/ml at 2 days and 2.7 + 0.7 IU/ml at 5 days) but lessthan PHA-activated adult cells (8 1 i 20 IU/ml at 2 days and 270 ? 161 IU/ml at 5 days). Placental and adult cells, but not cord cells, also synthesized significant quantities of IFN-y following incubation with interleukin-2 (IL-2). There was synergism between IL-2 and PHA activation for IFN-7 production for some cord samples. After a 5- to 7-day incubation with IL-2, the lymphocyte-activated killer (LAK cell) cytotoxicity of placental cells (measured in a 3-hr chromium-release assay at an E:T ratio of 4O:l) was enhanced 13-fold against K562 target cells (6 + 2% to 77 f 4%) compared to a 4-fold increase in cord cells (16 -+ 4% to 68 f 3%) and a 2-fold increase in normal adult cells (35 f 4% to 65 + 3%). Against the natural killer (NK)-resistant Raji target, placental cells increased their LAK cytotoxic activity (3 + 1%to 59 f 7%) compared to a 7-fold increase with cord cells (6 f 1%to 43 + 3%) and a 3-fold increase with adult cells (11 + 2% to 38 + 4%). A notable degree of cytotoxic activity in the absence of IL-2 against Molt targets was noted in 11 of 14 (79%) placental cell samples at 5 days. Only 10 of 24 (42%) adult and 17 of 37 (40%) cord samples showed spontaneous cytotoxic activity equal to or greater than 10%.Some placental samples actually showed an increase in cytotoxic activity when incubated without IL-2. The ability of placental cells to produce significant levels of IFN-y, to develop considerable LAK activity, and to maintain or develop cytotoxic activity in the absenceof IL-2 suggesisa vigorous, active immune system of the placenta compared to the relatively dormant immune system of the neonate. These observations suggestthat placental cells may have a primary role in fetal defense. 0 1988 Academic Press, inc.
INTRODUCTION Several immunologic functions can be attributed to the placenta. It must protect the fetus from maternal rejection, intercept microbial agents present in the maternal circulation, and limit accessto circulating antigens that could activate the fetal immune system. Immunocompetent cells isolated from human as well as murine placental tiss,ue( 1,2) have been shown to eliminate antibodies to fetal antigens by forma’ Supponted in part by NIH Grants HD-09800, AI-07008, and AI-15332. ’ Now at Miller Childrens Hospital, University of California, Irvine. 3To whom requests for reprints should be addressed.
0008-8749/88 $3.00 Copyright 0 1988 by Academic Press, Inc. All rights of reproduction in any form reserved
2
CHIN
ET AL.
tion of immune complexes (3) synthesize various lymphokines such as interferons and interleukin- 1 (4,5), demonstrate immunoregulatory suppressive activity (6), and phagocytize and kill various microorganisms (7). In our previous studies of the phenotypic and functional capacities of immune cells from the human placenta (S), we showed that despite the presence of large quantities of Leu-7+ and Leu- 1I+ cells, the natural killer (NK) cell activity of the placenta was considerably lessthan that of either cord or maternal mononuclear cells. Other cytotoxic functions of placenta (and cord blood), including antibody-dependent cellular cytotoxicity (ADCC) and lectin-dependent cellular cytotoxicity (LDCC), were also lessthan those of adult cells. By contrast, we have recently reported that lymphokine (interleukin-2)-activated killer cell (LAK) cytotoxicity in cord blood was equivalent to and often greater than that in adult cells (9). In addition, we have found that phytohemagglutinin (PHA)-activated cord lymphocytes synthesize significantly less immune or interferon-y (IFN-7) than do adult cells ( 11). In this paper, we examine some properties of activated placental mononuclear cells compared to those of neonatal cord cells. We noted enhanced PHA-induced IFN--r production and interleukin-2-induced cytotoxic function, suggesting a distinct placental immune system that has unique functional capabilities when compared to the neonate’s immune system. METHODS AND MATERIALS Efictor cells. Fourteen human placentas were obtained from uncomplicated term pregnancies by either vaginal delivery or Cesarian section. Within 5 hr of delivery, mononuclear cell preparations were obtained as previously described (8). Briefly, placental tissue rich in chorionic villi was minced and washed with 0.9% sterile saline. The tissue pieces were then digested with 150 units/ml collagenase (Type II, Sigma) and 60 units/ml DNase (Type 1, Sigma) in RPM1 1640 medium supplemented with 2% fetal calf serum (FCS) (M. A. Bioproducts), 100 u/ml penicillin, 100 pg/ml streptomycin, and 0.25 pg/ml amphotericin B. After 3 hr at 37°C the mixture was decanted through fine-mesh gauzeand mononuclear cells were obtained using Ficoll-Hypaque and discontinuous Percoll gradients. The resultant band was washed and resuspended in RPM1 1640 medium supplemented with 10%AB serum. Approximately 1O6cells were obtained per gram of placental tissue, with greater than 80% viability by trypan blue exclusion. This procedure yielded a preparation rich in lymphocytes (60-70%) and monocytes ( 15-30%) and about 8% granulocytes and trophoblastic cells. Chromosomal analysis of PHA-activated placental cells from male infants indicated they were of male phenotype, demonstrating the fetal origin of the immune cells isolated. Whole mononuclear cell preparations were also obtained by Ficoll-Hypaque density centrifugation from 37 cord bloods (usually within 5 hr of delivery) and from 25 normal adult controls (9). Lymphokine-activated killer cellular cytotoxicity. Mononuclear cell preparations at a concentration of 1 X lo6 cells/ml were incubated for 5-7 days at 37°C in 5% CO* in the presence of 50-100 units/ml of recombinant interleukin-2 (IL-2) (Amgen, Thousand Oaks, CA, or Cetus, Emeryville, CA). Control samples were incubated without IL-2. After centrifugation at 400g for 10 min, the supernatants were removed to assayfor the presenceof IFN-7. The cells were resuspended in RPM1 1640 medium supplemented with 10% human AB serum, recounted, and assayedfor cytotoxic activity.
IFN AND LAK CYTOTOXICITY
OF HUMAN
PLACENTAL
CELLS
TABLE 1 Interferon-y Production (W/ml f SEM) of Cord and Placental Mononuclear Cells following Activation with Phytohemagglutinin (PHA) or Interleukin-2 (IL-2)” Activating agent Supernatants A. Two Days Cord Placenta B. Five Days Adult Cord Placenta
None
IL-2
0.2 f 0.04 (24)b 0.5 kO.1 (16)
8.5+ 4.1(11) 6.9 f 1.6 (9)
0.7kO.l (19) 0.5kO.l (12) 0.8 f 0.2 (9)
42 f 17 (19) 11 f 7.0(12)’ 30 f 4.9 (9)
PHA 3.6 + 29 f
0.6 (22)’ 5.5(15)d
270 f 161 (17) 2.7 + 0.7 (12)’ 46 + 8.5(8)
’ Mononuclear cells at 1 X lo6 cells/ml were cultured with either IL-2 (100 U/ml) or PHA (1:40) at 37’C for 2-3 or 5-7 days. After centrifuging at 400g for 10 min, the supematants were removed and interferony levels were determined by radioimmunoassay (Centocor). ’ The numbers in parenthesesrepresent the number of subjects studied. ’ Interferon llevel of cord cell supernatants was significantly lessthan that of adult and placental supematants (P i 0.00 1). d Interferon level of placental cell supernatants was significantly lessthan that of adult cells (P < 0.01). eInterferon level of cord cell supematants was significantly lessthan that of placental cells (P < 0.001).
Cytotoxic assays. Baseline natural killer (NK) cell and LAK cell activities were assayedin a.3-hr chromium-release assayagainst labeled K562, Molt 4f, and Raji cell targets. Specifically, 4 X 1O5effector cells were added to 1O4target cells in RPM1 1640 medium with 10%AB serum (effector (E):target (T) cell ratio 40: 1) in triplicate in 96well, U-bottom microtiter plates. The plates were then subjected to a final centrifugation at 450~:for 10 min, and the supernatants were counted on a gamma-scintillation counter. The percentage specific lysis (SL) was calculated as follows: SL = [(experimental cpm - spontaneous cpm)/ (total cpm - spontaneous cpm)] X 100. Interfero,rl assay. Whole mononuclear cells were cultured with IL-2 at a concentration of 50-100 U/ml, phytohemagglutinin (PHA-M, Difco, 1:40), or both IL-2 and PHA at 37°C in 5% CO2 for 2-3 or 5-7 days. Supernatants were then removed and stored in aliquots at -70°C until the interferon assayswere run. Interferon-y levels were determined by radioimmunoassay (Centocor) (10). The validity of the assaywas confirmed by simultaneous bioassay (results not shown). RESULTS Interferon Production Table 1 shows that placental mononuclear cells following activation with PHA are capable of producing significant quantities of IFN-7 (29 f 5.5 IU/ml at 2 days and 46 + 8.5 IU/ml at 5 days). These values are less than levels observed by similarly activated adult cells (8 1 + 20 IU/ml at 2 days and 270 k 161 IU/ml at 5 days) but significantly greater (P < 0.01) than the minimal quantities of IIN- produced by
4
CHIN ET AL. TABLE 2 Synergistic Effect of Phytohemagglutinin (PHA) and Interleukin-2 (IL-2) on Interferon-y Production (IU/ml + SEM) by Cord Mononuclear Cells“ Activating agent Cord sample Ja We He Sh Summary
None
IL-2
PHA
IL-2 + PHA
0.2 0.3 1.5 0.1
20 0.1 0.3 2.2
9.5 3.0 3.2 3.1
84 7.2 8.4 26
5.1 f 5
4.1 k 1.6
31 f 18
0 Mononuclear cells were cultured with or without IL-2 (100 U/ml) and PHA (1:40 dilution) under conditions described for Table 1. Supematants were removed after 5 days and interferon-y levels were determined by radioimmunoassay.
most cord cells. In 6 out of I5 (40%) placental samples the interferon level was less than 20 IU/ml after 2 days of stimulation compared to 2 out of 22 (9%) normal adult samples. After 5 days of incubation with PHA, 2 of 8 (25%) placental samples and 2 of 17 (12%) adult subjects made less than 20 IU/ml of 1FN-r. Equivalent levels of IFN-7 were synthesized during a 2-day incubation with IL-2 by placental (6.9 f 1.6 IU/ml) or adult cells (13 k 8.1 IU/ml). More IFN-7 was synthesized after a 5-day incubation (30 k 4.9 IU/ml for placenta and 42 t- 17 IU/ml for adults). Stimulation with both PHA and IL-2 for 5 days resulted in an additive effect with an interferon level of 85 + 16 II-J/ml for placental cells (n = 10) compared to 3 17 f 169 IU/ml for adult cells (n = 16). IL-2 alone was able to stimulate significant (i.e., >20 IU/ml) IFN production in only 2 of 11 (18%) cord samples after 2 days and 1 of 12 (8%) cord samples after 5 days. The mean value of IFN-y production after 5 days of IL-2 incubation of cord cells ( 11 k 7.0 IU/ml, II = 12) was statistically less(P < 0.00 1) than the value observed with placental cells (30 + 4.9 IU/ml, yt = 9). The combination of PHA and IL-2 resulted in minimal II+&-/ production by cord mononuclear cells after a 5-day incubation ( 13 f 6.8 II-J/ml, n = 12), not significantly different from that of IL-2 stimulation alone (11 f. 7.0 IU/ml). In 4 cords, a synergistic effect of PHA and IL-2 was observed, as shown in Table 2. All cord, placental, and adult cells treated with PHA or PHA + IL-2 had significant mitogenesis, as indicated by incorporation of tritiated thymidine (stimulation index > 10). No relationship was observed between the stimulation indices and IIN-7 production (data not shown). Also, there was no relationship between IFN-y production after a 5-day incubation with IL-2 and the level of cytotoxicity observed. Cytotoxicity
As observed previously (8, 9), baseline NK cell activity of cord cells against K562 was significantly less than that observed for adult cells (35 4 4% versus 16 f 4%, P < 0.01) (Fig. 1). Whole mononuclear cells from placentas exhibited even lower levels of cytotoxicity toward K562 target cells (6 f 2%, P < 0.001) than did cord cells. There
IFN AND LAK CYTOTOXICITY
OF HUMAN
PLACENTAL
CELLS
5
”
10090-
.
80-
.
.
70-
.$ . aj.GO- ‘
. .
$
[jo-
:
0 ::
.
. ‘,O-
30 - I:
8
:
l
:4”
+
.
+ o-
l
’
.
.
.
:
.
l .
.! Yp
:
12
’
Adult
Cord Placenta Day 0 WIthout I L-2
Adult
Cord Placenta Day 5 Without IL-2
Adult
Cord Plocento Day 5 With IL-2
FIG. I. Natural killer and lymphokine-activated killer cell cytotoxicity of normal human adult, human cord, and human placental mononuclear cells against K562 target cells. Cells (1 X IO’) were incubated with 100 units/ml of interleukin-2 (Cetus) for 5 days at 37°C. Cytotoxic activity was then assayedin a 3-hr chromium-relseaseassay at an effector (E): target (T) cell ratio of 40: 1. Baseline NK activity was obtained at Day 0 at the same E:T ratio. Results are expressed as the mean percentage of specific lysis f standard error.
was no significant difference between cord and placental cells at Day 0. After a 5- to 7-day incubation in the absenceof IL-2, cord and adult cells have decreasedcytotoxic activity (most lost it completely). By contrast, half of the placental cell preparations retained or increased their cytotoxicity (mean 6 f 2% on Day 0 compared to 14 f 3% on Day 5). After activation with IL-2 for 5-7 days, the cytotoxicity of placental cells was enhanced 13-fold against K562 target (to 77 f 4%) compared with a 4-fold increase in cord cells (to 68 + 3%) and a 2-fold increase in normal adult cells (to 65 + 3%). The: greater percentage enhancement observed for placental cells was due to their lower baseline NK cell activity, rather than to a significantly higher postincubation cytotoxicity. The retention of some cytotoxic activity in the absence of IL-2 was noted against Molt targets for all three cell populations (Table 3). Of 14 placental cell samples, 11 (79%) had cytotoxicity equal to or greater than 10% after a 5-day incubation, compared to IO of 24 (42%) adult and 17 of 37 (46%) cord samples. Adult cells incubated with IL-2 for 5-7 days increased their cytotoxic activity 2-fold (36 +_4% to 63 + 5%) against Molt targets, whereas cord cells increased their cytotoxicity 4-fold ( 18 k 5% to 74 k 1%) as did placental cells (17 f 2% to 75 -t 4%) (Fig. 2). There was some correlation between initial baseline NK activity and the subsequent spontaneous or LAK activity. In addition, there was clearly a significant increase in cytotoxic activity of some placenta samples (P < 0.05 by paired Student’s t test) (Fig. 3). This was not observed in any of the cord or adult cell samples. Whole mononuclear cells from adults, cords, and placentas all had low cytotoxic activity toward the NK-resistant Raji cell line. After IL-2 activation, placental cells increased their cytotoxic activity 3 t 1%to 59 f 7%, compared to cord cells (6 f 1% to 43 + 3%) and adult cells (11 -+ 2% to 38 -+ 4%). LAK activity for all three cell
6
CHIN ET AL. TABLE 3 Summary of Spontaneous Induced Cytotoxicity (SIC) and Lymphokine-Activated Killer (LAK) Cytotoxicity toward Molt Target Cells” SIC (Day 5 without IL-2)
Adult Cord Placenta
LAK (Day 5 with IL-2)
Percentage specific lysis (mean f SEM)
Number elevated/ number studiedb
Percentage specific lysis (mean + SEM)
Number elevated/ number studied b
I1 *2 13~2 34~6
lo/24 (42%) 17/31(46%) I l/14 (79%)
63 i 5 74+ 1 742 1
25125(100%) 37/37 (100%) 10/10(l00%)
’ Mononuclear cell preparations (1 X lo6 cells/ml) were incubated for 5-7 days with or without interleukin-2 (100 U/ml). After centrifugation at 400g for 10 min, the supematants were removed and cells were adjusted to 4 X lo6 cells/ml. bNumber elevated represents those samples which showed cytotoxic activity equal to or greater than a specific lysis of 10%(as calculated under Methods and Materials).
populations was also observed for two other NK-resistant cell lines, EL-4 and Daudi (data not shown). DISCUSSION The ability of human placental cells to synthesize significant quantities of IFN-7 following PHA or IL-2 stimulation is in marked contrast to umbilical cord cells, which make little IFN-7 following PHA activation ( 11). This difference may be attributable to the presence of activated macrophages in the placenta. Taylor and Bryson ( 12) have suggestedthat the defect in IFN--y production by cord cells resides with the ..
.!. . A
.> :. . .
! .. . .
. .
. ;
.I2:
..
i I:4 : 1;: :
*It50 i . :
63
l
:
l
Y -a-
l . .
. Adult
Cord Placenta Day 0 Without IL-2
Adult
Cord Plocento Doy 5 Wlthout IL-2
Adult
Cord Pkxento Day 5 With IL-2
FIG. 2. Natural killer and lymphokine-activated killer cell cytotoxicity of normal human adult, cord, and placental mononuclear cells against Molt target cells. The conditions are similar to those for Fig. 1.
IFN AND LAK CYTOTOXICITY
OF HUMAN
PLACENTAL
CELLS
loo-
80.
OL
Day 0 Without IL-2
Day 5 Without IL-2
W%‘I:-2
FIG. 3. Natural killer and lymphokine-activated killer activity of placental mononuclear cells against Molt target cells. The lines connect the same placental preparation, indicating the alteration after a 5-day incubation with or without IL-2. There is correlation (P < 0.05) between spontaneously induced cytotoxicity (middle dots) and LAK activity (right dots). No correlation, however, was noted between the initial NK activity and the subsequent development of LAK or spontaneous cytotoxicity.
immature neonatal macrophage, reversible in the presence of adult macrophages. Cord cells can be stimulated to synthesize IFN--y by irradiation, indomethacin, or staphylococcal endotoxin, suggesting dysregulation ( 13, 14). IL-2 production may also be implicated since IL-2 can augment IFN-y production by some but not all cord cells. However, cord cells, in responseto mitogen or allogeneic cells, can synthesize IL-2 equivalent to that of adult cells (15, 16). In addition, cord serum has normal quantities of soluble IL-2 receptor ( 17). In any case,the ability of the fetal cells within the placenta to produce significant amounts of EN-7 suggeststhat placental cells are in an activated state, possibly because of increased antigen exposure, including maternal alloantigens. Similar findings were obtained by Bocci et al. (18) who found the presence of both IFN-a and IFN-/3 in isolated and perfused intact human placentas. A role of IFN--r in controlling viral infections is well established. There is also evidence for its participation in the host defense toward other pathogens such as LeishmarCu(19), Listeria (20), and Rickettsia (2 1). The antimicrobial effect may be by macrophage activation (22) with increased intracellular killing (23). Wilson and Haas (7) have attributed the increased neonatal susceptibility to Toxoplasma infection to a deficiency in the production of macrophage-activating lymphokines such as IFN-y. They also found that the anti-Toxoplasma activity of placental monocytes was similar to that of adult monocytes. Regulatory interactions between IL-2 and IFN-y have been described. Purified IL2 induces secretion of EN--r by NK cells (24) and T lymphocytes (25). Although Wakasugi et al. have claimed no IFN--y production following addition of exogenous IL-2 to cord cells ( 14), a close inspection of their data suggestsa significant increase in some salmpleswhich is greater than that induced by PHA alone. This is consistent with our results indicating that about 20% of cord blood samples produce IF%y after stimulation with IL-2. Recently, Seki et al. (33) also found that some cord cells were able to be stimulated to produce IFN-7 with IL-2.
8
CHIN ET AL.
We have previously demonstrated minimal cytotoxic activity in both cord and placental mononuclear cells when compared to normal adult or maternal controls (8). These activities include NK, antibody-dependent cellular cytotoxicity, and lectindependent cellular cytotoxicity. Recently, a new type of cytotoxic cell was described and termed lymphokine-activated killer (LAK) cell (26). Cells which are not B, T, or mature NK lymphocytes develop cytotoxic activity after long-term incubation with IL-2. The development of equivalent LAK activity in placental tissue, despite very low levels of NK activity, is clearly shown in this paper. Since Itoh et al. (27) have shown that some LAK cells are derived from Leu- 1I+ cells, the placental LAK activity may be attributable to the abundant Leu- 1If cells in placental mononuclear cell preparations (8). Placental mononuclear cell preparations retain or can develop considerable spontaneous cytotoxicity during the 5- to 7-day culture in the absence of IL-2. This is somewhat unexpected since both cord and adult cells show significant diminution of activity with prolonged culture. In fetal calf serum-supplemented medium, however, NK-like cytotoxic activity does develop (28). However, such activity does not develop when the serum source is pooled from humans with AB type blood. It is of note that placental spontaneous cytotoxicity appears to be primarily directed toward Molt targets. The importance of the target cell has been previously noted, since cord NK activity was equivalent to adult NK activity against Molt targets but depressedagainst K562 cells (29). Part of this spontaneous cytotoxic activity may be attributed to in vitro production of IFN--r (30). As noted previously, upon mitogen stimulation, placental cells produce significantly greater quantities of IFN-7 compared to cord cells. However, minimal levels of IFN--y were measured in the supematants of cells from placenta, as well as adult and cord, at two points during the incubation (Days 2 and 5) in the absenceof mitogen. This could imply that only small amounts of IFN-y are secretedby placental mononuclear cells, which then binds and stimulates spontaneous NK activity. This processmay be accelerated by other soluble factors releasedduring labor and delivery. Since lymphocyte cytotoxic activity may have a role in host defensesagainst various pathogens such as viruses (31) as well as in graft or transplant rejection (32), the enhanced cytotoxic ability of placental cells shown in this paper is of interest. Protection but not rejection of the developing fetus may represent a fine balance between the production of various lymphokines and development of cellular cytotoxicity. In conclusion, there are significant differences between cord and placental mononuclear cells in terms of IFN-7 production and cytotoxic activity. Placental cells appear to be in a state of increased activation, manifested by the presence of LAK cytotoxicity, persistent and spontaneous development of cytotoxic activity, and increased synthesis of IFN-+y. All of these features may reflect a placental immune defense system which is active in protecting the fetus. ACKNOWLEDGMENTS We are indebted to Cetus Corporation for giving us supplies of interleukin-2 and to the nursery staff of UCLA Hospital and St. Johns and Santa Monica Hospitals of Santa Monica, California, who helped obtain clinical samples.
REFERENCES 1. Wilson, C. B., Haas, J. E., and Weaver, W. M., J. Immunol. Methods%, 305, 1983. 2. Chaouat, G., Kolb, J. P., and Wegmann, T. G., Immunol. Rev. 7531, 1983.
IFN AND LAK CYTOTOXICITY
OF HUMAN
PLACENTAL
CELLS
9
3. Wood, 0. W., and King, C. R., Cell. Imrnunol. 69,341, 1982. 4. Flynn, A., Finke, J. H., and Hilfiker, M. L., Science218,475, 1982. 5. Due-Goiran, P., Robert-Galliot, B., Chudzio, T., and Chany, C., Proc. Natl. Acad. Sci. USA 80,2628, 1983. 6. Kolb, J. IP., Chaouat, G., and Chassoux, D., J. Irnrnunol. 132,2305, 1984. 7. Wilson, C. B., and Haas, J. E., J. Clin. Invest. 73, 1606, 1984. 8. Goldsobel, A., Ank, B., Spina, C., Giorgi, J., and Stiehm, E. R., Cell. Immunol. 97,335, 1986. 9. Chin, T. W., Ank, B. J., Murakami, D., Gill, M., Spina, C., Strom, S., and Stiehm, E. R., Cell. Immuno/. 103,241, 1986. 10. Chang, T. W., McKinney, S., Liu, V., Kung, P. C., Vilcek, J., and Le, J., Proc. Natl. Acad. Sci. USA 81,52 19, 1984. 11. Bryson, Y. J., Winter, H. S., Gard, S. E., Fischer, T. J., and Stiehm, E. R., Cell. Immunol. 55, 191, 1980. 12. Taylor, S., and Bryson, Y. J., J. Immunol. 134, 1493, 1985. 13. Wakasugi, N., and Virelizier, J. L., J. Immunol. 134, 167, 1985. 14. Wakasugi, N., Virelizier, J. L., Arenzana-Seisdedos, F., Rothhut, B., Huerta, J.-M. M., Russo-Marie, F., anedFiers, W., J. Immunol. 134, 172, 1985. 15. Miyawald, T., Seki, H., Taga, K., Sato, H., and Taniguchi, N., Clin. Exp. Immunol. 59,505, 1985. 16. Hauser, G. J., Zakuth, V., Rosenberg, H., Bing, T., and Spirer, Z., J C/in. Lab. Immunol. 16, 37, 1985.
17. Nelson, D. L., Kurman, C. C., Fritz, M. E., Bontin, B., and Rubin, L. A., Ped. Rex 20, 136, 1986. 18. Bocci, V., Paulesu, L., and Ricci, M. G., Proc. Sot. Exp. Biol. Med. 180, 137, 1985. 19. Sadick, M. D., Locksley, R. M., Tubbs, C., and Raff, H. V., J. Immunol. 136,655, 1986. 20. Kiderlen, A. F., Kaufmann, S. H. E., and Lohmann-Matthes, M.-L. L., Eur. J. Immunol. 14, 964, 1984. 21. Wisseman, C. L., Jr., and Waddell, A., J. Exp. Med. 157, 1780, 1983. 22. Schultz, R. M., and Kleinschmidt, W. J., Nature (London) 305,239, 1983. 23. Nathan, C. F., Murray, H. W., Weibe, M. E., and Rubin, B. Y., J. Exp. Med. 158,670, 1983. 24. Ortaldo, J. R., Mason, A. T., Gerard, J. P., Henderson, L. E., Farrar, W., Hopkins, R. F., Herberman, R. B., and Rabin, H., J. Immunol. 133,179, 1984. 25. Kasahar,a,T., Hooks, J. J., Dougherty, S. F., and Oppenheim, J. J., J. Immunol. 130, 1784, 1983. 26. Grimm, E. A., Mazunder, A., Zhang, H., and Rosenberg, S. A., J. Exp. Med. 155, 1823, 1982. 27. Itoh, K., Tilden, A. B., Kumagai, K., and Balch, C. M., J. Immunol. 134,802, 1985. 28. Golub, S. H., Golightly, M. G., and Zieloke, J. V., Int. J. Cancer 24,273, 1979. 29. Lubens, R. G., Gard, S. E., Soderberg-Warner, M., and Stiehm, E. R., Ceil. Immunoi. 74,40, 1982. 30. Itoh, K., Shiba, K., Shimizu, Y., Suzuki, R., and Kumagai, K., J. Immunol. 134,3124, 1985. 31. Welsh, F!. M., Curr. Top. Micro. Immunol. 92,83, 1981. 32. Lotzova., E., McCredie, K. B., Muesse, L., Dicke, K. A., and Freireich, E. J., Exp. Hematol. Today, 207-Z! 13, 1979. 33. Seki, U., Taga, K., Matsuda, A., Uwadana, N., Hasui, M., Miyawaki, T., and Taniguchi, N., J. Immunol. 1.37, 3 158, 1986.