Celular immune responses to malaria antigens are suppressed during acute Plasmodium falciparum infection, and evidence from both murine and human ...
Vol. 57, No. 10
INFECTION AND IMMUNITY, OCt. 1989, p. 3181-3188 0019-9567/89/103181-08$02.00/0
Copyright © 1989, American Society for Microbiology
Plasmodium falciparum Schizont Sonic Extracts Suppress Lymphoproliferative Responses to Mitogens and Antigens in Malaria-Immune Adults ELEANOR M. RILEY,* OUSMAN JOBE, MICHAEL BLACKMAN, HILTON C. WHITTLE, AND BRIAN M. GREENWOOD Medical Research Council Laboratories, P.O. Box 273, Fajara, Near Banjul, The Gambia Received 27 October 1988/Accepted 22 June 1989
Celular immune responses to malaria antigens are suppressed during acute Plasmodium falciparum infection, and evidence from both murine and human studies suggests that parasite-derived factors may be directly immunosuppressive. In this study we have shown that P. falciparum schizont sonic extract will suppress in vitro lymphoproliferative responses to purified malaria antigens and other soluble antigens. The degree of suppression appears to correlate with the level of the lymphoproliferative response to the schizont preparation and is correspondingly more marked in malaria-immune donors than in nonimmune individuals. The effect can be transferred with primed mononuclear cells and is partially abrogated by removal of CD8' lymphocytes. The suppressive component of the schizont preparation is nondialyzable and partially heat labile and comigrates with hemoglobin-derived proteins in the molecular mass range 10 to 20 kilodaltons. been taking regular proguanil chemoprophylaxis, and who had not experienced symptomatic malaria infection. Lymphocyte cultures. Blood (20 ml) was diluted in heparinized RPMI 1640 (Flow Laboratories, Irvine, United Kingdom), and mononuclear cells were isolated by buoyantdensity centrifugation (Lymphoprep, Nygaard, Norway) as described previously (11). Cells (2 x 105), in 150 ,ul of complete tissue culture medium containing 10% nonimmune (European) human serum, were placed in each well of a round-bottom microdilution plate (Linbro Chemical Co., New Haven, Conn.). Then 20 ,ul of schizont sonic extract and 20 pd of mitogen or purified soluble antigen were added to triplicate wells. Control wells received 20 1±l of culture medium or 20 ,ul of uninfected erythrocyte antigen. Plates were incubated at 37°C in 5% CO2 for 5 days, and then the samples were radiolabeled; harvesting took place 18 h later. Cellular incorporation of [3H]thymidine was measured by liquid scintillation counting. For preincubation of cells with schizont antigen, 1.5 x 106 cells were suspended in 1 ml of culture medium and 100 ,ul of schizont sonic extract or uninfected erythrocyte sonic extract was added. After 2 days, cells were washed three times in medium and 103 preincubated cells were added to each well of an autologous cell suspension, together with 20 ,ul of antigen or mitogen. The plates were then incubated for another 6 days before harvesting was carried out. Cell- and
Nonspecific immunosuppression is a well-recognized feature of acute falciparum malaria infection in humans (20) and is associated with increased susceptibility to concomitant bacterial infections and reduced humoral immune responses to vaccination (3, 10). Recent evidence suggests that there is also specific suppression of T-cell-mediated responses to malaria antigens during acute infection, but that T-cell responses to many other antigens are not affected (2, 6, 11, 14). Lymphocytotoxic antibody formation (21), increased frequency of CD8+ T (suppressor) cells (19), and defective activation of T helper cells (15, 17) have all been implicated in specific suppression of T-cell reactivity, and serum from malaria patients has been shown to suppress cell-mediated immune responses in vitro (11, 14). Circulating malaria antigens have been found in acutely ill and convalescent malaria patients (22), and it has been suggested that these contribute to suppression of in vitro lymphoproliferation (1). We have shown previously (12) that lymphocytes of malariaimmune adults respond in an antigen-specific manner to purified soluble malaria antigens but that crude Plasmodium falciparum schizont sonic extract is nonspecifically mitogenic for immune and nonimmune cells. To determine whether this mitogenic preparation is also immunosuppressive, we have investigated the effects of adding a sonicated schizont preparation to mononuclear cell cultures in an in vitro lymphoproliferation assay.
antigen-free supernatants were prepared by incubating washed, preincubated cells in fresh culture medium for 24 h. Depletion of CD8 cells. CD8-positive cells were removed from the mixed mononuclear cell preparation by rosetting with antibody-coated bovine erythrocytes (9). Briefly, mononuclear cells were incubated for 45 m with mouse monoclonal anti-CD8 antibody (OKT8; Ortho Diagnostics, Inc., Raritan, N.J.), washed thoroughly, and combined with goat anti-mouse immunoglobulin-coated bovine erythrocytes. After incubation for 45 min on ice, the cell pellet was resuspended in fresh culture medium, and CD8+ rosetted cells were separated from CD8- nonrosetted cells by Lymphoprep centrifugation. The CD8-depleted cell fraction was harvested from the interface. The efficiency of CD8+ deple-
MATERIALS AND METHODS Subjects. Venous blood samples were collected from 22 malaria-immune adults living in a rural Gambian village where transmission of falciparum malaria is intense and seasonal (4). Samples were collected at the end of the dry season, when malaria transmission is low. All subjects were aparasitemic at the time of sampling. Each individual was bled on at least two occasions. Control samples were obtained from six nonimmune Europeans who had been living in malaria-endemic countries for less than 3 years, who had *
Corresponding author. 3181
INFECT. IMMUN.
RILEY ET AL.
3182
0
20
0
0 0
16
0
12 E 0
0
0
8
0
0* 0
4
0 0 00 000
0
L
0 0* 0O 0 0
0
0
0
0
*00 _8.MA
*so
0 U-
--
0
0
0@@ 0 0
10
50
U
0 0
so
8
0
0
10
.
I
0
0 0 0
50
U
FIG. 1. Lymphoproliferative response to schizont sonic extract and control uninfected erythrocyte sonic extract. Each point represents the mean counts in triplicate wells for each individual. Horizontal bars represent mean counts for each group. Symbols: 0, immune donors; 0, nonimmune donors. Numbers along the bottom represent the amount of schizont sonic extract added: 0, none (i.e., background counts of unstimulated wells); 10, 10 ±g/ml added; 50, 50 ,ug/ml added; U, uninfected erythrocyte antigen (50 ,ug/ml) added. *, Insufficient cells were available from three individuals for testing with uninfected sonic extract.
tion was checked by immunofluorescent staining of the cell fractions, and the percentage of CD8+ cells was typically reduced from 25 to 30% in the undepleted preparation to approximately 3 to 4% in the depleted preparation. The proportion of CD8+ cells in the rosetted fraction ranged from 81 to 93%. Antigens and mitogens. Phytohemagglutinin (PHA) (Difco Laboratories, Detroit, Mich.) was added to cells at a final concentration of 12 ,ug/ml. Candida albicans extract (10%, wt/vol; Hollister Stier, Elkhart, Ind.) was used at a final dilution of 1:80. Purified soluble malaria antigen, affinity purified against polyclonal immune serum from a continuous culture of a Tanzanian isolate of P. falciparum (7), was a gift from P. H. Jakobsen, Statens Seruminstitut, Copenhagen, Denmark, and was used at a final concentration of 37.5 ,ug/ml. This antigen preparation consists of at least seven separate soluble exoantigens (molecular masses ranging from 30 to 300 kilodaltons). The preparation has previously been shown to induce proliferation of cells from immune individuals but not of cells from nonimmune individuals (12). Crude schizont antigen was obtained from a continuous culture of a Gambian isolate (G372) of P. falciparum. Schizont enrichment was performed by Plasmagel (Laboratoire Roger Bellon, Nevilly Sur Seine, France) sedimentation, and the enriched fraction was washed three times in serum-free parasite culture medium, sonicated for 30 s in a PG100 Ultrasonic disintegrator at 150 W (MSE, Crawley, United Kingdom), and stored at -20°C. The sonic extract was used at a final protein concentration of either 50 or 10 ,ug/ml. A
similar preparation of uninfected erythrocytes was used as a control. Column fractionation of schizont preparation. Schizont sonic extract was centrifuged at high speed to remove any precipitated components, and the supernatant (with 10% sucrose added) was run over a Sephacryl S-300 SF column (Pharmacia, Uppsala, Sweden) in phosphate-buffered saline-0.5% fetal bovine serum at a speed of 5 ml/h. Thirty 10-min fractions were collected, and every four consecutive fractions were pooled and sterilized by filtration through a filter (pore size, 0.22 ,um). The pooled fractions were compared with crude sonic extract, sonic extract supernatant, and sonic extract precipitate in standard proliferation assays. The molecular weights of the pooled fractions were estimated by protein electrophoresis. Analysis of data. Responses are expressed either as thousands of counts per minute (kcpm) or as Akcpm, which represents the counts in the wells with antigen minus the counts in the control wells without antigen. The effect of adding schizont sonic extract to the culture is expressed as the counts in the wells with antigen and sonic extract minus the counts in the wells with antigen only. Differences between results for wells with and without schizont sonic extract are analyzed by a paired t test.
Cells from
a group
experiments 1 to
Sa,
RESULTS of 14 immune donors were used in inclusive. Cells from two of these
VOL . 57 1989
IMMUNOSUPPRESSION BY P. FALCIPARUM SCHIZONT ANTIGENS
,
3183
48 0 0
40
0
0
0 0
32
0
0
.
0
E
.1
0
24
0* 0
0 0
0
.
00
0 0
@0
S
0
0 0
16
F
0
0 0
*
* 0
0
0
0
0 0
0
0
0 0
0
8
0
00
0
S
0
0
0 0
0
0
0
.
r
r-0
0I
. 50
10
U
0
[0
50 s0
10
'U
U
*
FIG. 2. Effect of schizont sonic extract on the lymphoproliferative response to PHA. Akcpm represents [(kcpm PHA + schizont) minus (kcpm schizont)]. Each point represents the mean counts in triplicate wells for each individual. Horizontal bars represent the mean counts for each group. Symbols: immune donors; 0, nonimmune donors. Numbers along the bottom represent the amount of schizont sonic extract added: 0, none (PHA alone); background values are 1.6 + 0.9 kcpm in immune donor cells and 2.2 0.6 kcpm in nonimmune donor cells); 10, 10 ,ug/ml added to PHA; 50, 50 ,ug/ml added to PHA; U, PHA plus uninfected erythrocyte antigen (50 ,ug/ml). *, Insufficient cells were available from two donors for testing with uninfected sonic extract. 0,
±
subjects and from eight
new
subjects
were
used in experi-
ment Sb.
(i) Lymphoproliferative
response
to schizont sonic extract
(experiment 1). At a concentration of 50 ,ug/ml, the schizont sonic extract causes significant proliferation of cells from both immune and nonimmune donors (P < 0.005) (Fig. 1). At a concentration of 10 ,ug/ml, the sonic extract induces proliferation of cells from some immune donors but not of cells from nonimmune controls. The control sonic extract of uninfected erythrocytes did not cause proliferation of cells from either group of donors, regardless of whether the high or the low concentration was used. (ii) Effects of adding schizont sonic extract to antigen- or mitogen-stimulated lymphocyte cultures (experiment 2). Experiment 2a measured the response to PHA. The effect of adding schizont sonic extract to PHA-stimulated cells was studied by comparing the counts in the wells containing PHA and sonic extract with those in the wells containing PHA alone (Fig. 2). For cells from nonimmune donors, addition of 50 ,ug of schizont sonic extract per ml (but not 10 ,ug/ml) significantly depressed the response to PHA (P < 0.01). For cells from immune donors, addition of either 10 or 50 ,ug of schizont sonic extract per ml significantly depressed the response to PHA (P < 0.025 and P < 0.005 respectively). The control sonic extract had no suppressive effect. Responses to PHA alone were somewhat higher in malariaimmune African donors than in the nonimmune European
donors, but values for both groups were within the normal range for 7-day PHA stimulations in our laboratory (E. M. Riley et al., unpublished data). Experiment 2b measured the response to C. albicans antigen. For cells from the nonimmune donors, responses to C. albicans were significantly decreased by the addition of 50 ,ug of schizont sonic extract per ml (P < 0.05); addition of 10 ,ug of schizont sonic extract per ml also decreased the response to C. albicans antigen, but the effect was not statistically significant (P > 0.05) (Fig. 3). The control uninfected-erythrocyte sonic extract had no effect on the response to C. albicans. For cells from the immune group, both concentrations of schizont antigen, but not the control preparation, markedly depressed the responses to C. albicans antigen (P < 0.005). This is not simply an effect of adding excess antigen to the culture, since higher doses of C. albicans antigen or mixtures of C. albicans antigen and PHA did not cause suppression of the response (data not shown). Experiment 2c measured the response to purified soluble malaria antigen. Responses to the purified P. falciparum antigen were low in the nonimmune control cells, and the difference between antigen-stimulated and unstimulated cells was not significant (P > 0.1). However, cells from three individuals in this group did give minimal responses to the antigen, and these responses were completely abrogated by addition of schizont antigen (Fig. 4). For the cells from the immune group, responses to the purified P. falciparum
INFECT. IMMUN.
RILEY ET AL.
3184
48 0
40 0
32
S 0 0
.
.
E
:L
24 0
.
0
16 0 0
8
S 0
0
0 S
0@
0 00
0
00 00
*0
* 00
0
0
a-
0
0
0
0
0 0
000
0
0 0
0
0
LL
0
0
0 00
50
10
0
00
50
10
U
FIG. 3. Effect of schizont sonic extract on the lymphoproliferative response to C. albicans antigen. Akcpm represents [(kcpm C. albicans schizont) minus (kcpm schizont)]. Each point represents the mean counts in triplicate wells for each individual. Horizontal bars represent the mean counts for each group. Background values are as in Fig. 2. Symbols: 0, immune donors; 0, nonimmune donors. Numbers along the bottom represent the amount of schizont sonic extract added: 0, none (C. albicans antigen alone); 10, 10 ,ug/ml added to C. albicans antigen; 50, 50 ,ug/ml added to C. albicans antigen; U, C. albicans antigen plus uninfected erythrocyte antigen (50 ,ug/ml). *, Insufficient cells were available from one individual for testing with uninfected sonic extract. +
antigen varied. However, addition of either concentration of sonic extract strongly depressed these responses (P < 0.005). Again, addition of the control sonic extract had no suppressive effect. (iii) Preincubation of cells with schizont sonic extract and their effects on antigen-stimulated lymphocytes (experiment 3). To investigate the possibility that suppression was cell mediated, we investigated the effects of preincubating cells with schizont sonic extract (or a control antigen from uninfected erythrocytes) and adding these cells to autologous cultures. We added 103 preincubated cells to 2 x 105 autologous cells and assessed the responses to C. albicans or malaria antigens after 6 days. Cells from eight immune
donors were used, and responses were analyzed by a paired t test (Fig. 5). The proliferative responses to both C. albicans and purified malaria antigens were significantly decreased by the addition of cells preincubated with schizont sonic extract (P < 0.025). Cells preincubated with control erythrocyte sonic extract had no effect; antigen responses in these cultures were not significantly different from those in cultures in which cells were preincubated in medium alone or in which no preincubated cells were added (data not shown). Supernatants from the same preincubated cells were also tested for suppressive activity as follows: 100 p1l of cell- and antigen-free supernatant from washed, preincubated cells was added to 2 x 105 autologous cells, and the proliferative
TABLE 1. Effect of removing CD8+ cells from the cell population prior to culture with schizont antigen CD8+ cells present Antigen
PHA C. albicans
Malaria (purified) a
n
=
Mean Akcpm + Control (erythrocytes)
33.3 + 3.0 27.3 ± 9.3 26.3 ± 5.9
CD8+ cells depleted
SEM' +Schizont antigen
Suppression
15.9 + 4.7 17.8 ± 7.6 3.9 ± 2.9
47.7 34.8 85.1
Mean Akcpm Control (erythrocytes)
46.8 ± 7.3 22.6 ± 7.8 18.4 ± 3.9
+
SEM' +Schizont antigen
Suppression
23.9 ± 2.7 21.5 ± 6.7 10.3 ± 5.1
48.9 4.8 44.0
6. Mean ± SEM background values in unstimulated wells are 3.9 ± 0.4 with CD8+ cells present and 3.5 ± 1.1 with CD8+ cells depleted.
VOL. 57, 1989
E
IMMUNOSUPPRESSION BY P. FALCIPARUM SCHIZONT ANTIGENS
20
0
16 .
0
3185
0 0
12 .
0
c. u
0
.
8
0 0
4
0
0
0
0
-00@
0 -a
0*
0
0
50
0
0
0
-0-
00
*
10
00 oo 00 U
-a0 00
'o
10
50
U
FIG. 4. Effect of schizont antigen on the lymphoproliferative response to purified P. falciparum soluble antigen. Akcpm represents [(kcpm purified P. falciparum + schizont) minus (kcpm schizont alone)]. Each point represents the mean counts in triplicate wells for each individual. Horizontal bars represent the mean counts for each group. Background values are as in Fig. 2. Symbols: 0, immune donors; 0, nonimmune donors. Numbers along the bottom represent the amount of schizont sonic extract added: 0, none (P. falciparum antigen alone); 10, 10 ,ug/ml added to P. falciparum antigen; 50, 50 ,ug/ml added to P. falciparum antigen; U, P. falciparum antigen plus uninfected erythrocyte antigen (50 ,ug/ml). *, Insufficient cells were available from two donors for testing with uninfected sonic extract. response to purified antigen was assessed after 6 days. There was no evidence that such supematants were able to suppress antigen-induced proliferation (data not shown).
(iv) Effects of removing CD8+ cells from the mononuclear cell population prior to antigen stimulation (experiment 4). To determine whether the cells involved in mediating suppression were typical CD8+ (T suppressor) cells, we removed CD8+ cells from the cell preparations from malaria-immune donors before setting up the lymphoproliferative assays. Removal of CD8+ cells made no overall difference to the degree of suppression of responses to PHA, but suppression of responses to purified soluble malaria antigen and, to a lesser extent, to C. albicans antigen, was markedly reduced in the absence of CD8+ cells (Table 1). (v) Preliminary characterization of the immunosuppressive component of P. falciparum (experiment 5). In experiment 5a,
lymphoproliferative assays were repeated with cells from six individuals by using schizont sonic extract which had been dialyzed (dialysis membrane molecular mass cutoff, 12 to 14 kDa) or heat treated (100°C for 5 min) instead of using the untreated schizont sonic extract. The results are shown in Table 2. Dialysis of schizont antigen made no difference to its suppressive potential, but heat treatment of the antigen did partially abrogate both its mitogenic and suppressive effects. In experiment 5b, Sephacryl S-300 column fractions were tested for their ability to induce lymphoproliferation and to suppress PHA responses in cells from 10 individuals. Ultracentrifugation of the sonic extract produced a highly suppressive supernatant and a somewhat less suppressive precipitate (Table 3). Proliferative responses to most of the column fractions were minimal, and when these fractions
TABLE 2. Effect of dialyzed and heat-treated schizont antigen on the lymphoproliferative response to mitogens and other purified antigens Mean Akcpm + SEM (% suppression)a
Antigen
None 50 ,ug of schizont antigen/ml Heat-treated schizont antigen Dialyzed schizont antigen
PHA
14.4 5.8 9.2 6.6
± 2.5 ± 2.9 (59.9%) ± 5.6 (35.9%) ± 2.4 (53.9%)
C. albicans antigen
14.2 7.5 10.0 6.6
+ 3.0 ± 3.1 (47.1%) ± 2.4 (29.8%) t 2.2 (53.9%)
a Akcpm, Counts in stimulated wells minus background counts in unstimulated wells for cells from six immune donors. is 2.8 + 0.6.
Purified malaria antigen
7.5 0.1 2.9 0.2
+ 3.2 ± 0.1 (98.7%) ± 0.9 (61.3%) ± 0.2 (97.5%)
Background value (with no antigen)
3186
INFECT. IMMUN.
RILEY ET AL.
48
40 1
32
E Li .X
24
0 0 0
16 0
0
S
8
0
0
m 10
U
I
I
I
U
I
i
10
b.
a.
FIG. 5. Effect of adding cells preincubated with either sonicated, schizont-infected erythrocytes or with uninfected erythrocyte sonic mean counts in triplicate wells for each individual. (a) C. albicans antigen-stimulated cultures; (b) purified P. falciparum antigen-stimulated cultures. U, Cells preincubated with 10 ,ug of uninfected erythrocyte sonic extract per ml; 10, cells preincubated with 10 ,ug of schizont sonic extract per ml. Akcpm is value for antigen-stimulated wells minus value for unstimulated control wells.
extract to lymphoproliferation assays. Each point represents the
added to PHA-stimulated cells, no suppression of the PHA response was observed. However, fractions 15 to 18 (which contained most of the hemoglobin and other pigments) induced significant proliferative responses in cells from some donors and markedly suppressed the response to were
PHA (Table 3). The magnitude of the suppressive effect was similar to that seen with crude sonic extract or with sonic extract supernatant. Suppression of PHA responses by fractions 15 to 18 (or the crude schizont sonic extract) correlated with the ability of the cells from each individual to
TABLE 3. Effect of ultracentrifuged and column-fractionated schizont sonic extract on PHA-induced lymphoproliferation for 10 individual donorsa Mean kcpm after addition of fractionated' sonic extract
Mean kcpm after addition of ultracentrifuged sonic extract
designation
AG SS DB SC YS BB JJ JS MS FB
PHHA +A e extract alone sonic extract alone
etactn
19.2 25.5 23.8 28.3 15.6 9.7 11.5 38.2 12.6 37.3
8.6 9.3 4.6 10.8 5.2 0.5 0.2 0.9 1.0 0.2
11.3 15.4 13.8 14.5 17.1 11.0 11.7 40.9 13.6 32.5
a Values are means of triplicate wells (SEM was less than bF15/18, Pool of fractions 15 through 18.
sonic extract supernatant sueaat
sonic extract precipitate
F15/18 alone
PHA + F15/18
% Suppression
9.0 7.4 6.8 7.8 9.0 13.6 14.7 33.1 9.9 36.3
10.7 16.3 14.0 11.3 16.6 11.6 11.6 38.0 13.3 30.6
1.0 6.8 18.7 16.5 14.4 0.1 0.3 0.2 6.3 0.1
12.5 12.0 8.4 10.7 10.8 9.9 9.8 34.8 10.5 31.9
35.0 52.9 64.8 61.9 30.6 +2.6 15.0 8.9 17.0 14.3
peiiaeby F15/18
10%o). Mean + SEM background counts for unstimulated cultures, 2.02 + 0.69.
VOL. 57, 1989
IMMUNOSUPPRESSION BY P. FALCIPARUM SCHIZONT ANTIGENS
recognize the fraction (or the sonic extract) in a direct proliferation assay (Table 3). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of fractions 15 to 18 (data not shown) revealed several distinct bands in the molecular mass range 10 to 20 kDa, with a major band at 14 kDa, which probably represents reduced hemoglobin-derived proteins. DISCUSSION Antigen-specific suppression of lymphoproliferative responses to purified malaria antigens has been found in both adults and children who are acutely infected with P. falciparum malaria (6, 11, 14). In addition, serum or plasma from infected or recently recovered individuals has been shown to suppress responses to malaria antigens in both autologous (11) and homologous (16) systems. Since soluble malaria antigens are known to be present in the serum of malaria patients during infection and for some time after treatment (22), it is possible that a soluble parasite-derived substance is directly responsible for both these effects. Indeed, Ballet et al. were able to correlate the suppression of in vitro responses to PHA with the presence of circulating malaria antigens (1). In this study we have examined the immunosuppressive activity of a crude preparation of erythrocytes infected with mature P. falciparum schizonts. During a 7-day culture period, the schizont sonic extract severely depressed lymphoproliferative responses to malaria antigens, as well as (but less severely) responses to unrelated antigens and mitogens. Not only were these effects dependent upon the dose of schizont sonic extract used, but also they were more marked in cells from malaria-immune donors than in cells from nonimmune donors. The suppressive effect cannot be attributed to a direct toxic effect of the schizont preparation, since cells from many immune donors proliferated well when cultured with the schizont preparation alone. The suppressive effect was not seen with a control sonic extract of uninfected erythrocytes. Minimal responses to purified malaria antigens were seen in the cells from three of six nonimmune but malariaexposed control donors, suggesting that these subjects may have experienced subclinical malaria infection while taking schizonticidal antimalarial drugs. Also, responses to purified malaria antigens have been found previously in nonimmune donors and may indicate recognition of cross-reacting antigens (5). In the murine P. berghei system, a factor has been isolated from infected erythrocytes which suppresses primary humoral immune responses to T-cell-dependent (but not Tcell-independent) antigens in vivo (8). This factor appears to be a protein or glycoprotein of relatively low molecular mass (27 kDa). Theander et al. (16) describe a suppressive factor
in human malaria serum with a molecular mass 30 to 100 kDa which is heat labile at 56°C. Our data indicate that the suppressive factor is soluble, nondialyzable, and partially heat sensitive. The suppressive factor comigrates with hemoglobin and hemoglobin-derived pigments and other parasite components in the molecular mass range of 10 to 20 kDa. Further work is now required to define the suppressive
molecule and determine whether it is entirely parasite derived or whether altered host components are also involved. In murine studies, in vivo immunosuppression has been attributed to T-suppressor-cell activity in some systems and to suppressor macrophages in others (reviewed in reference 20). The mechanism of immunosuppression in human P.
3187
falciparum malaria has not been elucidated, but the data presented here suggest that activation of primed CD8+ suppressor cells may be one mechanism. Suppression was more complete (with lower doses of schizont antigen) in cultures from immune donors than in those from nonimmune donors. Also, even in the cells from the malaria-immune donors, the degree of suppression seemed to correlate with the level of cellular proliferation induced by the schizont preparation (Table 3). The suppressive effect could be partially transferred with prestimulated cells and was abrogated by CD8+ cell depletion. It seems paradoxical that the fraction of the schizont preparation which is most active in suppressing proliferative responses to mitogens such as PHA is also the fraction which is most potent in activating cells when presented alone. This raises the possibility that fractions 15 to 18 may be acting on more than one subset of cells, some of which proliferate while others (after a lag phase during which induction may occur) inhibit proliferation, via some kind of negative feedback mechanism. If this is so, the active component of fractions 15 to 18 may in fact represent a parasite-derived immunoregulatory factor with a role in limiting the T-cell response to malaria antigens. In the absence of definitive in vitro assays for estimation of clinical immunity to malaria, the extent to which lymphoproliferation is representative of in vivo cell-mediated immunity is uncertain. However, suppression of in vitro responses does appear to correlate with in vivo immunosuppression in acute malaria infection (reviewed in reference 18), and activation of T suppressor cells has been implicated in the hyporesponsiveness to malaria antigens seen in parasitemic individuals (14, 18, 19) and in the regulation of cellular immune responses in malaria-immune adults (13; H. C. Whittle, K. Marsh, and J. Brown, Clin. Exp. Immunol. in press). Attempts to transfer suppression with cell-free supematants of prestimulated cells were not successful, suggesting that if soluble, cell-derived suppressor factors are involved, either they are very labile or their concentration is critical. Mediators such as prostaglandin, released from activated macrophages or other antigen-presenting cells, may also be involved, since we have demonstrated that addition of indomethacin to cultures from acutely malariainfected donors will enhance proliferative responses to purified malaria antigens (E. M. Riley, C. MacLennan, D. Kwiatkowski, and B. M. Greenwood, Parasite Immunol., in press). The ability of P. falciparum-derived antigens to depress cellular immune responses in humans, both in vitro and in vivo, has serious implications for the development of a malaria vaccine. Potential vaccine antigens should be screened not only for their ability to induce specific antibody formation but also for their effects on the cellular immune system. In particular, the effects of specific purified malaria antigens on the immune response to other, unrelated antigens should be investigated, both in vitro and in vivo. ACKNOWLEDGMENTS We thank the people of Brefet for their participation and cooperation in this study; Idrissa Sambou, Muhammed Singateh, Lang Bayo, and Lamin Kuyateh for their technical assistance; and Kevin Marsh and Dominic Kwiatkowski for useful discussions and criticisms.
LITERATURE CITED 1. Ballet, J. J., P. Druilhe, P. Brasseur, S. Looareesuwan, P. Chantavanch, and S. Tharavanl. 1986. Influence of circulating malaria antigens on cell mediated immunity in acute Plasmo-
3188
RILEY ET AL.
dium falciparum malaria. Acta Trop. 43:255-262. 2. Brasseur, P., M. Agrapart, J. J. Ballet, P. Druilhe, M. J. Warreil, and S. TharavaniJ. 1983. Impaired cell-mediated immunity in Plasmodium falciparum-infected patients with high parasitaemia and cerebral malaria. Clin. Immunol. Immunopathol. 27:38-50. 3. Greenwood, B. M. 1974. Immunosuppression in malaria and trypanosomiasis, p. 137-146. In R. Porter and J. Knight (ed.), Parasites in the immunised host: mechanisms of survival. Associated Scientific Publishers, Amsterdam. 4. Greenwood, B. M., A. K. Bradley, A. M. Greenwood, P. Byass, K. Jammeh, K. Marsh, S. Tulloch, F. S. J. Oldfield, and R. Hayes. 1987. Mortality and morbidity from malaria among children in a rural area of The Gambia, West Africa. Trans. R. Soc. Trop. Med. Hyg. 81:478-486. 5. Guttinger, M., P. Caspers, B. Takacs, A. Trzeciak, D. Gilleson, J. R. Pink, and F. Sinigaglia. 1988. Human T cells recognise polymorphic and non polymorphic regions of Plasmodium falciparum circumsporozoite protein. EMBO J. 7:2555-2558. 6. Ho, M., H. K. Webster, S. Looareesuwan, W. Supanaranond, R. E. Phillips, P. Chanthavanich, and D. A. Warrell. 1986. Antigen-specific immunosuppression in human malaria due to Plasmodium falciparum. J. Infect. Dis. 153:763-771. 7. Jepsen, S., and B. Andersen. 1981. Immunoadsorbent isolation of antigens from the culture medium of in vitro cultivated Plasmodiumfalciparum. Acta Pathol. Microbiol. Scand. Sect C 89:99-103. 8. Khansari, N., M. Segre, and D. Segre. 1981. Immunosuppression in murine malaria: a soluble immunosuppressive factor derived from Plasmodium berghei infected blood. J. Immunol. 127:1889-1893. 9. Lane, H. C., J. M. Depper, W. C. Greene, G. Whalen, T. A. Waldmann, and A. S. Fauci. 1985. Qualitative analysis of immune function in patients with the acquired immunodeficiency syndrome. N. Engl. J. Med. 313:79-84. 10. Mabey, D. C. W., A. Brown, and B. M. Greenwood. 1987. Plasmodium falciparum malaria and Salmonella infections in Gambian children. J. Infect. Dis. 155:1319-1321. 11. Riley, E. M., G. Andersson, L. N. Otoo, S. Jepsen, and B. M. Greenwood. 1988. Cellular immune responses to Plasmodium falciparum antigens in Gambian children during and after an attack of acute falciparum malaria. Clin. Exp. Immunol. 73: 17-22. 12. Riley, E. M., S. Jepsen, G. Andersson, L. N. Otoo, and B. M.
INFECT. IMMUN.
13.
14.
15.
16.
17.
18.
19.
20. 21.
22.
Greenwood. 1988. Cell-mediated responses to Plasmodium falciparum antigens in adult Gambians. Clin. Exp. Immunol. 71:377-382. Riley, E. M., 0. Jobe, and H. C. Whittle. 1989. CD8+ T cells inhibit Plasmodium flaciparum-induced lymphoproliferation and gamma-interferon production in cell preparations from some malaria-immune individuals. Infect. Immun. 57:12811284. Theander, T. G., I. C. Bygbjerg, B. J. Andersen, S. Jepsen, A. Kharazmi, and N. Odum. 1986. Suppression of parasite-specific response in Plasmodium falciparum malaria. A longitudinal study of blood mononuclear cell proliferation and subset composition. Scand. J. Immunol. 24:73-81. Theander, T. G., I. C. Bygbjerg, S. Jepsen, M. Svenson, A. Kharazmi, P. B. Larsen, and K. Bendtzen. 1986. Proliferation induced by Plasmodium falciparum antigen and interleukin-2 production by lymphocytes isolated from malaria-immune individuals. Infect. Immun. 53:221-225. Theander, T. G., M. Svenson, I. C. Bygbjerg, A. Kharazmi, S. Jepsen, B. J. Andersen, and P. B. Larsen. 1987. Inhibition of human lymphocyte proliferative response by serum from Plasmodium falciparum infected patients. Acta Pathol. Microbiol. Immunol. Scand. Sect. C 95:257-263. Troye-Blomberg, M., G. Andersson, M. Stoczkowska, R. Shabo, P. Romero, E. Patarroyo, H. Wigzell, and P. Perlmann. 1985. Production of IL-2 and IFN--y by T cells from malaria patients in response to Plasmodium falciparum or erythrocyte antigens in vitro. J. Immunol. 135:3498-3504. Troye-Blomberg, M., and P. Perlmann. 1988. T cell functions in P. falciparum and other malarias. Progr. Allergy 41:253-287. Troye-Blomberg, M., P. Romero, M. E. Patarroyo, A. Bjorkmann, and P. Perlmann. 1984. Regulation of the immune response in Plasmodium falciparum malaria. III. Proliferative response to antigen in vitro and subset composition of T cells from patients with acute infection or from immune donors. Clin. Exp. Immunol. 58:380-387. Weidanz, W. P. 1982. Malaria and alterations in immune reactivity. Br. Med. Bull. 38:167-172. Wells, R. A., K. Pavanand, S. Zolyomi, B. Permpanich, and R. P. MacDermott. 1980. Antilymphocytotoxic antibodies in sera of Thai adults infected with Plasmodium falciparum or Plasmodium vivax. Clin. Exp. Immunol. 39:663-667. Wilson, R. J. M., I. A. McGregor, P. Hall, K. Williams, and R. Bartholomew. 1969. Antigens associated with Plasmodium falciparum infections in man. Lancet ii:201-205.
