Eur. J. Immunol. 1991. 21: 2145-2152
Philippe Velge, Jean-Pierre Kusnierz, Ali Ouaissi, Bkatrice Marty, Bach Nga Pham and Andrk Capron Centre d’Immunologie et de Biologie Parasitaire, Unitk mixte INSERM U167 - CNRS 624, Institut Pasteur, Lille
ADCC against I: cvuzi-infected T lymphocytes
2145
Trypanosoma cruzi: infection of T lymphocytes and their destruction by antibody-dependent cell-mediated cytotoxicity We have demonstrated, with optical and transmission electron microscopy, that Trypanosoma cruzi trypomastigotes infect and multiply inside T lymphocytes. The infection rate we have observed inT cells was similar to that seen in the case of macrophage or polymorphonuclear cell infection. Flow cytofluorometric analysis of T lymphocytes purified from mice in the acute phase of the disease, revealed the presence of parasite-derived antigens on their surface. These antigens appear to be specific to 7: cruzi and they could be the result of intracellular parasite antigens as well as adsorption of T cruzi antigens on the surface of noninfected T cells. Antibodies recognizing these surface antigens were present in both 7: cruzi-infected mouse and human sera. They were able to induce antibody-dependent cell-mediated cytotoxicity (ADCC) in the presence of nonimmune mononuclear cells both in autologous and in heterologous combinations. Consequently, we provided evidence suggesting that T lymphocytes could be destroyed during the acute phase of Chagas’ disease either by cell infection or by an ADCC mechanism against cells bearing parasite antigens on their surface. Thus, the ability of trypomastigotes to invade T cells may play a crucial role in the immunopathogenesis characteristic of Chagas’ disease.
1 Introduction The ability of parasites to survive and multiply in their host often depends upon their ability to inhibit or to suppress the host immune response. Chagas’ disease, the result of infection by the protozoan parasite Trypanosoma cruzi, is characterized by an acute phase with high parasitemia, during which trypomastigotes invade a variety of cell types including M a , polymorphonuclear and muscle cells. During the chronic phase, parasites are not readily detectable in the circulation but are sequestered in various tissues[11.The acute phase is associated with an immunosuppression state, believed to facilitate the dissemination of the parasite in the host [2]. The suppression is related to the appearance of suppressor M a [3-41, suppressor T cells [5] or deficient production of IL 2 [6]. Impaired lymphocyte proliferation in response to mitogens [7] or parasite antigens [8] has also been reported. The mechanisms underlying this immune dysfunction are poorly understood, but it appears that lymphocyte function is directly or indirectly altered during 7: cruzi infection [9].We have previously reported [lo] that trypomastigotes can infect human and murine T lymphocytes in vitro but not B lymphoma cells. The 7: cruzi antigens detected on the surface of in vitro infected T lymphoma cells could, therefore, represent putative targets for cytotoxic mechanisms [101. In the present report we show in vivo destruction, during the acute phase of Chagas’disease, of T lymphocytes either [I 90791 Correspondence: PhilippeVelge, Institut National de la Recherche Agronomique, Centre de Tours-Nouzilly, Laboratoire de Pathologie Infectieuse et dImmunologie, F-37380 Nouzilly, France Abbreviations: ADCC: Antibody-dependent cell-mediated cytotoxicity NMS: Normal mouse serum NHS: Normal human serum, IMS: 7: cruzi-infected mouse serum GAM: Goat IgG to mouse immunoglobulins 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1991
as a result of the trypomastigote infection or by an antibody-dependent cell-mediated cytotoxicity (ADCC) mechanism which was raised against 7: cruzi-specific antigens. These results could in part explain the processes that lead to impaired human lymphocyte functions and that might be involved in the immunosuppression of Chagas’ disease.
2 Materials and methods 2.1 Parasites and animals
The Y strain of T cruzi was used throughout this work. It was maintained on 3T3 fibroblast cell line as described [l 11. Eight-week-old BALBk male mice (bred at the Pasteur Institute, Lille) were infected by i.p. injection of 1 x lo4 to 1 x lo5 Y blood trypomastigotes. 2.2 Cells and sera
Human T lymphoma Jurkat or MOLT 4 cells were a gift of Dr. C. Auriault (Institut Pasteur, Lille). HumanT cell clone (LDG4) was obtained from the culture of peripheral blood cells purified from a chronic SLE patient [12]. Cells were maintained in culture by stimulation with irradiated syngenic APC and the N-terminal peptide SS-A [13] in the presence of 1% of Jurkat cell SN. Cells were cloned at a ratio of one cell per well following by a second cloning at 0.3 cell per well. The clonal cells used in this report were maintained 2 weeks after stimulation without the presence of APC and expressed the following markers: CD2, CD3, CD8, CD45R0, CD25, HLA-DR [14]. Splenic Tcells were prepared from normal or infected BALBk mice and were cultured in R10 medium following the procedure described previously [15]. The T cellenriched population was analyzed by FCM using rat 0014-2980/91/0909-2145$3.50+.25/0
2146
Eur. J. Immunol. 1991. 21: 2145-2152
PVelge, J.-P. Kusnierz, A . Ouaissi et al.
anti-Thy-1.2 mAb (Sera Lab, Crawley Down, GB), FITCGoat anti-mouse (GaM) F(ab')2 and rat anti-Mac1 mAb (Boehringer Mannheim, Mannheim, FRG). Immune sera from T cruzi-infected Bolivian and Argentinian patients were provided by Dr. F. Breniere and Dr. U. 0. Martin, Goat hyperimmune sera against cultured forms of 7: cruzi were prepared as described [16]. Rat immune sera against 7: cruzi excretedhecreted antigens and mAb 155D3 against trypomastigote firbonectin-collagen receptor were prepared as described [17, 181.
30 min at 4°C with FITC-labeled of goat anti-human Ig (FITC-GaHu, Cappel). Finally, the lymphocytes were washed twice in HBSS, fixed with 1% paraformaldehyde and examined by FCM (50-H cytofluorograph, Ortho Instrument, Westwood, MA). Double labeling was performed after incubation with human sera. Washed cells were incubated for 30 min at 4°C with PE- labeled goat anti-human Ig (PE-GaHu, Southern Biotechnology Associates, Birmingham, AL) diluted 1/50 together with FITClabeled mouse anti-Thy-1.2 mAb (Miles, Naperville, IL) diluted 1/100. After washing, labeled cells were fixed with 1% paraformaldehyde and analyzed.
2.3 Transmission EM
2.7 Cytotoxicity assay The in v i m T cell infection was carried out following the procedure described previously [lo]. After 36 h of infection, cells were treated for EM analysis as described in previous reports [ 191.
2.4 Immunoperoxidase labeling Leukocytes were characterized by using antibodies against cell surface markers: rat anti-Mac-1 mAb; rat anti-Thy-1.2 mAb; goat IgG to mouse Ig (GaM). In the case of M@ and T lymphocytes cytocentrifuged cells were fixed with methanol, treated for 30 min at room temperature with normal mouse serum (NMS) diluted 1/50 in HBSS, washed in HBSS and then incubated 1h at room temperature with anti-Mac-1 or anti-Thy-1.2 mAb diluted 1/50 in HBSS. Washed cells were incubated 30 min with peroxidaselabeled F(ab'): fragments of goat anti-rat Ig [peroxidaseGaR F(ab')z; Cappel Lab., Cochranville, PA]. The slides were finally washed twice, stained with the diaminobenzidine tetrahydrochloride (Fluka AG, Buchs, Switzerland) and counterstained with RAL 555 stain. B lymphocytes were identified by incubating the slide with peroxidaselabeled GaM (peroxidase-GaM) (Pasteur Institut, Paris, France) diluted 1/100 in HBSS. The enzyme reaction was revealed as above.
2.5 I n vivo-infected leukocyte counts The in vivo infection level of leukocytes was determined in the spleen, during the acute phase of 7: cruzi-infected BALB/c mice. RBC were depleted by an osmotic shock and leukocytes were characterized by immunostaining with peroxidase as already described. One thousand to five thousand cells were counted on each slide (magnification X 1000). Levels of parasitemia were determined in peripheral blood as described by Rivera-Vanderpas et al.) [20]). The percentage of infected cells was calculated according t o the formula: number of infected cells x 100/total number of cells of the same type counted.
Ten million purified mouse T cells were labeled with 7.4 Ml3q (200 pCi) of Naz5'Cr04 (Amersham Int., Amersham, GB) for 3 h at 37 "C in R10 and washed three times before use. Target cells (1 x lo5) in SO pl of R10 medium were incubated in wells of Falcon 3040 flat-bottom microplates with 50 pl of various dilutions of heat-inactivated human or mouse sera or purified human IgG (10 pglwell) for 30 rnin at 37°C. Then 2 x loh mononuclear cells in 100 pl (E/T = 20/1) were added. Human mononuclear cells were purified from blood on Ficoll-Hypaque, whereas mouse mononuclear cells were purified from spleen and LN. After 12 h at 37 "C, 100 p1 from each well were removed and counted in a gamma counter. Triplicates of each experimental group were run and the percentage of cytotoxicity were calculated according to the formula: (cpm of target + effector + serum) - (cpm of target serum) x 100/(cpm of target lysed by detergent - (cpm of target serum).
+
+
3 Results 3.1 I n vitro infection of T lymphocytes by T. cruzi Trypomastigotes incubated with T cells overnight at 37 "C were able to infect humanT lymphoma cells (Fig. 1A) and purified mouse or human normal T lymphocytes (Fig. 1B). Infected cells were shown to be T cells by indirect immunoperoxidase labeling with anti-Thy-1.2 mAb (Fig. 1C-D). The same percentage of infected cells was obtained when purified CD4+ or CD8+ lymphocyte subpopulations were used (1.93 ?c 0.32% and 1.45 ? 0.51%, respectively). The majority of infected cells are blast cells even if we have evidenced trypomastigotes inside quiescent T cells. When peroxidase-GaM was used, no infected B cells were observed. Intracellular localization of parasite in in v i m infected human T lymphocytes, a Jurkat cell line and a human T cell clone was determined by EM (Fig. 2). Parasites were observed both in the cytosol and in vacuoles.
3.2 In vivo infected leukocyte number 2.6 FCM analysis
T cell infection was examined during the acute phase of 7: Suspensions of 1 x lo6purified mouse splenic lymphocytes were washed three times in HBSS at 4°C and incubated 30 min at 4°C with NMS diluted 1/50 to block FcR. After washing, cells were incubated for 1 h at 4°C with human sera diluted 1/50 in HBSS and revealed by incubation
cruzi-infected BALB/c mice. Morphological examination combined with immunocytochemical staining to identify immune cells demonstrated the presence of in vivo infected T lymphocytes. In contrast, the use of peroxidase-GaM revealed that 7: cvuzi does not infect B cells (Fig. 3). The
Eur. J. Immunol. 1991. 21: 2145-2152
ADCC against T cruzi-infected T lymphocytes
2147
Figure 1. Optical microscopy observation of in vitroT cell infection. Human Jurkat lymphoma cells (A), purified human T lymphocyte (B) or normal mouse lymphocytes (C and D) were incubated with 7: cruzi trypomastigotes using a parasite :cell ratio of ten organisms per cell. Slides were prepared with a cytocentrifuge and stained with the RAL555 stain (A and B). In (C) and (D), cell preparations were first subject to an indirect immunoperoxidase labeling by using anti-Thy-1.2 mAb before staining. (A) and (B) magnification X 1200; (C) and (D) magnification X 1OOO.A- Amastig0tes;TLc:T lymphocyte; B Lc: B lymphocyte; LN: lymphocyte nucleus; PN: parasite nucleus; K: kinetoplast.
2148
Eur. J. Immunol. 1991. 21: 2145-2152
P.Velge, J.-l? Kusnierz, A. Ouaissi et al.
'I
Macrophagem
7
6-
54-
T Lymphoyteo
c
0
8
1
5
10
15
20
25
30
Parasitaemia (x lo6)
Figure 2. Transmission EM of human T cell clone infected in vitro by 7: cruzi. A human T cell clone was infected in vitro by 7: cruzi trypomastigotes during 36 h at 37 "C. Cells were processed for EM as described in Sect. 2.3. Electron micrograph of a T cell clone shows four parasites inside an intact vacuole and the presence of other parasites free in the cytosol (magnification X 10 120). A: Amastigote; CC: cell cytoplasm; LN: lymphocyte nucleus: PN: parasite nucleus; K: kinetoplast; F: flagellum; PV: parasitophorus vacuole.
level of T cell infection increases with the parasitemia and the infection rate of T lymphocytes is not very different from that obtained with splenic PMN or M a . In LN, where parasites are not resident [2], the T cell infection level is lower. For example, at a parasitemia of 12 X lo6 parasites/ml, 2% of splenic T lymphocytes and 0.1% of LN T cells were infected by 7: crrrzi. In peripheral blood the percentage of infected T lymphocytes was also similar to that of the infected monocytes. 3.3 Reactivity of immune sera against T lymphocytes from mice in the acute phase
Figure 3. In vivo-infected leukocyte number. Four types of leukocytes were examined: PMN; T lymphocytes (Thy-1.2+ cells); B lymphocytes (Ig+ cells); and M@ and monocytes (Mac-l+ cells). The ability of cells to be infected in vivo was examined in T cruzi-infected BALB/c mice during the acute phase of the disease. FU3C were depleted from splenic cells by osmotic shock. Slides were prepared with a cytocentrifuge. Leukocytes were characterized by an indirect peroxidase labeling, counterstained with the RAL555 stain and 1000 to 5000 cells were counted (magnification x 1000). Levels of parasitemia were determined in peripheral blood. The percentage of infected cells was calculated following the formula: number of infected cells X 100/total number of cells of type counted.
from infected mice but few or n o T cells from normal mice (Fig. 4C). Of 13 IHS tested, only 2 sera from the chronic phase did not recognize the lymphocyte surface (Fig. 4A and C). It can be noted that the fluorescence intensity observed was bimodal (Fig. 4A) and that the number of infected T lymphocytes was comparable to the number of cells presenting the high fluorescence intensity. However. until now, we have not enough data to correlate T cell infection and this high fluorescence intensity. Considering the total fluorescence intensity, the presence of antigens on the cell membrane could not be related to the leukocyte infection level. Indeed, although a maximum of only 4% of cells were infected, up to 70% showed fluorescence with IHS. We also observed a plurimodal fluorescence intensity with anti-7: cruzigoat serum (Fig. 4B) suggesting that there is a heterogeneity of parasite antigens present on the T cells during the acute phase of the disease. However, a rat serum raised against excretedkecreted products of trypomastigotes did not react withT cells from infected mice as well as a mAb raised against 7: cruzi fibronectin-collagen receptor (Fig. 4D).
We examined, by FCM, the presence of 7: cruzi antigens on T lymphocytes purified from infected mice (parasitemias of 106-107 parasites/ml). The murine acute and chronic phase sera that were tested reacted against T lymphocytes purified during the acute phase (data not shown). However, a weak fluorescence level could be related to a binding of antibodies to the FcR of contaminating B lymphocytes.?~ 3.4 Analysis of the reactivity of IHS against murine lymphocytes by a double-staining test block this binding, the experiments were performed in heterologous combination in which murine cells were preincubated with NMS. Under these conditions, the Double labeling was performed using FITC-labeled mouse majority of infected human sera (IHS) recognized T cells anti-Thy-1.2 mAb and PE-GaHu. Results presented in
Eur. J. Immunol. 1991. 21: 2145-2152
ADCC against I: cruzi-infected T lymphocytes
2149
Fluorescence Intensity
(Dl
(C) 1001
loo
e
2e
0
60
60
0
0
-
40
$
20
a
1
80
=
3
40
c
0
8
0
N HS
P14
P66
P95
P35
PC
2o NOS
AP
IGS
NRS
IRS
mAb 1 5 5 D 3
Figure 4. Reactivity of human and goat immune sera against murineT lymphocytes.The four panels show (A)the fluorescence intensity vs. relative cell number, obtained using Tcells from the acute phase with NHS and with the chronic phase serum (P66); (B) the fluorescence intensity vs. relative cell number, obtained using T cells from the acute phase with normal goat serum (NGS) and with a I: cruzi-immunized goat serum (IGS); (C) the percentage of fluorescent cells with human sera from the acute phase (AP) or from the chronic phase (P14, P95, P66, P35, Pc) or from healthy donors (NHS); (D) the percentage of fluorescent cells with normal goat serum (NGS), I: cruzi-immunized goat serum (IGS), normal rat serum (NRS), serum from rat immunized against excreted secreted I: cruzi products (IRS), unrelated IgG mAb (mAb), mAb raised against T. cruzi fibronectin-collagen receptor (155D3).T lymphocytes (1 x lo6) purified during the acute phase or from normal mice (a)were incubated 30 min at 4 "C with FITC-GaHu. Cells were then washed, fixed and analyzed by FCM.The low- and high-fluorescence intensity peak represent about 90% and 10% of fluorescent cells, respectively. Cells were 95% Thy-1.2+ and 7% Ig+.
Fig. 5 show that T cells from the acute phase were well recognized by the acute-phase serum, whereas no specific double labeling was obtained with the same serum on cells purified during the chronic phase or from noninfected mice. In addition no doubly labeled cells were observed when using a NHS or a negative human chronic phase serum (P14). We also observed here, ab bimodal fluorescence intensity with GaHu and acute phase Tcells (Fig. 5B).
3.5 Immune mouse sera induced cytotoxicity of normal mouse mononuclear cells against T lymphocytes from mice in the acute phase Fig. 6 shows that mouse sera from both the acute and the chronic phases induced cytotoxic activity of normal mouse mononuclear cells against T lymphocytes purified during the acute phase.The cytotoxicity was observed with 70% of the acute-phase sera diluted up to 1/200 and with 30% of
Cells purified during the acute phase
Cells purified during the chronic phase
Cells purified from normal mice
1
2
1
2
1
2
APS
07'
42.5
92
10.6
95
8.1
CPS
90
9.7
94
2.3
98
1.7
NHS
02
21.3
90
10.7
92
8.1
-
'percentage of fluorescence
3 0 .
I f
Figure 5. Analysis of the reactivity of human immune sera against mouseT lymphocytes by a double-labelingtest. Purified spleenic lymphocytes were prepared from normal or infected mice (during the acute or the chronic phase of the disease). Lymphocytes (lo6) were incubated 30min at 4°C with NMS, washed and incubated 1 h at 4°C with human sera. After washing, cells were exposed to both PE-GaHu and FITC-labeled mouse anti-Thy-1.2 mAb. Cells were washed, fixed and analyzed by FCM (gating on all nuclear cells). The table represents the percentage of fluorescence obtained with human sera (APS: acute-phase serum: CPS: chronic-phase serum; NHS: normal human serum). 1: Thy-1.2+ cells; 2: doublestained cel1s.The three panels show, (A): green fluorescence (anti-Thy-1.2 mAb) vs. cell number; (B) red fluorescence (anti-human antibodies) vs. cell number and (C) red fluorescence (LFL2) vs. green fluorescence ( L E I ) .
Eur. J. lmmunol. 1991.21: 2145-2152
I? Velge, J.-I? Kusnierz. A. Ouaissi et al. 100-
sculephare
El
chmnkphea malnla normaleels
~
80 -
e3
60 40 20
-
0-
A
B
A
B
A
A
B
APS
A
B
~
B
~
NMS
CPS
the chronic-phase sera. Whereas infected mouse serum (IMS) did not induced cytotoxic activity against the other target cells tested, suggesting that the ADCC reaction was specific to lymphocytes from the acute phase of Chagas' disease. Interestingly, only sera recognizing the T lymphocytes surface by FCM could induce cytotoxic activity in the presence of normal mouse mononuclear cells. 3.6 IHS induced cytotoxic activity of human mononuclear cells against T cells from mice in the acute phase IHS, diluted u p to 1/200, promoted a high and significant cytotoxic activity of normal human mononuclear cells, whereas NHS did not induce ADCC either against T lymphocytes purified from normal mice or during the T cruzi infection (Table 1). In addition, these IHS (diluted 1/200), induced negligible cytotoxicity against T cells from normal mice. Experiments with IgG from human sera, purified on DEAE-Sepharose (Pharmacia, Uppsala, Sweden), showed the ability of IgG antibodies from immune sera to induce ADCC at a concentration of 10 pg/well (Table 1). It is noteworthy that no significant difference was observed between the cytotoxicity induced by IgG of cardiopathic immune serum (IgG P49) and asymptomatic immune sera (IgG P1.59, P98).
4 Discussion It is generally considered that the ability of parasites to survive in the immunocompetent host depends upon a variety of escape mechanisms. In the case of 7: cruzi infection, the question of how this protozoan parasite may affect the host immune response to the parasite itself is still a subject of controversy. In the present report, we studied how T cells could be affected by 7: cruzi. Using anti-Thy-1.2 mAb and peroxidase-GaR F(ab');! we observed that T but not B lymphocytes could be infected in vitro by trypomastigotes. Transmission EM studies demonstrate the presence of trypomastigotes inside a human T cell clone (CD2+, CD3+, CD8+, CD45RO+). It is interesting to note here that purified CD4+ and CD8+ T lymphocytes were in vitro infected to the same extent.
Infection of T cells during the course of the disease was examined in 7: cruzi-infected BALB/c mice. Morphological
Figure 6. ADCC against murine T lymphocytes by murine naive mononuclear cells in the presence of IMS. Mouse T cells were ) from noninfected mice: ( R ) during the chronic phase of Chagas' disease: (U) during the acute phase of Chagas'disease and ( 8 )during the acute phase of malaria. NazS"CrO4-labeled T cells (1 x lo5) were incubated for 12 h at 37 "C with 1 X loh naive mouse mononuclear cells ( E n = 20/1) and mouse sera (APS: acute-phase serum: CPS: chronic-phase serum diluted 1/50 (A) or 1/200 (B). Each determination was performed in triplicate. This figure represent a single experiment, which was repeated twice with similar results.
examination, after indirect peroxidase labeling, strongly suggest the presence of infected T lymphocytes. The T cell infection was not a trivial phenomenon, indeed the infection rate of Tcells was similar to that observed, in the spleen, with PMN or with the preferential host cells, i.e. the M a . In addition infected T cells were observed both in LN and in peripheral blood. The leukocyte infection levels observed in vivo in mice are consistent with the results reported by Mosca et al. [21] with human monocytes.These results suggest that T cells could be considered in Chagas' disease as a host cell for parasites. Since T cells are known
Tablel. ADCC against mouse T lymphocytes by nonimmune human mononuclear cells in the presence of IHSa)
Reciprocal dilution NHS
10
40
200 CPS P14
10 40 200
CPS
10
PYS
40 40 200
CPS P 49 NH IgG IgG P l y IgG P149 IgG P98
10 pg/well
10 &well
10 kg/well 10 (rg/well
Immune cclls Oh)
3 12 12 2 9
32*)
Normal cclls 1hl 4
x 7
h 6
67dJ 3x*)
204
59)
23d)
49~1)
18''
9
8
5,
Sod) 6od) W)
x 24dl 16L"
a) Na2s1Cr04-labeled mouse T cells ( 1 x lo5) from the acute phase (immune cells) or from normal mice (normal cells) were incubated for 12 h at 37°C with 1 x loh human mononuclear cells (E/T = 20/1) and human sera or purified human IgG (10 yg/well). NHS: Normal human serum: CPS: Chronic-phase sera. number P14, P95, P49: NHIgG: purificd human IgG from healthy donors; IgG P159, IgG P9X: purified human IgG from donors in the chronic phase of the disease which were classificd as asymptomatic; IgG P49: purified human IgG from donors in the cardiopathic chronic phase of the diseasc. b) Percentage of cytotoxicity. c) Students' t-test: p < 0.05. d) Students' f-test: p < 0.01.
Eur. J. Immunol. 1991. 22: 2145-2152
to recirculate [22], they might constitute a possible alternative for trypomastigotes to invade several tissues. It is clear that the T lymphopenia and B lymphocytosis observed during the acute phase of the disease [23-251, cannot simply be explained by the ability of trypomastigotes to infect and destroy a small percent of T cells (infected cells are destroyed in vitro when parasites are released after 120-144 h). Thus, we studied the destruction by the host immune response of noninfected T cells bearing parasite antigens. Preliminary experiments did not demonstrate CTL or NK activity against T lymphocytes during the acute phase of mouse infections (data not shown). The lack of Tcell cytotoxicity could be related to results demonstrating that, despite CTL proliferation and activation, these cells appear to be nonspecific for parasite antigens [26]. By FCM, we demonstrated that lymphocytes purified during the acute phase were specifically recognized by mouse and human immune sera. The T cell lineage of these lymphocytes was demonstrated by using double labeling with FITC-labeled anti-Thy-1.2 mAb and PE-GaHu. These results are consistent with our previous observations [lo] showing that IHS but not NHS recognized human T lymphoma cells infected in vitro by 1: cruzi. The capacity of the same IHS and heat-inactivated mouse sera to induce cytotoxic activity against T cells, collected from mice in the acute phase of the disease, in cooperation with nonimmune mononuclear cells was demonstrated.The majority of the infected sera that recognized the surface of T lymphocytes from the acute phase could also induce lysis of these cellsThis activity of the immune sera is recovered in the IgG fraction The antigens detected by IHS and by IMS on the surface of T cells could either represent parasite-derived antigens or host antigens which are altered by the parasite or, alternatively, proceed from an autoimmune reaction against antigens from activated T lymphocytes (the surfaces of normal lymphocytes were not recognized by IHS or by IMS). Evidence for the existence of parasite-derived antigens on Tcell surface came from three series of experiments. First, the chronic-phase sera did not recognize a chronic-phase T cell extract by Western blotting analysis, whereas the same sera bound to an extract from acute-phase T cells (data not shown). Second, activated T lymphocytes from mice infected by Plasmodium chabaudi for which the presence of parasite antigens on the infected cell surface has been described [27] were not lyzed by nonimmune mononuclear cells in cooperation with IMS. In the same way, IMS did not induce cytotoxic activity against T cells from a chronic phase of Chagas’ disease, in which polyclonal activation also occurs [l]. Third, using FCM an anti-T. cruzi goat serum only recognized T lymphocytes from the acute phase. The fact that immune sera did not recognize chronic-phase T cells could be explained by the lack of parasites in the blood during the chronic phase and, thus, the lack of parasite-antigen binding to the lymphocyte surface. Considering the level of T cell infection in vivo, the percentage of lymphocytes recognized by FCM and the percentage of ADCC, we think that parasite-derived antig-
ADCC against 7: cruzi-infected T lymphocytes
2151
ens are present on the surface of both infected and noninfected cells.These antigens could occur in at least two ways: by the integration of parasite components into the host membrane during intracellular parasite growth [28], or by the binding to the cell surface of extracellular parasite antigens released during host cell disruption or shed by the parasite [29]. It is highly probable that both hypotheses are true. Indeed, the bimodal fluorescence intensity obtained by FCM with infected sera suggests that parasite antigens are present on two different lymphocyte populations. The highly fluorescent population could correspond to the 3% of infected T cells observed by optical microscopy, Indeed the level obtained by optical microscopy is lower than in reality because cyto-centrifugation is more destructive for infected cells than for noninfected cells. The majority of parasite-derived antigens detected (weakly fluorescent population) could result from the adsorption of extracellular parasite components. This is supported by our binding experiments showing that 7: cruzi proteins could specifically bind to the T cell surface (data not shown). The presence of parasite antigens on the surface of infected and noninfected host cells cultivated in vitro has already been reported in the case of T. cruzi infections [30, 311 as well as with Theileria annulata-infected lymphocytes [32]. In the present study, we demonstrate that the parasite antigens are acquired in vivo by T lymphocytes, which play a crucial role in the immune response. Cells bearing such antigens could become targets for the host’s own immune response against the parasite, confirming reports which suggested that parasite antigens might also play an important role in the destruction of uninfected host cells [28, 301. In conclusion, we have provided evidence suggesting that T lymphocytes could be destroyed during the acute phase of Chagas’ disease by two mechanisms: (a) disruption of T cells as a result of intracellular parasite multiplication; (b) T cells bearing parasite antigens on their surface become targets for an ADCC mechanism. At the present time, we cannot directly associate T lymphocyte infection in vivo with the presence in vivo of parasite antigens on T cell surface, but synergy between cytotoxic activity raised against infected and uninfected cells and lymphocyte infection might play a crucial role in the immunopathogenesis characteristic of Chagas’disease.The low level of T cell infection does not diminish the importance of the above observations. Indeed, it is well known that in the case of HIV infection the very low T cell infection level can have dramatic consequences on immune regulation [33]. Not only do these mechanisms modify the normal physiological functions of T cells but also their lymphokine synthesis and, thus, the physiological functions of other immune cells. These hypotheses could explain reports demonstrating that soluble factors from infected spleen cell cultures have an inhibitory impact on IL 2-dependent lymphocyte proliferation [34, 351. We thank M . Loyens for excellent technical assistance and we are indebted to Dr. C. Mazingue, Prof. J. k! Dessaint, Dr. B. Kaeffer and Dr. D. Williams for helpful discussions and critical reading of the manuscript. Received November 26, 1990; in final revised form June 3 , 1991.
2152
P.Velge, J.-I? Kusnierz, A. Ouaissi et al.
5 References 1 D’Imperio Lima, M. R., Eisen. H., Minoprio, F!, Joskowicz, M. and Continho, A., J. lmmunol. 1986. 137: 353. 2 Brener, Z., Adv. Parasitol. 1980. 18: 247. 3 Rowland, E. C. and Kuhn, R. E., Infect. Immun. 1978. 20: 393, 4 Cunningham, D. S. and Kuhn, R. E . , J. Parasitol. 1980. 66: 390. 5 Ramos, C., Sch2idtler-Siwon,I. and Ortiz-Ortiz, L., J. lmmunol. 1979. 122: 1243. 6 Harel-Bellan, A., Joskowicz, M.. Fradelizi, D. and Eisen, H., Proc. Natl. Acad. Sci. USA 1983. 80: 3466. 3 Maleckar, J. R. and Kierszenbaum, R., J. Immunol. 1983.130: 908. 8 Ramos, C., Lamoyi, E., Feoli, M., Perez, R. M. and OrtizOrtiz, L., Exp. Parasitol. 1978. 45: 190. 9 Kierszenbaum. F. and Sztein, M. B., Parasitol. Today 1990. 6: 261. 10 Velge, Ph., Ouaissi, A. , Kusniez, J. P. and Capron, A., C. R. Acad. Sci. Paris, Ser. Ill, 1989. 309: 93. 11 Quaissi, M. A . , Afchain, D., Capron, A. and Grimaud, J. A,, Nature 1984. 308: 380. 12 Phan, B. N., Prin, L., Gosset, Ph., Hatron, PY., Devulder, B., Capron, A. and Dessaint, J. l?, Clin Exp. Immunol. 1989. 77: 168. 13 Lieu,T. S., Newkerk, M. M.. Capra, J. D. and Southeimer, R. D., J. CZin. Invest. 1989. 82: 96. 14 Pham, B. N., Doctoral Thesis, Lille University. 1990. 15 Pestel, J., Dissous, C., Dessaint, J. P., Louis, J., Engers, H. and Capron, A . , J. lmrnunol. 1985. 134: 4132. 16 Afchain, D., Le Rax, D., Fruit, J. and Capron, A., J. Parasitol. 1979. 65: 507. 17 Quaissi. M. A.. Taibi, A.. Cornette. J.,Velge, Ph., Marty, B., Loyens, M . , Esteva, M. Risvi. R. S. and Capron, A., Parasitology 1989. 40: 115.
Eur. J. Irnmunol. 1991. 21: 2145-2152 18 Velge, Ph., Quaissi, M. A., Cornette. J., Afchain, D. and Capron, A., Parasitology 1988. 97: 255. 19 Quaissi, M. A.. Santoro, F. and Capron, A., Exp. Parasitol. 1980. 50: 74. 20 Rivera-Vanderpas, M. T., Rodriguez, A. M., Afchain, D., Bazin, H. and Capron, A., Act. Trop. 1983. 40: 5. 21 Mosca,W., Castes, M., Ojeda, A. and El Homsi, A., Trans. R. SOC. Med. Hyg. 1986. 81: 975. 22 Hay, J. B. and Hobbs, B. B., J. Exp. Med. 1977. 145: 31. 23 Hofflin, J. M., Sadler, R. H. Araujo, F. G., Page, W. E. and Remmington, J. S., Trans. R. Soc. Trop. Med. Hyg. 1987. 81: 437. 24 Hayes, M. M. and Kierszenbaum, F., Infect. Immun. 31: 1117. 25 Minoprio, P. M., Eisen, H., Formi, L., D’Imperio Lima, L. R., Joskowicz, M. and Coutinho, A., Scand. J. lmmunol. 1986.24: 661. 26 Petry, K. and Eisen, H., Parasitol. Today 1989. 5: 111. 27 Leech, J. H., Barnwell, J. W., Miller, L. H. and Howard. R. J., J. Exp. Med. 1984. 159: 1567. 28 Arauyo, F. G., J. Immunol. 1985. 135: 4149. 29 Williams, G. T., Fielder, L., Smith, H. and Hudson, L., Act. Trop. 1985. 42: 33. 30 Ribeiro dos Santos, R. and Hudson, L.. Parasit. lmmunol. 1980. 2: 1. 31 Davis, C. D. and Kuhn, R. E., Znfect. lrnmun. 1990. 58: 1812. 32 Pearson, T. W., Dolan,T. T., Stagg, D. A. and Lundin, L. B., Nature 1979. 281: 678. 33 Edelrnan, A. S. and Zolla-Pazner, S., FASEB J. 1989. 3: 22. 34 Rottenberg, M., Lindqvist, C., Koman, A., Segura, E. L. and Om, A., Scand. J. Immunol. 1989. 30: 65. 35 Serrano, L. E. and ODaly, J. A., lnt. J. Parasitol. 1987. 17: 851.
Philippe Velge, Jean-Pierre Kusnierz, Ali Ouaissi, Bkatrice Marty, Bach Nga Pham and Andrk Capron Centre d’Immunologie et de Biologie Parasitaire, Unitk mixte INSERM U167 - CNRS 624, Institut Pasteur, Lille
ADCC against I: cvuzi-infected T lymphocytes
2145
Trypanosoma cruzi: infection of T lymphocytes and their destruction by antibody-dependent cell-mediated cytotoxicity We have demonstrated, with optical and transmission electron microscopy, that Trypanosoma cruzi trypomastigotes infect and multiply inside T lymphocytes. The infection rate we have observed inT cells was similar to that seen in the case of macrophage or polymorphonuclear cell infection. Flow cytofluorometric analysis of T lymphocytes purified from mice in the acute phase of the disease, revealed the presence of parasite-derived antigens on their surface. These antigens appear to be specific to 7: cruzi and they could be the result of intracellular parasite antigens as well as adsorption of T cruzi antigens on the surface of noninfected T cells. Antibodies recognizing these surface antigens were present in both 7: cruzi-infected mouse and human sera. They were able to induce antibody-dependent cell-mediated cytotoxicity (ADCC) in the presence of nonimmune mononuclear cells both in autologous and in heterologous combinations. Consequently, we provided evidence suggesting that T lymphocytes could be destroyed during the acute phase of Chagas’ disease either by cell infection or by an ADCC mechanism against cells bearing parasite antigens on their surface. Thus, the ability of trypomastigotes to invade T cells may play a crucial role in the immunopathogenesis characteristic of Chagas’ disease.
1 Introduction The ability of parasites to survive and multiply in their host often depends upon their ability to inhibit or to suppress the host immune response. Chagas’ disease, the result of infection by the protozoan parasite Trypanosoma cruzi, is characterized by an acute phase with high parasitemia, during which trypomastigotes invade a variety of cell types including M a , polymorphonuclear and muscle cells. During the chronic phase, parasites are not readily detectable in the circulation but are sequestered in various tissues[11.The acute phase is associated with an immunosuppression state, believed to facilitate the dissemination of the parasite in the host [2]. The suppression is related to the appearance of suppressor M a [3-41, suppressor T cells [5] or deficient production of IL 2 [6]. Impaired lymphocyte proliferation in response to mitogens [7] or parasite antigens [8] has also been reported. The mechanisms underlying this immune dysfunction are poorly understood, but it appears that lymphocyte function is directly or indirectly altered during 7: cruzi infection [9].We have previously reported [lo] that trypomastigotes can infect human and murine T lymphocytes in vitro but not B lymphoma cells. The 7: cruzi antigens detected on the surface of in vitro infected T lymphoma cells could, therefore, represent putative targets for cytotoxic mechanisms [101. In the present report we show in vivo destruction, during the acute phase of Chagas’disease, of T lymphocytes either [I 90791 Correspondence: PhilippeVelge, Institut National de la Recherche Agronomique, Centre de Tours-Nouzilly, Laboratoire de Pathologie Infectieuse et dImmunologie, F-37380 Nouzilly, France Abbreviations: ADCC: Antibody-dependent cell-mediated cytotoxicity NMS: Normal mouse serum NHS: Normal human serum, IMS: 7: cruzi-infected mouse serum GAM: Goat IgG to mouse immunoglobulins 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1991
as a result of the trypomastigote infection or by an antibody-dependent cell-mediated cytotoxicity (ADCC) mechanism which was raised against 7: cruzi-specific antigens. These results could in part explain the processes that lead to impaired human lymphocyte functions and that might be involved in the immunosuppression of Chagas’ disease.
2 Materials and methods 2.1 Parasites and animals
The Y strain of T cruzi was used throughout this work. It was maintained on 3T3 fibroblast cell line as described [l 11. Eight-week-old BALBk male mice (bred at the Pasteur Institute, Lille) were infected by i.p. injection of 1 x lo4 to 1 x lo5 Y blood trypomastigotes. 2.2 Cells and sera
Human T lymphoma Jurkat or MOLT 4 cells were a gift of Dr. C. Auriault (Institut Pasteur, Lille). HumanT cell clone (LDG4) was obtained from the culture of peripheral blood cells purified from a chronic SLE patient [12]. Cells were maintained in culture by stimulation with irradiated syngenic APC and the N-terminal peptide SS-A [13] in the presence of 1% of Jurkat cell SN. Cells were cloned at a ratio of one cell per well following by a second cloning at 0.3 cell per well. The clonal cells used in this report were maintained 2 weeks after stimulation without the presence of APC and expressed the following markers: CD2, CD3, CD8, CD45R0, CD25, HLA-DR [14]. Splenic Tcells were prepared from normal or infected BALBk mice and were cultured in R10 medium following the procedure described previously [15]. The T cellenriched population was analyzed by FCM using rat 0014-2980/91/0909-2145$3.50+.25/0
2146
Eur. J. Immunol. 1991. 21: 2145-2152
PVelge, J.-P. Kusnierz, A . Ouaissi et al.
anti-Thy-1.2 mAb (Sera Lab, Crawley Down, GB), FITCGoat anti-mouse (GaM) F(ab')2 and rat anti-Mac1 mAb (Boehringer Mannheim, Mannheim, FRG). Immune sera from T cruzi-infected Bolivian and Argentinian patients were provided by Dr. F. Breniere and Dr. U. 0. Martin, Goat hyperimmune sera against cultured forms of 7: cruzi were prepared as described [16]. Rat immune sera against 7: cruzi excretedhecreted antigens and mAb 155D3 against trypomastigote firbonectin-collagen receptor were prepared as described [17, 181.
30 min at 4°C with FITC-labeled of goat anti-human Ig (FITC-GaHu, Cappel). Finally, the lymphocytes were washed twice in HBSS, fixed with 1% paraformaldehyde and examined by FCM (50-H cytofluorograph, Ortho Instrument, Westwood, MA). Double labeling was performed after incubation with human sera. Washed cells were incubated for 30 min at 4°C with PE- labeled goat anti-human Ig (PE-GaHu, Southern Biotechnology Associates, Birmingham, AL) diluted 1/50 together with FITClabeled mouse anti-Thy-1.2 mAb (Miles, Naperville, IL) diluted 1/100. After washing, labeled cells were fixed with 1% paraformaldehyde and analyzed.
2.3 Transmission EM
2.7 Cytotoxicity assay The in v i m T cell infection was carried out following the procedure described previously [lo]. After 36 h of infection, cells were treated for EM analysis as described in previous reports [ 191.
2.4 Immunoperoxidase labeling Leukocytes were characterized by using antibodies against cell surface markers: rat anti-Mac-1 mAb; rat anti-Thy-1.2 mAb; goat IgG to mouse Ig (GaM). In the case of M@ and T lymphocytes cytocentrifuged cells were fixed with methanol, treated for 30 min at room temperature with normal mouse serum (NMS) diluted 1/50 in HBSS, washed in HBSS and then incubated 1h at room temperature with anti-Mac-1 or anti-Thy-1.2 mAb diluted 1/50 in HBSS. Washed cells were incubated 30 min with peroxidaselabeled F(ab'): fragments of goat anti-rat Ig [peroxidaseGaR F(ab')z; Cappel Lab., Cochranville, PA]. The slides were finally washed twice, stained with the diaminobenzidine tetrahydrochloride (Fluka AG, Buchs, Switzerland) and counterstained with RAL 555 stain. B lymphocytes were identified by incubating the slide with peroxidaselabeled GaM (peroxidase-GaM) (Pasteur Institut, Paris, France) diluted 1/100 in HBSS. The enzyme reaction was revealed as above.
2.5 I n vivo-infected leukocyte counts The in vivo infection level of leukocytes was determined in the spleen, during the acute phase of 7: cruzi-infected BALB/c mice. RBC were depleted by an osmotic shock and leukocytes were characterized by immunostaining with peroxidase as already described. One thousand to five thousand cells were counted on each slide (magnification X 1000). Levels of parasitemia were determined in peripheral blood as described by Rivera-Vanderpas et al.) [20]). The percentage of infected cells was calculated according t o the formula: number of infected cells x 100/total number of cells of the same type counted.
Ten million purified mouse T cells were labeled with 7.4 Ml3q (200 pCi) of Naz5'Cr04 (Amersham Int., Amersham, GB) for 3 h at 37 "C in R10 and washed three times before use. Target cells (1 x lo5) in SO pl of R10 medium were incubated in wells of Falcon 3040 flat-bottom microplates with 50 pl of various dilutions of heat-inactivated human or mouse sera or purified human IgG (10 pglwell) for 30 rnin at 37°C. Then 2 x loh mononuclear cells in 100 pl (E/T = 20/1) were added. Human mononuclear cells were purified from blood on Ficoll-Hypaque, whereas mouse mononuclear cells were purified from spleen and LN. After 12 h at 37 "C, 100 p1 from each well were removed and counted in a gamma counter. Triplicates of each experimental group were run and the percentage of cytotoxicity were calculated according to the formula: (cpm of target + effector + serum) - (cpm of target serum) x 100/(cpm of target lysed by detergent - (cpm of target serum).
+
+
3 Results 3.1 I n vitro infection of T lymphocytes by T. cruzi Trypomastigotes incubated with T cells overnight at 37 "C were able to infect humanT lymphoma cells (Fig. 1A) and purified mouse or human normal T lymphocytes (Fig. 1B). Infected cells were shown to be T cells by indirect immunoperoxidase labeling with anti-Thy-1.2 mAb (Fig. 1C-D). The same percentage of infected cells was obtained when purified CD4+ or CD8+ lymphocyte subpopulations were used (1.93 ?c 0.32% and 1.45 ? 0.51%, respectively). The majority of infected cells are blast cells even if we have evidenced trypomastigotes inside quiescent T cells. When peroxidase-GaM was used, no infected B cells were observed. Intracellular localization of parasite in in v i m infected human T lymphocytes, a Jurkat cell line and a human T cell clone was determined by EM (Fig. 2). Parasites were observed both in the cytosol and in vacuoles.
3.2 In vivo infected leukocyte number 2.6 FCM analysis
T cell infection was examined during the acute phase of 7: Suspensions of 1 x lo6purified mouse splenic lymphocytes were washed three times in HBSS at 4°C and incubated 30 min at 4°C with NMS diluted 1/50 to block FcR. After washing, cells were incubated for 1 h at 4°C with human sera diluted 1/50 in HBSS and revealed by incubation
cruzi-infected BALB/c mice. Morphological examination combined with immunocytochemical staining to identify immune cells demonstrated the presence of in vivo infected T lymphocytes. In contrast, the use of peroxidase-GaM revealed that 7: cvuzi does not infect B cells (Fig. 3). The
Eur. J. Immunol. 1991. 21: 2145-2152
ADCC against T cruzi-infected T lymphocytes
2147
Figure 1. Optical microscopy observation of in vitroT cell infection. Human Jurkat lymphoma cells (A), purified human T lymphocyte (B) or normal mouse lymphocytes (C and D) were incubated with 7: cruzi trypomastigotes using a parasite :cell ratio of ten organisms per cell. Slides were prepared with a cytocentrifuge and stained with the RAL555 stain (A and B). In (C) and (D), cell preparations were first subject to an indirect immunoperoxidase labeling by using anti-Thy-1.2 mAb before staining. (A) and (B) magnification X 1200; (C) and (D) magnification X 1OOO.A- Amastig0tes;TLc:T lymphocyte; B Lc: B lymphocyte; LN: lymphocyte nucleus; PN: parasite nucleus; K: kinetoplast.
2148
Eur. J. Immunol. 1991. 21: 2145-2152
P.Velge, J.-l? Kusnierz, A. Ouaissi et al.
'I
Macrophagem
7
6-
54-
T Lymphoyteo
c
0
8
1
5
10
15
20
25
30
Parasitaemia (x lo6)
Figure 2. Transmission EM of human T cell clone infected in vitro by 7: cruzi. A human T cell clone was infected in vitro by 7: cruzi trypomastigotes during 36 h at 37 "C. Cells were processed for EM as described in Sect. 2.3. Electron micrograph of a T cell clone shows four parasites inside an intact vacuole and the presence of other parasites free in the cytosol (magnification X 10 120). A: Amastigote; CC: cell cytoplasm; LN: lymphocyte nucleus: PN: parasite nucleus; K: kinetoplast; F: flagellum; PV: parasitophorus vacuole.
level of T cell infection increases with the parasitemia and the infection rate of T lymphocytes is not very different from that obtained with splenic PMN or M a . In LN, where parasites are not resident [2], the T cell infection level is lower. For example, at a parasitemia of 12 X lo6 parasites/ml, 2% of splenic T lymphocytes and 0.1% of LN T cells were infected by 7: crrrzi. In peripheral blood the percentage of infected T lymphocytes was also similar to that of the infected monocytes. 3.3 Reactivity of immune sera against T lymphocytes from mice in the acute phase
Figure 3. In vivo-infected leukocyte number. Four types of leukocytes were examined: PMN; T lymphocytes (Thy-1.2+ cells); B lymphocytes (Ig+ cells); and M@ and monocytes (Mac-l+ cells). The ability of cells to be infected in vivo was examined in T cruzi-infected BALB/c mice during the acute phase of the disease. FU3C were depleted from splenic cells by osmotic shock. Slides were prepared with a cytocentrifuge. Leukocytes were characterized by an indirect peroxidase labeling, counterstained with the RAL555 stain and 1000 to 5000 cells were counted (magnification x 1000). Levels of parasitemia were determined in peripheral blood. The percentage of infected cells was calculated following the formula: number of infected cells X 100/total number of cells of type counted.
from infected mice but few or n o T cells from normal mice (Fig. 4C). Of 13 IHS tested, only 2 sera from the chronic phase did not recognize the lymphocyte surface (Fig. 4A and C). It can be noted that the fluorescence intensity observed was bimodal (Fig. 4A) and that the number of infected T lymphocytes was comparable to the number of cells presenting the high fluorescence intensity. However. until now, we have not enough data to correlate T cell infection and this high fluorescence intensity. Considering the total fluorescence intensity, the presence of antigens on the cell membrane could not be related to the leukocyte infection level. Indeed, although a maximum of only 4% of cells were infected, up to 70% showed fluorescence with IHS. We also observed a plurimodal fluorescence intensity with anti-7: cruzigoat serum (Fig. 4B) suggesting that there is a heterogeneity of parasite antigens present on the T cells during the acute phase of the disease. However, a rat serum raised against excretedkecreted products of trypomastigotes did not react withT cells from infected mice as well as a mAb raised against 7: cruzi fibronectin-collagen receptor (Fig. 4D).
We examined, by FCM, the presence of 7: cruzi antigens on T lymphocytes purified from infected mice (parasitemias of 106-107 parasites/ml). The murine acute and chronic phase sera that were tested reacted against T lymphocytes purified during the acute phase (data not shown). However, a weak fluorescence level could be related to a binding of antibodies to the FcR of contaminating B lymphocytes.?~ 3.4 Analysis of the reactivity of IHS against murine lymphocytes by a double-staining test block this binding, the experiments were performed in heterologous combination in which murine cells were preincubated with NMS. Under these conditions, the Double labeling was performed using FITC-labeled mouse majority of infected human sera (IHS) recognized T cells anti-Thy-1.2 mAb and PE-GaHu. Results presented in
Eur. J. Immunol. 1991. 21: 2145-2152
ADCC against I: cruzi-infected T lymphocytes
2149
Fluorescence Intensity
(Dl
(C) 1001
loo
e
2e
0
60
60
0
0
-
40
$
20
a
1
80
=
3
40
c
0
8
0
N HS
P14
P66
P95
P35
PC
2o NOS
AP
IGS
NRS
IRS
mAb 1 5 5 D 3
Figure 4. Reactivity of human and goat immune sera against murineT lymphocytes.The four panels show (A)the fluorescence intensity vs. relative cell number, obtained using Tcells from the acute phase with NHS and with the chronic phase serum (P66); (B) the fluorescence intensity vs. relative cell number, obtained using T cells from the acute phase with normal goat serum (NGS) and with a I: cruzi-immunized goat serum (IGS); (C) the percentage of fluorescent cells with human sera from the acute phase (AP) or from the chronic phase (P14, P95, P66, P35, Pc) or from healthy donors (NHS); (D) the percentage of fluorescent cells with normal goat serum (NGS), I: cruzi-immunized goat serum (IGS), normal rat serum (NRS), serum from rat immunized against excreted secreted I: cruzi products (IRS), unrelated IgG mAb (mAb), mAb raised against T. cruzi fibronectin-collagen receptor (155D3).T lymphocytes (1 x lo6) purified during the acute phase or from normal mice (a)were incubated 30 min at 4 "C with FITC-GaHu. Cells were then washed, fixed and analyzed by FCM.The low- and high-fluorescence intensity peak represent about 90% and 10% of fluorescent cells, respectively. Cells were 95% Thy-1.2+ and 7% Ig+.
Fig. 5 show that T cells from the acute phase were well recognized by the acute-phase serum, whereas no specific double labeling was obtained with the same serum on cells purified during the chronic phase or from noninfected mice. In addition no doubly labeled cells were observed when using a NHS or a negative human chronic phase serum (P14). We also observed here, ab bimodal fluorescence intensity with GaHu and acute phase Tcells (Fig. 5B).
3.5 Immune mouse sera induced cytotoxicity of normal mouse mononuclear cells against T lymphocytes from mice in the acute phase Fig. 6 shows that mouse sera from both the acute and the chronic phases induced cytotoxic activity of normal mouse mononuclear cells against T lymphocytes purified during the acute phase.The cytotoxicity was observed with 70% of the acute-phase sera diluted up to 1/200 and with 30% of
Cells purified during the acute phase
Cells purified during the chronic phase
Cells purified from normal mice
1
2
1
2
1
2
APS
07'
42.5
92
10.6
95
8.1
CPS
90
9.7
94
2.3
98
1.7
NHS
02
21.3
90
10.7
92
8.1
-
'percentage of fluorescence
3 0 .
I f
Figure 5. Analysis of the reactivity of human immune sera against mouseT lymphocytes by a double-labelingtest. Purified spleenic lymphocytes were prepared from normal or infected mice (during the acute or the chronic phase of the disease). Lymphocytes (lo6) were incubated 30min at 4°C with NMS, washed and incubated 1 h at 4°C with human sera. After washing, cells were exposed to both PE-GaHu and FITC-labeled mouse anti-Thy-1.2 mAb. Cells were washed, fixed and analyzed by FCM (gating on all nuclear cells). The table represents the percentage of fluorescence obtained with human sera (APS: acute-phase serum: CPS: chronic-phase serum; NHS: normal human serum). 1: Thy-1.2+ cells; 2: doublestained cel1s.The three panels show, (A): green fluorescence (anti-Thy-1.2 mAb) vs. cell number; (B) red fluorescence (anti-human antibodies) vs. cell number and (C) red fluorescence (LFL2) vs. green fluorescence ( L E I ) .
Eur. J. lmmunol. 1991.21: 2145-2152
I? Velge, J.-I? Kusnierz. A. Ouaissi et al. 100-
sculephare
El
chmnkphea malnla normaleels
~
80 -
e3
60 40 20
-
0-
A
B
A
B
A
A
B
APS
A
B
~
B
~
NMS
CPS
the chronic-phase sera. Whereas infected mouse serum (IMS) did not induced cytotoxic activity against the other target cells tested, suggesting that the ADCC reaction was specific to lymphocytes from the acute phase of Chagas' disease. Interestingly, only sera recognizing the T lymphocytes surface by FCM could induce cytotoxic activity in the presence of normal mouse mononuclear cells. 3.6 IHS induced cytotoxic activity of human mononuclear cells against T cells from mice in the acute phase IHS, diluted u p to 1/200, promoted a high and significant cytotoxic activity of normal human mononuclear cells, whereas NHS did not induce ADCC either against T lymphocytes purified from normal mice or during the T cruzi infection (Table 1). In addition, these IHS (diluted 1/200), induced negligible cytotoxicity against T cells from normal mice. Experiments with IgG from human sera, purified on DEAE-Sepharose (Pharmacia, Uppsala, Sweden), showed the ability of IgG antibodies from immune sera to induce ADCC at a concentration of 10 pg/well (Table 1). It is noteworthy that no significant difference was observed between the cytotoxicity induced by IgG of cardiopathic immune serum (IgG P49) and asymptomatic immune sera (IgG P1.59, P98).
4 Discussion It is generally considered that the ability of parasites to survive in the immunocompetent host depends upon a variety of escape mechanisms. In the case of 7: cruzi infection, the question of how this protozoan parasite may affect the host immune response to the parasite itself is still a subject of controversy. In the present report, we studied how T cells could be affected by 7: cruzi. Using anti-Thy-1.2 mAb and peroxidase-GaR F(ab');! we observed that T but not B lymphocytes could be infected in vitro by trypomastigotes. Transmission EM studies demonstrate the presence of trypomastigotes inside a human T cell clone (CD2+, CD3+, CD8+, CD45RO+). It is interesting to note here that purified CD4+ and CD8+ T lymphocytes were in vitro infected to the same extent.
Infection of T cells during the course of the disease was examined in 7: cruzi-infected BALB/c mice. Morphological
Figure 6. ADCC against murine T lymphocytes by murine naive mononuclear cells in the presence of IMS. Mouse T cells were ) from noninfected mice: ( R ) during the chronic phase of Chagas' disease: (U) during the acute phase of Chagas'disease and ( 8 )during the acute phase of malaria. NazS"CrO4-labeled T cells (1 x lo5) were incubated for 12 h at 37 "C with 1 X loh naive mouse mononuclear cells ( E n = 20/1) and mouse sera (APS: acute-phase serum: CPS: chronic-phase serum diluted 1/50 (A) or 1/200 (B). Each determination was performed in triplicate. This figure represent a single experiment, which was repeated twice with similar results.
examination, after indirect peroxidase labeling, strongly suggest the presence of infected T lymphocytes. The T cell infection was not a trivial phenomenon, indeed the infection rate of Tcells was similar to that observed, in the spleen, with PMN or with the preferential host cells, i.e. the M a . In addition infected T cells were observed both in LN and in peripheral blood. The leukocyte infection levels observed in vivo in mice are consistent with the results reported by Mosca et al. [21] with human monocytes.These results suggest that T cells could be considered in Chagas' disease as a host cell for parasites. Since T cells are known
Tablel. ADCC against mouse T lymphocytes by nonimmune human mononuclear cells in the presence of IHSa)
Reciprocal dilution NHS
10
40
200 CPS P14
10 40 200
CPS
10
PYS
40 40 200
CPS P 49 NH IgG IgG P l y IgG P149 IgG P98
10 pg/well
10 &well
10 kg/well 10 (rg/well
Immune cclls Oh)
3 12 12 2 9
32*)
Normal cclls 1hl 4
x 7
h 6
67dJ 3x*)
204
59)
23d)
49~1)
18''
9
8
5,
Sod) 6od) W)
x 24dl 16L"
a) Na2s1Cr04-labeled mouse T cells ( 1 x lo5) from the acute phase (immune cells) or from normal mice (normal cells) were incubated for 12 h at 37°C with 1 x loh human mononuclear cells (E/T = 20/1) and human sera or purified human IgG (10 yg/well). NHS: Normal human serum: CPS: Chronic-phase sera. number P14, P95, P49: NHIgG: purificd human IgG from healthy donors; IgG P159, IgG P9X: purified human IgG from donors in the chronic phase of the disease which were classificd as asymptomatic; IgG P49: purified human IgG from donors in the cardiopathic chronic phase of the diseasc. b) Percentage of cytotoxicity. c) Students' t-test: p < 0.05. d) Students' f-test: p < 0.01.
Eur. J. Immunol. 1991. 22: 2145-2152
to recirculate [22], they might constitute a possible alternative for trypomastigotes to invade several tissues. It is clear that the T lymphopenia and B lymphocytosis observed during the acute phase of the disease [23-251, cannot simply be explained by the ability of trypomastigotes to infect and destroy a small percent of T cells (infected cells are destroyed in vitro when parasites are released after 120-144 h). Thus, we studied the destruction by the host immune response of noninfected T cells bearing parasite antigens. Preliminary experiments did not demonstrate CTL or NK activity against T lymphocytes during the acute phase of mouse infections (data not shown). The lack of Tcell cytotoxicity could be related to results demonstrating that, despite CTL proliferation and activation, these cells appear to be nonspecific for parasite antigens [26]. By FCM, we demonstrated that lymphocytes purified during the acute phase were specifically recognized by mouse and human immune sera. The T cell lineage of these lymphocytes was demonstrated by using double labeling with FITC-labeled anti-Thy-1.2 mAb and PE-GaHu. These results are consistent with our previous observations [lo] showing that IHS but not NHS recognized human T lymphoma cells infected in vitro by 1: cruzi. The capacity of the same IHS and heat-inactivated mouse sera to induce cytotoxic activity against T cells, collected from mice in the acute phase of the disease, in cooperation with nonimmune mononuclear cells was demonstrated.The majority of the infected sera that recognized the surface of T lymphocytes from the acute phase could also induce lysis of these cellsThis activity of the immune sera is recovered in the IgG fraction The antigens detected by IHS and by IMS on the surface of T cells could either represent parasite-derived antigens or host antigens which are altered by the parasite or, alternatively, proceed from an autoimmune reaction against antigens from activated T lymphocytes (the surfaces of normal lymphocytes were not recognized by IHS or by IMS). Evidence for the existence of parasite-derived antigens on Tcell surface came from three series of experiments. First, the chronic-phase sera did not recognize a chronic-phase T cell extract by Western blotting analysis, whereas the same sera bound to an extract from acute-phase T cells (data not shown). Second, activated T lymphocytes from mice infected by Plasmodium chabaudi for which the presence of parasite antigens on the infected cell surface has been described [27] were not lyzed by nonimmune mononuclear cells in cooperation with IMS. In the same way, IMS did not induce cytotoxic activity against T cells from a chronic phase of Chagas’ disease, in which polyclonal activation also occurs [l]. Third, using FCM an anti-T. cruzi goat serum only recognized T lymphocytes from the acute phase. The fact that immune sera did not recognize chronic-phase T cells could be explained by the lack of parasites in the blood during the chronic phase and, thus, the lack of parasite-antigen binding to the lymphocyte surface. Considering the level of T cell infection in vivo, the percentage of lymphocytes recognized by FCM and the percentage of ADCC, we think that parasite-derived antig-
ADCC against 7: cruzi-infected T lymphocytes
2151
ens are present on the surface of both infected and noninfected cells.These antigens could occur in at least two ways: by the integration of parasite components into the host membrane during intracellular parasite growth [28], or by the binding to the cell surface of extracellular parasite antigens released during host cell disruption or shed by the parasite [29]. It is highly probable that both hypotheses are true. Indeed, the bimodal fluorescence intensity obtained by FCM with infected sera suggests that parasite antigens are present on two different lymphocyte populations. The highly fluorescent population could correspond to the 3% of infected T cells observed by optical microscopy, Indeed the level obtained by optical microscopy is lower than in reality because cyto-centrifugation is more destructive for infected cells than for noninfected cells. The majority of parasite-derived antigens detected (weakly fluorescent population) could result from the adsorption of extracellular parasite components. This is supported by our binding experiments showing that 7: cruzi proteins could specifically bind to the T cell surface (data not shown). The presence of parasite antigens on the surface of infected and noninfected host cells cultivated in vitro has already been reported in the case of T. cruzi infections [30, 311 as well as with Theileria annulata-infected lymphocytes [32]. In the present study, we demonstrate that the parasite antigens are acquired in vivo by T lymphocytes, which play a crucial role in the immune response. Cells bearing such antigens could become targets for the host’s own immune response against the parasite, confirming reports which suggested that parasite antigens might also play an important role in the destruction of uninfected host cells [28, 301. In conclusion, we have provided evidence suggesting that T lymphocytes could be destroyed during the acute phase of Chagas’ disease by two mechanisms: (a) disruption of T cells as a result of intracellular parasite multiplication; (b) T cells bearing parasite antigens on their surface become targets for an ADCC mechanism. At the present time, we cannot directly associate T lymphocyte infection in vivo with the presence in vivo of parasite antigens on T cell surface, but synergy between cytotoxic activity raised against infected and uninfected cells and lymphocyte infection might play a crucial role in the immunopathogenesis characteristic of Chagas’disease.The low level of T cell infection does not diminish the importance of the above observations. Indeed, it is well known that in the case of HIV infection the very low T cell infection level can have dramatic consequences on immune regulation [33]. Not only do these mechanisms modify the normal physiological functions of T cells but also their lymphokine synthesis and, thus, the physiological functions of other immune cells. These hypotheses could explain reports demonstrating that soluble factors from infected spleen cell cultures have an inhibitory impact on IL 2-dependent lymphocyte proliferation [34, 351. We thank M . Loyens for excellent technical assistance and we are indebted to Dr. C. Mazingue, Prof. J. k! Dessaint, Dr. B. Kaeffer and Dr. D. Williams for helpful discussions and critical reading of the manuscript. Received November 26, 1990; in final revised form June 3 , 1991.
2152
P.Velge, J.-I? Kusnierz, A. Ouaissi et al.
5 References 1 D’Imperio Lima, M. R., Eisen. H., Minoprio, F!, Joskowicz, M. and Continho, A., J. lmmunol. 1986. 137: 353. 2 Brener, Z., Adv. Parasitol. 1980. 18: 247. 3 Rowland, E. C. and Kuhn, R. E., Infect. Immun. 1978. 20: 393, 4 Cunningham, D. S. and Kuhn, R. E . , J. Parasitol. 1980. 66: 390. 5 Ramos, C., Sch2idtler-Siwon,I. and Ortiz-Ortiz, L., J. lmmunol. 1979. 122: 1243. 6 Harel-Bellan, A., Joskowicz, M.. Fradelizi, D. and Eisen, H., Proc. Natl. Acad. Sci. USA 1983. 80: 3466. 3 Maleckar, J. R. and Kierszenbaum, R., J. Immunol. 1983.130: 908. 8 Ramos, C., Lamoyi, E., Feoli, M., Perez, R. M. and OrtizOrtiz, L., Exp. Parasitol. 1978. 45: 190. 9 Kierszenbaum. F. and Sztein, M. B., Parasitol. Today 1990. 6: 261. 10 Velge, Ph., Ouaissi, A. , Kusniez, J. P. and Capron, A., C. R. Acad. Sci. Paris, Ser. Ill, 1989. 309: 93. 11 Quaissi, M. A . , Afchain, D., Capron, A. and Grimaud, J. A,, Nature 1984. 308: 380. 12 Phan, B. N., Prin, L., Gosset, Ph., Hatron, PY., Devulder, B., Capron, A. and Dessaint, J. l?, Clin Exp. Immunol. 1989. 77: 168. 13 Lieu,T. S., Newkerk, M. M.. Capra, J. D. and Southeimer, R. D., J. CZin. Invest. 1989. 82: 96. 14 Pham, B. N., Doctoral Thesis, Lille University. 1990. 15 Pestel, J., Dissous, C., Dessaint, J. P., Louis, J., Engers, H. and Capron, A . , J. lmrnunol. 1985. 134: 4132. 16 Afchain, D., Le Rax, D., Fruit, J. and Capron, A., J. Parasitol. 1979. 65: 507. 17 Quaissi. M. A.. Taibi, A.. Cornette. J.,Velge, Ph., Marty, B., Loyens, M . , Esteva, M. Risvi. R. S. and Capron, A., Parasitology 1989. 40: 115.
Eur. J. Irnmunol. 1991. 21: 2145-2152 18 Velge, Ph., Quaissi, M. A., Cornette. J., Afchain, D. and Capron, A., Parasitology 1988. 97: 255. 19 Quaissi, M. A.. Santoro, F. and Capron, A., Exp. Parasitol. 1980. 50: 74. 20 Rivera-Vanderpas, M. T., Rodriguez, A. M., Afchain, D., Bazin, H. and Capron, A., Act. Trop. 1983. 40: 5. 21 Mosca,W., Castes, M., Ojeda, A. and El Homsi, A., Trans. R. SOC. Med. Hyg. 1986. 81: 975. 22 Hay, J. B. and Hobbs, B. B., J. Exp. Med. 1977. 145: 31. 23 Hofflin, J. M., Sadler, R. H. Araujo, F. G., Page, W. E. and Remmington, J. S., Trans. R. Soc. Trop. Med. Hyg. 1987. 81: 437. 24 Hayes, M. M. and Kierszenbaum, F., Infect. Immun. 31: 1117. 25 Minoprio, P. M., Eisen, H., Formi, L., D’Imperio Lima, L. R., Joskowicz, M. and Coutinho, A., Scand. J. lmmunol. 1986.24: 661. 26 Petry, K. and Eisen, H., Parasitol. Today 1989. 5: 111. 27 Leech, J. H., Barnwell, J. W., Miller, L. H. and Howard. R. J., J. Exp. Med. 1984. 159: 1567. 28 Arauyo, F. G., J. Immunol. 1985. 135: 4149. 29 Williams, G. T., Fielder, L., Smith, H. and Hudson, L., Act. Trop. 1985. 42: 33. 30 Ribeiro dos Santos, R. and Hudson, L.. Parasit. lmmunol. 1980. 2: 1. 31 Davis, C. D. and Kuhn, R. E., Znfect. lrnmun. 1990. 58: 1812. 32 Pearson, T. W., Dolan,T. T., Stagg, D. A. and Lundin, L. B., Nature 1979. 281: 678. 33 Edelrnan, A. S. and Zolla-Pazner, S., FASEB J. 1989. 3: 22. 34 Rottenberg, M., Lindqvist, C., Koman, A., Segura, E. L. and Om, A., Scand. J. Immunol. 1989. 30: 65. 35 Serrano, L. E. and ODaly, J. A., lnt. J. Parasitol. 1987. 17: 851.
