Eur. J. Immunol. 1991. 21: 89-95
Dirk A. E. Dobbelaere, Isabel J. Roditi ThCrGse M. Coquerelle, Christina Kelke, Margarete Eichhorn and Richard 0.Williams Kernforschungszentrum Karlsruhe, lnstitut fur Genetik und Toxikologie, Karlsruhe
Cell-cell contact-dependent proliferation of 7: parva-infected cells
89
Lymphocytes infected with Theileria parva require both cell-cell contact and growth factor to proliferate* Lymphocytes infected with the intracellular parasite Theileria parva proliferate continuously as lymphoblastoid cell lines. We have previously shown that the continuous proliferation of the 7: parva-infected (Tpi) cell line TpM(803) is mediated in part by an autocrine mechanism (Dobbelaere, D. A. E. et al., Proc. Natl. Acad. Sci. USA 1988.85: 4730).We now report that continuous proliferation also requires surface stimulation through cell-cell contact. Under standard culture conditions this surface stimulus is provided by the infected cells themselves, but it can also be provided by uninfected lymphocytes or macrophages.The ability to respond to surface stimulation is critically dependent on the presence of the parasite in the host cell and is lost within 48 h after the elimination of the parasite from the host cell cytoplasm by treatment with the theilericidal drug BW720c. Tpi cells also secrete a growth factor which is able to support the proliferation of diluted Tpi cells. Growth factor secretion is rapidly lost upon elimination of the parasite. Moreover, inhibition experiments using antiinterleukin 2 (IL 2) antibodies show that IL 2 is involved in the proliferation of theTpi cell linesTpM(803) and IN10.Tcell proliferation is dependent on a number of costimulatory signals which are normally provided by accessory cells. The finding that Tpi cells can mutually stimulate each other to grow in the absence of conventional accessory cells helps to explain how they can escape the normal constraints on T cell growth, allowing them to invade and multiply in nonlymphoid as well as lymphoid tissues.
1 Introduction The protozoan parasite Theileria parva is transmitted by ticks and causes an acute and usually fatal disease of cattle known as East Coast fever (Theileriosis) (reviewed in [ l , 21). The disease is characterized by a massive lymphoproliferation followed by widespread lymphocytolysis with profound depletion of lymphocytes both in the solid lymphoid tissues and in the recirculating pool [3, 41. The latter is mediated by a nonspecific cytotoxic reaction directed against both infected and uninfected cells. 7: parva-infected (Tpi) cells express parasite-specific determinants which are recognized by MHC-restricted CTL [5-71 or Th cells [8-101 and it has been proposed that in immune animals infected cells are selectively destroyed by MHC class I-restricted CTL recognizing these antigens. When infected cells are isolated by LN biopsy, they can be grown indefinitely in culture as lymphoblastoid cell lines [ l l ] . Infected cell lines can also be generated in vitro by [I 87151
* This research was funded in part by a grant from the Deutsche Gescllschaft fur Technische Zusammenarbeit, supplied as Special Project Funding (No. 87.7860.7.01.100), supplement to the core research program of the International Laboratory for Research on Animal Diseases. Nairobi. Kenya. Correspondence: Dirk A. E. Dobbelaere, Universitat Bern, Institut fur Tierpathologie, Abt. Parasitologie, Langgass-StraBe 122. CH-3012 Bern, Switzerland Abbreviation: Tpi: 7: parva-infected (cells) 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1991
incubating lymphocytes with sporozoites isolated from infected ticks [ 121. Treatment of cultures with the theilericidal drug BW720c [13],which specifically kills the parasite, leads to growth arrest and reversion of the host cell phenotype to that of a normal resting lymphocyte [14]. When Tpi cells are injected into nude mice tumors are formed which can metastasize and invade different tissues [15]. In the infected animal,Tpi cells are first found in the regional LN draining the site of parasite inoculation. Parasitised cells spread not only to secondary lymphoid tissues, but also to the thymus, BM and a variety of non-lymphoid tissues, being particularly numerous in the lamina propria of the gastrointestinal tract and the interstitial tissues of the lungs. Foci of infected cells called “pseudoinfarcts” can sometimes also be found in the kidneys [ 161.
It has been shown that Tcell clones become independent of exogenous growth factors after infection with 7: parva [ 171. Tpi cells are sensitive to dilution, but their growth rate can largely be restored by the addition of human rIL 2 to the culture medium [14]. Consistent with this is the observation that all Tpi cell lines constitutively express functional IL 2R as long as the parasite is present in the host cell cytoplasm [14, 181. These findings, together with the report that SN from Tpi cell cultures stimulate DNA synsthesis in uninfected Tcell lines [ 171, all indicate that Tpi cells grow via an autocrine loop involving IL 2 and the IL 2R. Several observations, however, suggest that proliferation of Tpi does not depend On the concentration Of lymphokines in the culture medium, but also requires cell-cell contact. Cells from which the parasite had been eliminated by BW720c treatment continued to grow as OO14-2980/9 1/0101-O089$3.50 + .25/0
YO
D. A . E. Dobbelaere. I . J. Roditi,T. M. Coquerelle, C. Kelke et al.
rapidly asTpi cells over a period of 5 days, provided that the contact between them was not broken. When contact was disturbed, cured cells rapidly stopped growing and died [ l4].The addition of conditioned medium from logarithmically growing Tpi cells was unable to restore growth, but merely prevented the cured cells from dying. Tpi cells possess many characteristics of continuously activated Tcells [ 191. They contain elevated levels of PKC activity (A. Mastro, D. Grove, R. Williams and D. Dobbelaere, unpublished observation) which is most likely responsible for the constitutive activation of the nuclear factor NF-xB inTpi cells [20]. NF-.xB itself has been shown to be involved in the regulation of transcription of the IL 2R gene [21] which is constitutively expressed in all Tpi cell lines ([MI and D. Dobbelaere et al., submitted). Tpi cells also express the IFN-y gene (D. Dobbelaere, unpublished observation). Under normal conditions, antigen-induced T cell activation and proliferation requires two sets of signals. These are: (a) activation through the antigen-specific TcR and its associated molecules such as CD3, CD4 and CD8, and (b) antigen-independent costimulatory signals provided by activated APC expressing MHC class I1 (reviewed in [22-291). These signals do not only involve the adhesion molecules and IL 1. but also less well-defined costimulatory signals, requiring cell-cell contact, that are essential for progression into division. Without these costimulatory signals, cells become activated but fail to proliferate because lymphokine secretion is not induced [27]. In this report we show that Tpi cells. in contrast to normal Tcells, seem to bypass the requirement for activation through specific TcR/antigen-MHC interactions and also provide each other with costimulatory signals, thereby eliminating the need for accessory cells. We further show that Tpi cells secrete a growth factor which stimulates proliferation of diluted cell cultures. It is proposed that both cell-cell contact and lymphokines are required for the continuous proliferation of Tpi cells.
2 Materials and methods 2.1 Cells
Eur. J. Immunol. 1991. 21: 89-95
2.2 Stimulation assays
The following were used as stimulator cells: Tpi cells, allogeneic Con A-stimulated LN cells, M a or Tpi cells treated with BW720c for different times. Stimulator cells were prepared by exposing them to y-irradiation (5000 rad), washing them twice in Dulbecco's L15 (without additives) followed by one wash in PBS and fixation in gluteraldehyde (0.5% w/v) as described [33,34]. After fixation, stimulator cells were washed twice in L15 without additives and twice in cL15. Tpi responder cells originated from exponentially growing cultures and were washed twice in cL15 prior to use. Stimulation experiments were carried out in flat-bottom microtiter plates. Responder cells consisted of either Tpi cells, originating from the same culture from which the Tpi stimulator cells were prepared, or of cured cells treated with BW720c for 1, 2 or 3 days as indicated. Responder cells (500, 1000 or 2000 cells/well in 200 pl cL15 medium) were cultured in the presence of fixedhrradiated stimulator cells as indicated. After 3 days, proliferation of the responder cells was measured by pulse labeling cultures for 4 h with 3.5 pCi/ml ['HIdThd. The Stimulation Index shown in Fig. 1 was calculated by dividing the counts incorporated in the presence of 1.28 x 105 stimulator cells by the counts incorporated in the absence of stimulator cells.
-
For the preparation of membranes, 5 x 10' Tpi cells were washed, resuspended in 2ml cold TBS (25 mM Tris, 137 mM NaCl, 5 mM KCI, 0.45 mM CaC12,0.49 mM MgC12, pH 7.4) and submitted to hypotonic shock for 10 min by addition of 5 ml distilled water. Cells were disrupted in a Dounce homogenizer and overlaid on a discontinuous metrizamide gradient (10 ml25% ; 8 ml3% metrizamide in TBS). After 30 min centrifugation at 12500 x g, the 3%/25% interphase material was collected and washed twice (12000 x g, 15 min) in TBS, twice in cL15 and resuspended in 1.5 ml of cL15. Aliquots of the membrane preparation were added to Tpi responder cells as indicated.
The origin and maintenance of the cloned Tpi cell line TpM(803) used in this study have been described previously [30]. TpM(803) is a cloned T cell line infected with the Muguga strain of 7: parva and expressing the following surface phenotype: CD2+, CD4+. CD8-, IL 2R/Tac+, MHC class 11+.The cell line IN 10 is an uncloned Tpi cell line of the same phenotype which was generated by infection in vitro of Con A-stimulated LN cells with 7:parva (Muguga) sporozoites (D. Dobbelaere et al., submitted). Cells were cultured in Dulbecco's L15 medium containing 20 mM Hepes (pH 7. l), benzylpenicillin (100 units/ml), streptomycin (100 pg/ml), L-glutamine (20 mg/ml) 10% (v/v) heatinactivated FBS (Gibco Ltd, Paisley, Scotland) and 50 pM ?-ME. This medium is further referred to as cL15.
The mAb W6/32 (IgG2a) reacts with a monomorphic determinant on HLA-A, -B and -C molecules and recognizes a similar determinant on bovine MHC class I molecules [5, 71. Similarly mAb DAKO-HLA-DR (IgG2a. Dakopatts, Hamburg, FRG) reacts in a nonpolymorphic fashion with bovine MHC class I1 molecules [8]. Both antibodies were tested for their ability to inhibit the induction of Tpi responder cell proliferation by Tpi stimulator cells. Antibody-treated Tpi stimulator cells were prepared as described above, with the exception that they were incubated on ice for 30 min with saturating amounts of anti-class I , anti-class I1 or control IgG2a mAb prior to glutaraldehyde fixation.
Con A-stimulated LN cells [31] and bovine M a (321 were prepared as described. Tpi cells were cured of the parasite by treatment with 50 ng/ml of the theilericidal drug BW720c [ 13. 14, 181 for the indicated length of time. For cultivation of cured cells for more than 8 days, 10 U/ml of rIL 2 and SO ng/ml PMA were added to the medium.
2.3 Conditioned medium and growth factor assay Conditioned medium was prepared by culturing TpM(803) cells seeded at 5 x lo5 cells/ml for 2 days in RPMI containing 20 mM Hepes (pH 7.1). benzylpenicillin (100 units/ml), streptomycin (100 pg/ml). L-glutamine (20
Eur. J. Immunol. 19Yl. 21: 89-95
Cell-cell contact-dependent proliferation of 7:parva-infected cells
pg/ml), to which 1.2 mg/ml BSA was added. Cultures were then centrifuged and the SN filtered through a sterile
0.22-pm filter. Conditioned medium was also prepared from cured cells; for this purpose,TpM(803) cells were first treated with BW720c for 1, 2 or 3 days, washed to remove the drug and subsequently cultured for 2 days in RPMI/BSA.To test for growth factor activity,Tpi cells of the INlO cell line were washed twice and resuspended (1 x 10.' cells/ml) in 1-ml volumes of medium which consisted of 50% cL15 and. in the case of the control cultures, 50% RPMI to which 1.2 mg/ml of BSA was added.Test cultures were grown in 50% cL15 and 50% RPMI/BSA containing increasing amounts of conditioned medium, as indicated. Cells were cultured for 4 days and culture densities measured by Coulter counter. 2.4 Anti-IL 2 inhibition assay
The antibodies used in the anti-IL 2 inhibition experiments consisted of Ig isolated by protein A purification from serum (Genzyme, Munchen, FRG) of a rabbit immunized with human rIL 2. Control Ig were isolated in the same way from rabbit preimmune serum. Prior to use, protein A-purified Ig were first dialyzed against RPMI culture medium. TpM(803) and INlO cells were cultured in microtiter plates at 2.5 x 10.' cells/ml in 200 pl of RPMI containing 20 mM Hepes (pH 7. I), benzylpenicillin (100 units/ml), streptomycin (100 pg/ml), L-glutamine (20 pg/ml), 10% (v/v) heat-inactivated FBS (Gibco) and 50 pM 2-ME, to which 100 pg/ml of anti-IL 2 or control antibody was added. After 24 h, proliferation of the cells was measured by pulse labeling cultures for 3 h with 2.5 pCi/ml ["IdThd.
responder cells (Fig. 1B). The addition of Tpi membranes to dilute cultures of Tpi responder cells again resulted in a dose-dependent increase in proliferation. In control experiments, SN from concentrated membrane preparation did not stimulate the proliferation of responder cells. These data confirm that interactions between membrane-associated molecules on the surface of Tpi cells strongly enhance their proliferation and further show that intact cells are not required for this effect. 3.2 Uninfected cells also support proliferation The stimulus provided by the fixed Tpi cells could be parasite specific causing the activation of the TcR; alternatively, it could be a more general stimulus which is not restricted by MHC or antigen specificity [21, 281. To
(A)
responder cells Stim. per well: Index:
7
40
8.37 10
4.42 8
0
16
32
3.1 Proliferation of Tpi cells requires cell-cell contact
Membrane preparations from Tpi cells were also tested for their ability to stimulate proliferation of diluted Tpi
64
128
No. of Tpi stimulator cells x 10-3
3 Results
Preliminary experiments had shown that the growth of a constant number of celldm1 could be influenced by changing the surface area on which they rested. To distinguish between local gradients of growth factor and the effect of cell-cell contact, the following experiments were performed: dilute cultures of Tpi cells (responder cells) were cultured in the presence of increasing numbers of glutaraldehyde-fixedirradiated Tpi cells originating from the same culture (stimulator cells). Glutaraldehyde fixation was carried out to prevent the secretion of soluble factors and the internalization of factors secreted by the responder cells. Fig. 1 A shows that highly diluted Tpi cells grew extremely poorly, but that the addition of stimulator cells resulted in a dramatic increase in proliferation. The ability to respond to stimulator cells was greatly enhanced by the presence of 2-ME in the culture medium (data not shown). The control stimulator cells alone did not incorporate ['HIdThd and no increase in proliferation could be detected when responder cells were cultured in the SN from 2.56 x lo5 stimulator cells which had been incubated overnight in cL15 at 37 "C.The magnitude of the response was not only dependent on the number of stimulator Tpi cells, but also on the number of responder cells.
91
responder cells per well: 0 X
E
8 . Ei 8
150 125 100 75
C .-
u E
-5
50 25
0 -
1/54 1/18
i/6
112
Dilution of membrane preparation Figure I. Stimulation of the proliferation of T parva-infected cells by fixed Tpi cells (A) or their membranes (B). Tpi responder cells were cultured in microtiter wells in the presence of fixedhrradiatcd Tpi stimulator cells (A) or increasing amounts of a Tpi cell membrane preparation (B). After 3 days. DNA synthesis was measured by pulse labeling with ['HIdThd. (*): incorporation by 2000 responder cells cultured in SN from 2.56 x lo5stimulator cells which had been incubated overnight in cL15 at 37°C. ( 0 ) : incorporation by 2000 responder cells cultured in SN from an undiluted membrane preparation which had been incubated overnight at 37°C. Error bars represent 1 SD for triplicate cultures.
Eur. J. Immunol. 1991. 21: 89-95
D. A. E. Dobbelaere, I. J. R0diti.T. M. Coquerelle, C. Kelke et al.
92
Table 1. Stimulation of diluted Tpi responder cells by a variety of cell types % of stimulationa)obtained
Stimulator cells (1.28 x 1p)
using Qi stimulator cells A Tpi ceU line TpM(803) Con A-stimulated allogeneic bovine LN cells Tpi ceUs treated with BW720c for 3 days Tpi cells treated with BW720c for 10 days Tpi cells treated with BW720c for 13 days Tpi cells treated with BW720c for 19 days Bovine alveolar M@ B Q i cells pretreatedb)with control antibodies Qi cells pretreated anti-MHC class I antibodies Tpi cells pretreated anti-MHC class I1 antibodies
100.0 94.5 96.4 69.8 57.0 57.5 %.6 92.4 87.4 93.0
a) Tpi responder cclls (2WO/wcll)were cultured for 3 days with 1.28 X los fixed/irradiated stimulator cells of different origin and [3HldThd incorporation measured. h) Tpi stimulator cells were incuhated with saturating amounts of anti-MHC class I. anti-MHC class II or control IgG2a antibodies prior t o glutaraldehyde fixation.
(A)
(B)
120'
c-
100-
E 0 P
80 -
0 .-c
8c -e u = 8 .-
R = Bw720~
S = Tpi
60-
c
-G z -5 s
40-
examine whether the stimulus is specific for infected cells we tested the response of Tpi cells to uninfected leucocytes and cells from which the parasite had been eliminated by treatment with the theilericidal drug BW72Oc (see Table 1 A). Con A-stimulated LN cells and alveolar MQ, were as effective as Tpi cells in stimulating proliferation. Cells which had been treated for 3 days were also able to stimulate equally well. but the extent of stimulation declined with increasing time after treatment. These data show unequivocally that cells d o not need to be infected to stimulate Tpi cell proliferation. The stimulus is also not MHC restricted, since the uninfected stimulator cells were derived from different animals. These results would be consistent with the responder cells receiving co-stimulatory signals which are known to be independent of MHC type [28]. I n support of this, coating stimulator cells with anti-MHCclass I mAb,which had previously been shown to inhibit parasite-specific cytotoxic killing of Tpi cells [S-7]. o r anti-MHC class I1 mAb, which were shown to block Tpi-specific T h cell proliferation [8], did not affect their ability to drive proliferation of the responder cells (Table 1B).
20 -
0-
-
#
3.3 The ability to respond to surface stimulation is parasite dependent Once it had been established that the capacity t o stimulate was not parasite dependent, experiments were carried out to examine whether the ability t o respond to surface stimulation is dependent on the presence of viable parasites in the host cell cytoplasm. Cells were first treated for 3 days with BW720c to eliminate the parasite and then used as either stimulator o r responder cells. Fig. 2 A shows that although BW720c-cured cells remained capable of stimuIatingTpi responder cells, they completely failed to respond to stimulation at all concentrations tested. To examine the rate at which the ability to respond was lost, cured cells were stimulated starting 1, 2 o r 3 days after the onset of treatment with drug. ['HIdThd incorporation in response t o stimulation was reduced t o SO% in cells treated with
Cell-cell contact-dependent proliferation of 7:parva-infected cells
Eur. J. Immunol. 1991. 21: 89-95
93
BW720c for 1 day and to 7% within 2 days. Importantly, the lack of response was not due to a loss of viability or the inability of cured cells to proliferate, since cultures of cured cells could be maintained for up to 3 months in the presence of PMA and rIL 2 (data not shown). These data show that contact-dependent proliferation is controlled by the parasite at the level of the responder cell rather than the stimulator cell.
3.4 Growth factor dependence and secretion It has previously been shown that Tpi cells secrete small quantities of an IL 2-like growth factor [ 171 and it has been proposed that Tpi cells proliferate via an autocrine mechanism [ 14, 171. The control of growth thus appears to be multifactorial, involving both cell-cell contact and the secretion of a growth factor. Having established that responsiveness to surface stimuli depends o n the continuous presence of the parasite, we also examined the influence of the parasite on growth factor secretion. Although exogenous IL 2 has successfully been used to support the growth of diluted Tpi cultures, no proof has been presented until now that conditioned medium from Tpi cell cultures can support the proliferation of diluted Tpi cells nor that secreted IL 2 itself contributes to the proliferation of Tpi cell lines. Tpi cell lines can differ considerably in their sensitivity to dilution (I. Baumann and T. Coquerelle, unpublished observation). Cells from theTpi cell line IN 10are especially sensitive and their growth rate is severely reduced when the culture is diluted below the optimal density of los cells/ml. Since dilute INlO cells respond particularly well to the addition of rIL2, they consistute a good target cell line for testing growth factor activity. Fig. 3 shows that SN from TpM(803) cultures stimulate the growth of diluted INlO cells in a dosedependent manner.The replacement of 20% of the culture medium by conditioned medium resulted in a 25-fold increase in cell number compared to a 7-fold increase for the diluted cells over a period of 4 days (Fig. 3).The former
0 LL
no
1
2
3
Days of BW72Oc treatment
Figure 4. SN from cells treated with BW720c progressively lose their ability to stimulate proliferation. INlO cells were cultured in medium containing 20% conditioned medium fromTpM(803) cells treated with BW720c for 0, 1, 2 or 3 days. The vertical axis represents the factor obtained after dividing the increase in cell numbers of the test culture by the increase in cell numbers obtained for the control cultures. Error bars represent 1 SD of triplicate cultures.
corresponds to a division rate equivalent to that observed for INlO cells growing at normal density (approximately once every 20 h), the latter to one division every 34 h.The SN of TpM(803) cells from which the parasite had been eliminated by treatment with BW720c was also tested for growth factor activity. Conditioned medium originating from cells treated with BW720c for 1 day only was already less efficient at stimulating the proliferation of INlO cells than SN from infected cells (Fig. 4). By the 2nd day of treatment, SN induced only a 30% higher increase in cell numbers than that measured for the diluted control cells and by day 3 no stimulation could be measured. To investigate whether IL 2 contributes to the growth of Tpi cells we tested the effect of anti-IL2 antibodies on the growth of both TpM(803) and IN10 cells. Table 2 shows that, in both cases, anti-IL 2 antibodies caused a growth inhibition of almost 50% within 24 h of addition to the culture medium. No inhibition could be observed using control antibodies. From this we conclude that both cell lines secrete IL 2 and that it is needed for their growth. Table 2. Inhibition of Tpi cell proliferation by anti-IL 2 antibodies
0
5
10
20
30
40
50
% conditioned medium Figure3. The effect of conditioned medium on the growth of diluted Tpi INlO cells. Dilution-sensitive INlO Tpi cells (starting density 5 X 1@ celldml) were cultured in the presence of increasing amounts of conditioned medium from TpM(803) cells.The vertical axis represents the factor by which the cell numbers increased over 4 days of culture. Error bars represent 1 SD of triplicate cultures.
[3H]dThd incorporation (+ SD)a)
Cell line
Antibody
w4(803)
Control Ab Anti-I1 2
31 037
INlO
Control Ab Anti-IL 2
15 140
17512 7 040
(+ 3097) (+ 610) (* 360) (+ 426)
a) Tpi cells (5 x 103/well)were cultured for 24 h in the presence of control antibodies or anti-IL 2 antibodies and ["H]dThd incorporation measured (values are expressed in cpm k SD for quadruple observations).
04
D. A. E. Dobbelaere, I. J. Roditi,T. M. Coquerelle, C. Kelke et al.
Together the data show that the secretion of growth factor, like the ability to respond to surface stimulation, is parasite dependent and rapidly ceases upon elimination of the parasite from the cytoplasm. The data also show that IL 2 itself is involved in the proliferation of the TpM(803) and INlO cell lines, although it is not the only requirement for their growth.
4 Discussion Two components, cell-cell contact and the secretion of growth factor(s), including IL 2, are important for the autocrine proliferation of TpM(803) and INlO cells in vitro. The ability of Tpi cells to respond to surface stimulatory signals and to secrete growth factor(s) are both strictly dependent on the presence of the parasite in the host cell cytoplasm. The nature of the surface molecule(s) involved in the cell-cell surface interactions that induce proliferation is as yet unclear. Both the fact that a variety of different cells can fulfil this requirement and the fact that the stimulus is not MHC restricted suggest that the signal belongs to the group of co-stimulatory signals. Lymphokine secretion by activated T cells requires co-stimulatory signaling through surface interactions [22, 23, 281, and by analogy it is tempting to speculate that growth factor secretion by Tpi cells is also dependent on surface signals provided via cell-cell contact. The growth inhibition caused by culture dilution is thus likely to be due to a combination of insufficient growth factor and reduced surface signaling required for growth factor secretion.Within this context it is easy to envisage how the growth inhibition caused by dilution could largely be prevented by the addition of exogenous IL 2 [ 141 or conditioned medium. The notion that surface interactions and lymphokine secretion are interrelated is further supported by the fact that the down-regulation of both responsiveness to surface stimulation and growth factor secretion upon BW720c treatment occurs with similar kinetics. In addition, the increase in stimulation index observed with increasing responder cell concentrations (Fig. 1) is also consistent with the induction of growth factor secretion by the responder cells. The fact that antibodies directed against IL 2 could significantly inhibit the proliferation of TpM(803) and INlO cells provides strong support for a crucial role of IL 2 and IL 2R in the autocrine growth of TpM(803) and INlO cells [ 14,171. Moreover.we have recently obtained evidence that the IL 2 gene is transcribed, albeit at low levels, in most of the T.parva-infected cell lines tested (M. Eichhorn, Volker Heussler and D. A. E. Dobbelaere, manuscript in preparation). SN of TpM cells have in addition been shown to stimulate the growth of IL 2-dependent Tcell clones [ 171. The fact that anti-IL 2 antibodies did not inhibit proliferation completely could have several explanations. One possibility is that the affinity of the antibodies for IL 2 is much lower than the affinity of IL 2 for its receptor [35]. Alternatively, a number of IL 2 molecules may already associate with the IL 2R before reaching the cell surface as has been shown for the growth factor I L 3 (361. Finally, other growth factors in addition to I L 2 may also be involved .
Eur. J. Immunol. 1991. 21: 89-05
Under normal circumstances, the ability to respond to nonspecific co-stimulatory signals is subject to the delivery of a specific signal supplied by an antigen/MHC complex on the APC which is recognized by a clonally restricted TcR [22,23].The mechanism by which the requirement for this “specific priming” is fulfilled or bypassed in Tpi cells is not known, but the parasite could potentially interfere with the host cell at many different levels. The possibility that a parasite kinase may be instrumental in proliferation has recently been proposed [37]. SpecificTcR stimulation might conceivably be bypassed by altering the phosphorylation state of key molecules involved in the TcR signal-transduction pathways. Alternatively, other Tcell activation pathways that have been described [38] could also form a target for parasite-induced modifications. Continuous activation could also be achieved by secreted molecules that can associate with the cell surface. One explanation for the fact that uninfected stimulator cells are also capable of supporting the proliferation of the Tpi responder cells could be that antigens, released from the infected responder cells. are captured and presented back to the responder cells. Glutaraldehyde fixation of the stimulator cells, however, excludes the classical path of internalization and processing followed by presentation [33]. In addition, the observation that anti-MHC class I and class I1 mAb, which inhibited cytotoxic killing of Tpi cells [5-71 and the proliferation of Tpi-specific Th cells [S-lo], respectively, did not inhibit stimulation, also argues against classical antigen presentation. If secreted antigens were involved, they would have to be presented without prior internalization and processing. Examples of lymphocyte activation have been described that occur by such a mechanism. Antigens such as staphylococcal enterotoxins. toxic shock toxins [39-411, a product of M y c o p f a m a arthritidis [42] and the MIS antigens [43] have potent polyclonal mitogen activity for both murine and human T cells. Although parasite-derived soluble factors may be involved in activation, they do not appear to effect the response to surface stimuli. Conditioned medium from logarithmically growing Tpi cells, which could be expected to contain putative parasite-derived antigens, is unable to stimulate the growth of BW720c-treated cells [14]. This observation suggests that responsiveness depends on intervention by the parasite within the host for signal transduction allowing proliferation to occur. The fact that 7: parva-transformed cells secrete their own growth factors and can reciprocally provide co-stimulatory signals implies that once a Tpi cell is committed to divide. the daughter cells can further stimulate each other. Timelapse video recordings and microscopic observations have shown that after mitosis Tpi daughter cells remain closely associated for several hours providing ample opportunity for mutual stimulation (H. Spring and D. Dobbelaere, unpublished observation). I t has been reported that Th cells stimulated by receptor cross-linking release lymphokines in the direction of the stimulus [44]. Such a mechanism. in addition to surface stimulation, could explain the clonal growth of Tpi cells in soft agar or as microfoci in both lymphoid and non-lymphoid tissues [ 161. Our data show that a clear association exists between the transformed state of Tpi cells and their ability to respond to
Eur. J. Immunol. 1991.21: 89-95
Cell-cell contact-dependent proliferation of 7:parva-infected cells
surface signals. In contrast to permanently transformed cell lines the transformed state of Tpi cells can be reversed, allowing the phenomena of stimulation and responsiveness to be separated. Surface stimulation may play a much wider role than has generally been assumed and the possibility that similar mechanisms are involved in the continuous proliferation of other transformed cells deserves investigation. Ina Baumatin is thanked for providing the IN10 cell line, Cooper's Animal Health ( U K ) for BW720c, and F! Herrlich, C. Liicke, H. Rahmsdorf arid J. Gerhards for critical reading of the manuscript. Received July 6, 1990; in revised form September 7, 1990.
5 References 1 Morrison,W. I . , Lalor, P. A,, Goddeeris, B. M. and Teale, A. J., in Pearson, T. W. (Ed.), Parasite Antigens: Towards New Strategies for Vaccines, Marcel Dekker Inc., New York 1986, p. 167. 2 Irvin, A. D. and Morrison, W. I., in Soulsby, E. J. L. (Ed.). Immunopathology, Immunology and Immunoprophylaxis of Theileria infections, CRC Press, Boca Ratan 1987, p. 223. 3 Emery, D. L., Vet. Pathol. 1981. 9: 1. 4 Morrison, W. I., Buscher, G., Murray, M., Emery, D. L., Masake. R. H.. Cook, R. H. and Wells, PW., Exp. Parasitol. 1981. 52: 248. 5 Goddeeris, B. M., Morrison,W. I.,Teale, A. J., Bensaid, A . and Baldwin, C.. Proc. Natl. Acad. Sci. USA 1986. 83: 5238. 6 Morrison, W. I., Goddeeris, W. I. and Teale, A. J., Eur. J. Immunol. 1987. 17: 1703. 7 Morrison, W. I.. Goddeeris, B. M., Teale, A. J.. Groocock, C. M.. Kemp. S. J. and Stagg, D. A., Parasite Immunol. 1987. 9: 563. 8 Baldwin, C. I., Goddeeris, B. M. and Morrison,W. I., Parasite Immunol. 1987. 9: 499. 9 Brown, W. C.. Sugimoto. C. and Grab, D. J., Exp. Parasitol. 1989. 69: 234. 10 Brown, W. C., Lonsdale-Eccles, J. D., DeMartini. J. C. and Grab, D. J., J. Immunol. 1990. 144: 271. 11 Malmquist, W. A , , Nyindo, M. B. A. and Brown, C. G. D., Trop. Anim. Health Prod. 1970. 2: 139. 12 Br0wn.C. G. D.,Stagg, D. A., Purnell, R. E.,Kahn,G. K. and Payne, R. C.. Nature 1975. 245: 101. 13 Hudson, A.T., Randall, A. W., Fry, M., Ginger, C. D., Hill, B., Latter,V. S., McHardy, N. and Williams, R. B., Parasitology 1985. 90: 45. 14 Dobbelaere, D. A. E., Coquerelle, T. M., Roditi, I. J., Eichhorn, M. and Williams, R. O., Proc. Natl. Acad. Sci. USA 1988. 85: 4730. 15 Irvin, A. D., Brown, C. G. D., Kanhai, G. K. and Stagg, D. A.. Nature 1975. 255: 713. 16 Irvin, A. D. and Mwamachi, D. M., Vet. Rec. 1983. 113: 192.
95
17 Brown, W. C. and Logan, K. S., Parasite Immunol. 1986. 8: 189. 18 Coquerelle,T. M., Eichhorn, M., Magnuson. N. S.. Reeves, R.. Williams, R. A. and Dobbelaere, D. A. E.. Eur. J. Immunol. 1989. 19: 655. 19 Naessens, J., Newson, J., Bensaid, A.. Teale. A. J.. Magondu. J. G. and Black, S. J., J. Immunol. 1985. 135: 4183. 20 Ivanov, V., Stein, B., Baumann, I., Dobbelaere. D. A. E.. Herrlich, €? and Williams, R. O., Mol. Cell. Biol. 1989. 9: 4677. 21 Leung, K. and Nabel, G. J., Nature 1988. X?3: 776. 22 Mueller, D. L., Jenkins, M. K. and Schwartz. R. H.. Atinu. Rev. Immunol. 1989. 7: 445. 23 Weaver. C. T. and Unanue, E. R.. Immunol. Today 1990. 11: 49. 24 Weaver, C.T., Hawrylowicz, M. and Unanue. E. R.. Proc. Natl. Acad. Sci. USA 1988. 85: 8181. 25 Shivdasani. R. A. and Thomas, D. W.. J. Immunol. 1988. 141: 1252. 26 Kawakami, K.,Yamamoto,Y., Kakimoto, K. and Onoue. K., J. Immunol. 1989. 142: 1818. 27 Mueller, D. L., Jenkins, M. K. and Schwartz. R. H., J. Immunol. 1989. 142: 2617. 28 Jenkins, M. K., Pardoll, D. M., Mizuguchi. J.. Chused.T. M. and Schwartz, R. H., Proc. Natl. Acad. Sci. USA 1987. 84: 5409. 29 Jenkins, M. K., Ashwell. J. D. and Schwartz, R. H.. J. Immunol. 1988. 140: 3324. 30 Emery, D. L., Morrison,W. I., Nelson, R.T. and Murray, M.. in Irvin, A . E., Cunningham, M. I? and Young, A. S. (Eds.). Advances in the Control of Theileriosis, Nijhof. Thc Hague 1981, p. 295. 31 Mastro, A. M. and Pepin, K. G.. Cancer Res. 1980. 40: 3307. 32 Fox, M. L., Can. J. Microbiol. 1973. 19: 1207. 33 Shimonkevitz, R., Kappler, J., Marrack. €? and Grey, H.. J. Exp. Med. 1983. 158: 303. 34 Goddeeris, B. M. and Morrison,W. I., Immunology 1987. 60: 63. 35 Smith, K. A., Science 1988. 240: 1169. 36 Browder,T. M., Abrams, J. S.,Wong, P. M. C. and Nienhuis. A. W., Mol. Cell. Biol. 1989. 9: 204. 37 Ole-MoiYoi, 0. K., Exp. Parasitol. 1989. 69: 204. 38 Testi, R., Phillips, J. H. and Lanier, L. L., J. Immunol. 1989. 143: 1123. 38 Carlsson, R., Fischer, H. and Sjogren, H. 0..J. Immunol. 1988. 140: 2484. 40 Fleisher, B. and Schrenzenmeier, H., J. Exp. Med. 1988. 167: 1697. 41 Choi,Y., Kotzin, B., Herron, L., Callahan, J., Marrack. I? and Kappler, J., Proc. Natl. Acad. Sci. USA 1989. 86: 8941. 42 Lynch, D., Cole, B., Bluestone, J. and Hodes, R.. Eur. J. Immunol. 1988. 16: 747. 43 MacDonald, H. R., Schneider, R., Lees, R. K.. Howe, R. C.. Acha-Orbea, H., Festenstein, H., Zinkernagel, R. M. and Hengartner, H., Nature 1988. 332: 40. 44 Poo, W.-J., Conrad, L. and Janeway, C. A., Nature 1988.332: 378.
Dirk A. E. Dobbelaere, Isabel J. Roditi ThCrGse M. Coquerelle, Christina Kelke, Margarete Eichhorn and Richard 0.Williams Kernforschungszentrum Karlsruhe, lnstitut fur Genetik und Toxikologie, Karlsruhe
Cell-cell contact-dependent proliferation of 7: parva-infected cells
89
Lymphocytes infected with Theileria parva require both cell-cell contact and growth factor to proliferate* Lymphocytes infected with the intracellular parasite Theileria parva proliferate continuously as lymphoblastoid cell lines. We have previously shown that the continuous proliferation of the 7: parva-infected (Tpi) cell line TpM(803) is mediated in part by an autocrine mechanism (Dobbelaere, D. A. E. et al., Proc. Natl. Acad. Sci. USA 1988.85: 4730).We now report that continuous proliferation also requires surface stimulation through cell-cell contact. Under standard culture conditions this surface stimulus is provided by the infected cells themselves, but it can also be provided by uninfected lymphocytes or macrophages.The ability to respond to surface stimulation is critically dependent on the presence of the parasite in the host cell and is lost within 48 h after the elimination of the parasite from the host cell cytoplasm by treatment with the theilericidal drug BW720c. Tpi cells also secrete a growth factor which is able to support the proliferation of diluted Tpi cells. Growth factor secretion is rapidly lost upon elimination of the parasite. Moreover, inhibition experiments using antiinterleukin 2 (IL 2) antibodies show that IL 2 is involved in the proliferation of theTpi cell linesTpM(803) and IN10.Tcell proliferation is dependent on a number of costimulatory signals which are normally provided by accessory cells. The finding that Tpi cells can mutually stimulate each other to grow in the absence of conventional accessory cells helps to explain how they can escape the normal constraints on T cell growth, allowing them to invade and multiply in nonlymphoid as well as lymphoid tissues.
1 Introduction The protozoan parasite Theileria parva is transmitted by ticks and causes an acute and usually fatal disease of cattle known as East Coast fever (Theileriosis) (reviewed in [ l , 21). The disease is characterized by a massive lymphoproliferation followed by widespread lymphocytolysis with profound depletion of lymphocytes both in the solid lymphoid tissues and in the recirculating pool [3, 41. The latter is mediated by a nonspecific cytotoxic reaction directed against both infected and uninfected cells. 7: parva-infected (Tpi) cells express parasite-specific determinants which are recognized by MHC-restricted CTL [5-71 or Th cells [8-101 and it has been proposed that in immune animals infected cells are selectively destroyed by MHC class I-restricted CTL recognizing these antigens. When infected cells are isolated by LN biopsy, they can be grown indefinitely in culture as lymphoblastoid cell lines [ l l ] . Infected cell lines can also be generated in vitro by [I 87151
* This research was funded in part by a grant from the Deutsche Gescllschaft fur Technische Zusammenarbeit, supplied as Special Project Funding (No. 87.7860.7.01.100), supplement to the core research program of the International Laboratory for Research on Animal Diseases. Nairobi. Kenya. Correspondence: Dirk A. E. Dobbelaere, Universitat Bern, Institut fur Tierpathologie, Abt. Parasitologie, Langgass-StraBe 122. CH-3012 Bern, Switzerland Abbreviation: Tpi: 7: parva-infected (cells) 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1991
incubating lymphocytes with sporozoites isolated from infected ticks [ 121. Treatment of cultures with the theilericidal drug BW720c [13],which specifically kills the parasite, leads to growth arrest and reversion of the host cell phenotype to that of a normal resting lymphocyte [14]. When Tpi cells are injected into nude mice tumors are formed which can metastasize and invade different tissues [15]. In the infected animal,Tpi cells are first found in the regional LN draining the site of parasite inoculation. Parasitised cells spread not only to secondary lymphoid tissues, but also to the thymus, BM and a variety of non-lymphoid tissues, being particularly numerous in the lamina propria of the gastrointestinal tract and the interstitial tissues of the lungs. Foci of infected cells called “pseudoinfarcts” can sometimes also be found in the kidneys [ 161.
It has been shown that Tcell clones become independent of exogenous growth factors after infection with 7: parva [ 171. Tpi cells are sensitive to dilution, but their growth rate can largely be restored by the addition of human rIL 2 to the culture medium [14]. Consistent with this is the observation that all Tpi cell lines constitutively express functional IL 2R as long as the parasite is present in the host cell cytoplasm [14, 181. These findings, together with the report that SN from Tpi cell cultures stimulate DNA synsthesis in uninfected Tcell lines [ 171, all indicate that Tpi cells grow via an autocrine loop involving IL 2 and the IL 2R. Several observations, however, suggest that proliferation of Tpi does not depend On the concentration Of lymphokines in the culture medium, but also requires cell-cell contact. Cells from which the parasite had been eliminated by BW720c treatment continued to grow as OO14-2980/9 1/0101-O089$3.50 + .25/0
YO
D. A . E. Dobbelaere. I . J. Roditi,T. M. Coquerelle, C. Kelke et al.
rapidly asTpi cells over a period of 5 days, provided that the contact between them was not broken. When contact was disturbed, cured cells rapidly stopped growing and died [ l4].The addition of conditioned medium from logarithmically growing Tpi cells was unable to restore growth, but merely prevented the cured cells from dying. Tpi cells possess many characteristics of continuously activated Tcells [ 191. They contain elevated levels of PKC activity (A. Mastro, D. Grove, R. Williams and D. Dobbelaere, unpublished observation) which is most likely responsible for the constitutive activation of the nuclear factor NF-xB inTpi cells [20]. NF-.xB itself has been shown to be involved in the regulation of transcription of the IL 2R gene [21] which is constitutively expressed in all Tpi cell lines ([MI and D. Dobbelaere et al., submitted). Tpi cells also express the IFN-y gene (D. Dobbelaere, unpublished observation). Under normal conditions, antigen-induced T cell activation and proliferation requires two sets of signals. These are: (a) activation through the antigen-specific TcR and its associated molecules such as CD3, CD4 and CD8, and (b) antigen-independent costimulatory signals provided by activated APC expressing MHC class I1 (reviewed in [22-291). These signals do not only involve the adhesion molecules and IL 1. but also less well-defined costimulatory signals, requiring cell-cell contact, that are essential for progression into division. Without these costimulatory signals, cells become activated but fail to proliferate because lymphokine secretion is not induced [27]. In this report we show that Tpi cells. in contrast to normal Tcells, seem to bypass the requirement for activation through specific TcR/antigen-MHC interactions and also provide each other with costimulatory signals, thereby eliminating the need for accessory cells. We further show that Tpi cells secrete a growth factor which stimulates proliferation of diluted cell cultures. It is proposed that both cell-cell contact and lymphokines are required for the continuous proliferation of Tpi cells.
2 Materials and methods 2.1 Cells
Eur. J. Immunol. 1991. 21: 89-95
2.2 Stimulation assays
The following were used as stimulator cells: Tpi cells, allogeneic Con A-stimulated LN cells, M a or Tpi cells treated with BW720c for different times. Stimulator cells were prepared by exposing them to y-irradiation (5000 rad), washing them twice in Dulbecco's L15 (without additives) followed by one wash in PBS and fixation in gluteraldehyde (0.5% w/v) as described [33,34]. After fixation, stimulator cells were washed twice in L15 without additives and twice in cL15. Tpi responder cells originated from exponentially growing cultures and were washed twice in cL15 prior to use. Stimulation experiments were carried out in flat-bottom microtiter plates. Responder cells consisted of either Tpi cells, originating from the same culture from which the Tpi stimulator cells were prepared, or of cured cells treated with BW720c for 1, 2 or 3 days as indicated. Responder cells (500, 1000 or 2000 cells/well in 200 pl cL15 medium) were cultured in the presence of fixedhrradiated stimulator cells as indicated. After 3 days, proliferation of the responder cells was measured by pulse labeling cultures for 4 h with 3.5 pCi/ml ['HIdThd. The Stimulation Index shown in Fig. 1 was calculated by dividing the counts incorporated in the presence of 1.28 x 105 stimulator cells by the counts incorporated in the absence of stimulator cells.
-
For the preparation of membranes, 5 x 10' Tpi cells were washed, resuspended in 2ml cold TBS (25 mM Tris, 137 mM NaCl, 5 mM KCI, 0.45 mM CaC12,0.49 mM MgC12, pH 7.4) and submitted to hypotonic shock for 10 min by addition of 5 ml distilled water. Cells were disrupted in a Dounce homogenizer and overlaid on a discontinuous metrizamide gradient (10 ml25% ; 8 ml3% metrizamide in TBS). After 30 min centrifugation at 12500 x g, the 3%/25% interphase material was collected and washed twice (12000 x g, 15 min) in TBS, twice in cL15 and resuspended in 1.5 ml of cL15. Aliquots of the membrane preparation were added to Tpi responder cells as indicated.
The origin and maintenance of the cloned Tpi cell line TpM(803) used in this study have been described previously [30]. TpM(803) is a cloned T cell line infected with the Muguga strain of 7: parva and expressing the following surface phenotype: CD2+, CD4+. CD8-, IL 2R/Tac+, MHC class 11+.The cell line IN 10 is an uncloned Tpi cell line of the same phenotype which was generated by infection in vitro of Con A-stimulated LN cells with 7:parva (Muguga) sporozoites (D. Dobbelaere et al., submitted). Cells were cultured in Dulbecco's L15 medium containing 20 mM Hepes (pH 7. l), benzylpenicillin (100 units/ml), streptomycin (100 pg/ml), L-glutamine (20 mg/ml) 10% (v/v) heatinactivated FBS (Gibco Ltd, Paisley, Scotland) and 50 pM ?-ME. This medium is further referred to as cL15.
The mAb W6/32 (IgG2a) reacts with a monomorphic determinant on HLA-A, -B and -C molecules and recognizes a similar determinant on bovine MHC class I molecules [5, 71. Similarly mAb DAKO-HLA-DR (IgG2a. Dakopatts, Hamburg, FRG) reacts in a nonpolymorphic fashion with bovine MHC class I1 molecules [8]. Both antibodies were tested for their ability to inhibit the induction of Tpi responder cell proliferation by Tpi stimulator cells. Antibody-treated Tpi stimulator cells were prepared as described above, with the exception that they were incubated on ice for 30 min with saturating amounts of anti-class I , anti-class I1 or control IgG2a mAb prior to glutaraldehyde fixation.
Con A-stimulated LN cells [31] and bovine M a (321 were prepared as described. Tpi cells were cured of the parasite by treatment with 50 ng/ml of the theilericidal drug BW720c [ 13. 14, 181 for the indicated length of time. For cultivation of cured cells for more than 8 days, 10 U/ml of rIL 2 and SO ng/ml PMA were added to the medium.
2.3 Conditioned medium and growth factor assay Conditioned medium was prepared by culturing TpM(803) cells seeded at 5 x lo5 cells/ml for 2 days in RPMI containing 20 mM Hepes (pH 7.1). benzylpenicillin (100 units/ml), streptomycin (100 pg/ml). L-glutamine (20
Eur. J. Immunol. 19Yl. 21: 89-95
Cell-cell contact-dependent proliferation of 7:parva-infected cells
pg/ml), to which 1.2 mg/ml BSA was added. Cultures were then centrifuged and the SN filtered through a sterile
0.22-pm filter. Conditioned medium was also prepared from cured cells; for this purpose,TpM(803) cells were first treated with BW720c for 1, 2 or 3 days, washed to remove the drug and subsequently cultured for 2 days in RPMI/BSA.To test for growth factor activity,Tpi cells of the INlO cell line were washed twice and resuspended (1 x 10.' cells/ml) in 1-ml volumes of medium which consisted of 50% cL15 and. in the case of the control cultures, 50% RPMI to which 1.2 mg/ml of BSA was added.Test cultures were grown in 50% cL15 and 50% RPMI/BSA containing increasing amounts of conditioned medium, as indicated. Cells were cultured for 4 days and culture densities measured by Coulter counter. 2.4 Anti-IL 2 inhibition assay
The antibodies used in the anti-IL 2 inhibition experiments consisted of Ig isolated by protein A purification from serum (Genzyme, Munchen, FRG) of a rabbit immunized with human rIL 2. Control Ig were isolated in the same way from rabbit preimmune serum. Prior to use, protein A-purified Ig were first dialyzed against RPMI culture medium. TpM(803) and INlO cells were cultured in microtiter plates at 2.5 x 10.' cells/ml in 200 pl of RPMI containing 20 mM Hepes (pH 7. I), benzylpenicillin (100 units/ml), streptomycin (100 pg/ml), L-glutamine (20 pg/ml), 10% (v/v) heat-inactivated FBS (Gibco) and 50 pM 2-ME, to which 100 pg/ml of anti-IL 2 or control antibody was added. After 24 h, proliferation of the cells was measured by pulse labeling cultures for 3 h with 2.5 pCi/ml ["IdThd.
responder cells (Fig. 1B). The addition of Tpi membranes to dilute cultures of Tpi responder cells again resulted in a dose-dependent increase in proliferation. In control experiments, SN from concentrated membrane preparation did not stimulate the proliferation of responder cells. These data confirm that interactions between membrane-associated molecules on the surface of Tpi cells strongly enhance their proliferation and further show that intact cells are not required for this effect. 3.2 Uninfected cells also support proliferation The stimulus provided by the fixed Tpi cells could be parasite specific causing the activation of the TcR; alternatively, it could be a more general stimulus which is not restricted by MHC or antigen specificity [21, 281. To
(A)
responder cells Stim. per well: Index:
7
40
8.37 10
4.42 8
0
16
32
3.1 Proliferation of Tpi cells requires cell-cell contact
Membrane preparations from Tpi cells were also tested for their ability to stimulate proliferation of diluted Tpi
64
128
No. of Tpi stimulator cells x 10-3
3 Results
Preliminary experiments had shown that the growth of a constant number of celldm1 could be influenced by changing the surface area on which they rested. To distinguish between local gradients of growth factor and the effect of cell-cell contact, the following experiments were performed: dilute cultures of Tpi cells (responder cells) were cultured in the presence of increasing numbers of glutaraldehyde-fixedirradiated Tpi cells originating from the same culture (stimulator cells). Glutaraldehyde fixation was carried out to prevent the secretion of soluble factors and the internalization of factors secreted by the responder cells. Fig. 1 A shows that highly diluted Tpi cells grew extremely poorly, but that the addition of stimulator cells resulted in a dramatic increase in proliferation. The ability to respond to stimulator cells was greatly enhanced by the presence of 2-ME in the culture medium (data not shown). The control stimulator cells alone did not incorporate ['HIdThd and no increase in proliferation could be detected when responder cells were cultured in the SN from 2.56 x lo5 stimulator cells which had been incubated overnight in cL15 at 37 "C.The magnitude of the response was not only dependent on the number of stimulator Tpi cells, but also on the number of responder cells.
91
responder cells per well: 0 X
E
8 . Ei 8
150 125 100 75
C .-
u E
-5
50 25
0 -
1/54 1/18
i/6
112
Dilution of membrane preparation Figure I. Stimulation of the proliferation of T parva-infected cells by fixed Tpi cells (A) or their membranes (B). Tpi responder cells were cultured in microtiter wells in the presence of fixedhrradiatcd Tpi stimulator cells (A) or increasing amounts of a Tpi cell membrane preparation (B). After 3 days. DNA synthesis was measured by pulse labeling with ['HIdThd. (*): incorporation by 2000 responder cells cultured in SN from 2.56 x lo5stimulator cells which had been incubated overnight in cL15 at 37°C. ( 0 ) : incorporation by 2000 responder cells cultured in SN from an undiluted membrane preparation which had been incubated overnight at 37°C. Error bars represent 1 SD for triplicate cultures.
Eur. J. Immunol. 1991. 21: 89-95
D. A. E. Dobbelaere, I. J. R0diti.T. M. Coquerelle, C. Kelke et al.
92
Table 1. Stimulation of diluted Tpi responder cells by a variety of cell types % of stimulationa)obtained
Stimulator cells (1.28 x 1p)
using Qi stimulator cells A Tpi ceU line TpM(803) Con A-stimulated allogeneic bovine LN cells Tpi ceUs treated with BW720c for 3 days Tpi cells treated with BW720c for 10 days Tpi cells treated with BW720c for 13 days Tpi cells treated with BW720c for 19 days Bovine alveolar M@ B Q i cells pretreatedb)with control antibodies Qi cells pretreated anti-MHC class I antibodies Tpi cells pretreated anti-MHC class I1 antibodies
100.0 94.5 96.4 69.8 57.0 57.5 %.6 92.4 87.4 93.0
a) Tpi responder cclls (2WO/wcll)were cultured for 3 days with 1.28 X los fixed/irradiated stimulator cells of different origin and [3HldThd incorporation measured. h) Tpi stimulator cells were incuhated with saturating amounts of anti-MHC class I. anti-MHC class II or control IgG2a antibodies prior t o glutaraldehyde fixation.
(A)
(B)
120'
c-
100-
E 0 P
80 -
0 .-c
8c -e u = 8 .-
R = Bw720~
S = Tpi
60-
c
-G z -5 s
40-
examine whether the stimulus is specific for infected cells we tested the response of Tpi cells to uninfected leucocytes and cells from which the parasite had been eliminated by treatment with the theilericidal drug BW72Oc (see Table 1 A). Con A-stimulated LN cells and alveolar MQ, were as effective as Tpi cells in stimulating proliferation. Cells which had been treated for 3 days were also able to stimulate equally well. but the extent of stimulation declined with increasing time after treatment. These data show unequivocally that cells d o not need to be infected to stimulate Tpi cell proliferation. The stimulus is also not MHC restricted, since the uninfected stimulator cells were derived from different animals. These results would be consistent with the responder cells receiving co-stimulatory signals which are known to be independent of MHC type [28]. I n support of this, coating stimulator cells with anti-MHCclass I mAb,which had previously been shown to inhibit parasite-specific cytotoxic killing of Tpi cells [S-7]. o r anti-MHC class I1 mAb, which were shown to block Tpi-specific T h cell proliferation [8], did not affect their ability to drive proliferation of the responder cells (Table 1B).
20 -
0-
-
#
3.3 The ability to respond to surface stimulation is parasite dependent Once it had been established that the capacity t o stimulate was not parasite dependent, experiments were carried out to examine whether the ability t o respond to surface stimulation is dependent on the presence of viable parasites in the host cell cytoplasm. Cells were first treated for 3 days with BW720c to eliminate the parasite and then used as either stimulator o r responder cells. Fig. 2 A shows that although BW720c-cured cells remained capable of stimuIatingTpi responder cells, they completely failed to respond to stimulation at all concentrations tested. To examine the rate at which the ability to respond was lost, cured cells were stimulated starting 1, 2 o r 3 days after the onset of treatment with drug. ['HIdThd incorporation in response t o stimulation was reduced t o SO% in cells treated with
Cell-cell contact-dependent proliferation of 7:parva-infected cells
Eur. J. Immunol. 1991. 21: 89-95
93
BW720c for 1 day and to 7% within 2 days. Importantly, the lack of response was not due to a loss of viability or the inability of cured cells to proliferate, since cultures of cured cells could be maintained for up to 3 months in the presence of PMA and rIL 2 (data not shown). These data show that contact-dependent proliferation is controlled by the parasite at the level of the responder cell rather than the stimulator cell.
3.4 Growth factor dependence and secretion It has previously been shown that Tpi cells secrete small quantities of an IL 2-like growth factor [ 171 and it has been proposed that Tpi cells proliferate via an autocrine mechanism [ 14, 171. The control of growth thus appears to be multifactorial, involving both cell-cell contact and the secretion of a growth factor. Having established that responsiveness to surface stimuli depends o n the continuous presence of the parasite, we also examined the influence of the parasite on growth factor secretion. Although exogenous IL 2 has successfully been used to support the growth of diluted Tpi cultures, no proof has been presented until now that conditioned medium from Tpi cell cultures can support the proliferation of diluted Tpi cells nor that secreted IL 2 itself contributes to the proliferation of Tpi cell lines. Tpi cell lines can differ considerably in their sensitivity to dilution (I. Baumann and T. Coquerelle, unpublished observation). Cells from theTpi cell line IN 10are especially sensitive and their growth rate is severely reduced when the culture is diluted below the optimal density of los cells/ml. Since dilute INlO cells respond particularly well to the addition of rIL2, they consistute a good target cell line for testing growth factor activity. Fig. 3 shows that SN from TpM(803) cultures stimulate the growth of diluted INlO cells in a dosedependent manner.The replacement of 20% of the culture medium by conditioned medium resulted in a 25-fold increase in cell number compared to a 7-fold increase for the diluted cells over a period of 4 days (Fig. 3).The former
0 LL
no
1
2
3
Days of BW72Oc treatment
Figure 4. SN from cells treated with BW720c progressively lose their ability to stimulate proliferation. INlO cells were cultured in medium containing 20% conditioned medium fromTpM(803) cells treated with BW720c for 0, 1, 2 or 3 days. The vertical axis represents the factor obtained after dividing the increase in cell numbers of the test culture by the increase in cell numbers obtained for the control cultures. Error bars represent 1 SD of triplicate cultures.
corresponds to a division rate equivalent to that observed for INlO cells growing at normal density (approximately once every 20 h), the latter to one division every 34 h.The SN of TpM(803) cells from which the parasite had been eliminated by treatment with BW720c was also tested for growth factor activity. Conditioned medium originating from cells treated with BW720c for 1 day only was already less efficient at stimulating the proliferation of INlO cells than SN from infected cells (Fig. 4). By the 2nd day of treatment, SN induced only a 30% higher increase in cell numbers than that measured for the diluted control cells and by day 3 no stimulation could be measured. To investigate whether IL 2 contributes to the growth of Tpi cells we tested the effect of anti-IL2 antibodies on the growth of both TpM(803) and IN10 cells. Table 2 shows that, in both cases, anti-IL 2 antibodies caused a growth inhibition of almost 50% within 24 h of addition to the culture medium. No inhibition could be observed using control antibodies. From this we conclude that both cell lines secrete IL 2 and that it is needed for their growth. Table 2. Inhibition of Tpi cell proliferation by anti-IL 2 antibodies
0
5
10
20
30
40
50
% conditioned medium Figure3. The effect of conditioned medium on the growth of diluted Tpi INlO cells. Dilution-sensitive INlO Tpi cells (starting density 5 X 1@ celldml) were cultured in the presence of increasing amounts of conditioned medium from TpM(803) cells.The vertical axis represents the factor by which the cell numbers increased over 4 days of culture. Error bars represent 1 SD of triplicate cultures.
[3H]dThd incorporation (+ SD)a)
Cell line
Antibody
w4(803)
Control Ab Anti-I1 2
31 037
INlO
Control Ab Anti-IL 2
15 140
17512 7 040
(+ 3097) (+ 610) (* 360) (+ 426)
a) Tpi cells (5 x 103/well)were cultured for 24 h in the presence of control antibodies or anti-IL 2 antibodies and ["H]dThd incorporation measured (values are expressed in cpm k SD for quadruple observations).
04
D. A. E. Dobbelaere, I. J. Roditi,T. M. Coquerelle, C. Kelke et al.
Together the data show that the secretion of growth factor, like the ability to respond to surface stimulation, is parasite dependent and rapidly ceases upon elimination of the parasite from the cytoplasm. The data also show that IL 2 itself is involved in the proliferation of the TpM(803) and INlO cell lines, although it is not the only requirement for their growth.
4 Discussion Two components, cell-cell contact and the secretion of growth factor(s), including IL 2, are important for the autocrine proliferation of TpM(803) and INlO cells in vitro. The ability of Tpi cells to respond to surface stimulatory signals and to secrete growth factor(s) are both strictly dependent on the presence of the parasite in the host cell cytoplasm. The nature of the surface molecule(s) involved in the cell-cell surface interactions that induce proliferation is as yet unclear. Both the fact that a variety of different cells can fulfil this requirement and the fact that the stimulus is not MHC restricted suggest that the signal belongs to the group of co-stimulatory signals. Lymphokine secretion by activated T cells requires co-stimulatory signaling through surface interactions [22, 23, 281, and by analogy it is tempting to speculate that growth factor secretion by Tpi cells is also dependent on surface signals provided via cell-cell contact. The growth inhibition caused by culture dilution is thus likely to be due to a combination of insufficient growth factor and reduced surface signaling required for growth factor secretion.Within this context it is easy to envisage how the growth inhibition caused by dilution could largely be prevented by the addition of exogenous IL 2 [ 141 or conditioned medium. The notion that surface interactions and lymphokine secretion are interrelated is further supported by the fact that the down-regulation of both responsiveness to surface stimulation and growth factor secretion upon BW720c treatment occurs with similar kinetics. In addition, the increase in stimulation index observed with increasing responder cell concentrations (Fig. 1) is also consistent with the induction of growth factor secretion by the responder cells. The fact that antibodies directed against IL 2 could significantly inhibit the proliferation of TpM(803) and INlO cells provides strong support for a crucial role of IL 2 and IL 2R in the autocrine growth of TpM(803) and INlO cells [ 14,171. Moreover.we have recently obtained evidence that the IL 2 gene is transcribed, albeit at low levels, in most of the T.parva-infected cell lines tested (M. Eichhorn, Volker Heussler and D. A. E. Dobbelaere, manuscript in preparation). SN of TpM cells have in addition been shown to stimulate the growth of IL 2-dependent Tcell clones [ 171. The fact that anti-IL 2 antibodies did not inhibit proliferation completely could have several explanations. One possibility is that the affinity of the antibodies for IL 2 is much lower than the affinity of IL 2 for its receptor [35]. Alternatively, a number of IL 2 molecules may already associate with the IL 2R before reaching the cell surface as has been shown for the growth factor I L 3 (361. Finally, other growth factors in addition to I L 2 may also be involved .
Eur. J. Immunol. 1991. 21: 89-05
Under normal circumstances, the ability to respond to nonspecific co-stimulatory signals is subject to the delivery of a specific signal supplied by an antigen/MHC complex on the APC which is recognized by a clonally restricted TcR [22,23].The mechanism by which the requirement for this “specific priming” is fulfilled or bypassed in Tpi cells is not known, but the parasite could potentially interfere with the host cell at many different levels. The possibility that a parasite kinase may be instrumental in proliferation has recently been proposed [37]. SpecificTcR stimulation might conceivably be bypassed by altering the phosphorylation state of key molecules involved in the TcR signal-transduction pathways. Alternatively, other Tcell activation pathways that have been described [38] could also form a target for parasite-induced modifications. Continuous activation could also be achieved by secreted molecules that can associate with the cell surface. One explanation for the fact that uninfected stimulator cells are also capable of supporting the proliferation of the Tpi responder cells could be that antigens, released from the infected responder cells. are captured and presented back to the responder cells. Glutaraldehyde fixation of the stimulator cells, however, excludes the classical path of internalization and processing followed by presentation [33]. In addition, the observation that anti-MHC class I and class I1 mAb, which inhibited cytotoxic killing of Tpi cells [5-71 and the proliferation of Tpi-specific Th cells [S-lo], respectively, did not inhibit stimulation, also argues against classical antigen presentation. If secreted antigens were involved, they would have to be presented without prior internalization and processing. Examples of lymphocyte activation have been described that occur by such a mechanism. Antigens such as staphylococcal enterotoxins. toxic shock toxins [39-411, a product of M y c o p f a m a arthritidis [42] and the MIS antigens [43] have potent polyclonal mitogen activity for both murine and human T cells. Although parasite-derived soluble factors may be involved in activation, they do not appear to effect the response to surface stimuli. Conditioned medium from logarithmically growing Tpi cells, which could be expected to contain putative parasite-derived antigens, is unable to stimulate the growth of BW720c-treated cells [14]. This observation suggests that responsiveness depends on intervention by the parasite within the host for signal transduction allowing proliferation to occur. The fact that 7: parva-transformed cells secrete their own growth factors and can reciprocally provide co-stimulatory signals implies that once a Tpi cell is committed to divide. the daughter cells can further stimulate each other. Timelapse video recordings and microscopic observations have shown that after mitosis Tpi daughter cells remain closely associated for several hours providing ample opportunity for mutual stimulation (H. Spring and D. Dobbelaere, unpublished observation). I t has been reported that Th cells stimulated by receptor cross-linking release lymphokines in the direction of the stimulus [44]. Such a mechanism. in addition to surface stimulation, could explain the clonal growth of Tpi cells in soft agar or as microfoci in both lymphoid and non-lymphoid tissues [ 161. Our data show that a clear association exists between the transformed state of Tpi cells and their ability to respond to
Eur. J. Immunol. 1991.21: 89-95
Cell-cell contact-dependent proliferation of 7:parva-infected cells
surface signals. In contrast to permanently transformed cell lines the transformed state of Tpi cells can be reversed, allowing the phenomena of stimulation and responsiveness to be separated. Surface stimulation may play a much wider role than has generally been assumed and the possibility that similar mechanisms are involved in the continuous proliferation of other transformed cells deserves investigation. Ina Baumatin is thanked for providing the IN10 cell line, Cooper's Animal Health ( U K ) for BW720c, and F! Herrlich, C. Liicke, H. Rahmsdorf arid J. Gerhards for critical reading of the manuscript. Received July 6, 1990; in revised form September 7, 1990.
5 References 1 Morrison,W. I . , Lalor, P. A,, Goddeeris, B. M. and Teale, A. J., in Pearson, T. W. (Ed.), Parasite Antigens: Towards New Strategies for Vaccines, Marcel Dekker Inc., New York 1986, p. 167. 2 Irvin, A. D. and Morrison, W. I., in Soulsby, E. J. L. (Ed.). Immunopathology, Immunology and Immunoprophylaxis of Theileria infections, CRC Press, Boca Ratan 1987, p. 223. 3 Emery, D. L., Vet. Pathol. 1981. 9: 1. 4 Morrison, W. I., Buscher, G., Murray, M., Emery, D. L., Masake. R. H.. Cook, R. H. and Wells, PW., Exp. Parasitol. 1981. 52: 248. 5 Goddeeris, B. M., Morrison,W. I.,Teale, A. J., Bensaid, A . and Baldwin, C.. Proc. Natl. Acad. Sci. USA 1986. 83: 5238. 6 Morrison, W. I., Goddeeris, W. I. and Teale, A. J., Eur. J. Immunol. 1987. 17: 1703. 7 Morrison, W. I.. Goddeeris, B. M., Teale, A. J.. Groocock, C. M.. Kemp. S. J. and Stagg, D. A., Parasite Immunol. 1987. 9: 563. 8 Baldwin, C. I., Goddeeris, B. M. and Morrison,W. I., Parasite Immunol. 1987. 9: 499. 9 Brown, W. C.. Sugimoto. C. and Grab, D. J., Exp. Parasitol. 1989. 69: 234. 10 Brown, W. C., Lonsdale-Eccles, J. D., DeMartini. J. C. and Grab, D. J., J. Immunol. 1990. 144: 271. 11 Malmquist, W. A , , Nyindo, M. B. A. and Brown, C. G. D., Trop. Anim. Health Prod. 1970. 2: 139. 12 Br0wn.C. G. D.,Stagg, D. A., Purnell, R. E.,Kahn,G. K. and Payne, R. C.. Nature 1975. 245: 101. 13 Hudson, A.T., Randall, A. W., Fry, M., Ginger, C. D., Hill, B., Latter,V. S., McHardy, N. and Williams, R. B., Parasitology 1985. 90: 45. 14 Dobbelaere, D. A. E., Coquerelle, T. M., Roditi, I. J., Eichhorn, M. and Williams, R. O., Proc. Natl. Acad. Sci. USA 1988. 85: 4730. 15 Irvin, A. D., Brown, C. G. D., Kanhai, G. K. and Stagg, D. A.. Nature 1975. 255: 713. 16 Irvin, A. D. and Mwamachi, D. M., Vet. Rec. 1983. 113: 192.
95
17 Brown, W. C. and Logan, K. S., Parasite Immunol. 1986. 8: 189. 18 Coquerelle,T. M., Eichhorn, M., Magnuson. N. S.. Reeves, R.. Williams, R. A. and Dobbelaere, D. A. E.. Eur. J. Immunol. 1989. 19: 655. 19 Naessens, J., Newson, J., Bensaid, A.. Teale. A. J.. Magondu. J. G. and Black, S. J., J. Immunol. 1985. 135: 4183. 20 Ivanov, V., Stein, B., Baumann, I., Dobbelaere. D. A. E.. Herrlich, €? and Williams, R. O., Mol. Cell. Biol. 1989. 9: 4677. 21 Leung, K. and Nabel, G. J., Nature 1988. X?3: 776. 22 Mueller, D. L., Jenkins, M. K. and Schwartz. R. H.. Atinu. Rev. Immunol. 1989. 7: 445. 23 Weaver. C. T. and Unanue, E. R.. Immunol. Today 1990. 11: 49. 24 Weaver, C.T., Hawrylowicz, M. and Unanue. E. R.. Proc. Natl. Acad. Sci. USA 1988. 85: 8181. 25 Shivdasani. R. A. and Thomas, D. W.. J. Immunol. 1988. 141: 1252. 26 Kawakami, K.,Yamamoto,Y., Kakimoto, K. and Onoue. K., J. Immunol. 1989. 142: 1818. 27 Mueller, D. L., Jenkins, M. K. and Schwartz. R. H., J. Immunol. 1989. 142: 2617. 28 Jenkins, M. K., Pardoll, D. M., Mizuguchi. J.. Chused.T. M. and Schwartz, R. H., Proc. Natl. Acad. Sci. USA 1987. 84: 5409. 29 Jenkins, M. K., Ashwell. J. D. and Schwartz, R. H.. J. Immunol. 1988. 140: 3324. 30 Emery, D. L., Morrison,W. I., Nelson, R.T. and Murray, M.. in Irvin, A . E., Cunningham, M. I? and Young, A. S. (Eds.). Advances in the Control of Theileriosis, Nijhof. Thc Hague 1981, p. 295. 31 Mastro, A. M. and Pepin, K. G.. Cancer Res. 1980. 40: 3307. 32 Fox, M. L., Can. J. Microbiol. 1973. 19: 1207. 33 Shimonkevitz, R., Kappler, J., Marrack. €? and Grey, H.. J. Exp. Med. 1983. 158: 303. 34 Goddeeris, B. M. and Morrison,W. I., Immunology 1987. 60: 63. 35 Smith, K. A., Science 1988. 240: 1169. 36 Browder,T. M., Abrams, J. S.,Wong, P. M. C. and Nienhuis. A. W., Mol. Cell. Biol. 1989. 9: 204. 37 Ole-MoiYoi, 0. K., Exp. Parasitol. 1989. 69: 204. 38 Testi, R., Phillips, J. H. and Lanier, L. L., J. Immunol. 1989. 143: 1123. 38 Carlsson, R., Fischer, H. and Sjogren, H. 0..J. Immunol. 1988. 140: 2484. 40 Fleisher, B. and Schrenzenmeier, H., J. Exp. Med. 1988. 167: 1697. 41 Choi,Y., Kotzin, B., Herron, L., Callahan, J., Marrack. I? and Kappler, J., Proc. Natl. Acad. Sci. USA 1989. 86: 8941. 42 Lynch, D., Cole, B., Bluestone, J. and Hodes, R.. Eur. J. Immunol. 1988. 16: 747. 43 MacDonald, H. R., Schneider, R., Lees, R. K.. Howe, R. C.. Acha-Orbea, H., Festenstein, H., Zinkernagel, R. M. and Hengartner, H., Nature 1988. 332: 40. 44 Poo, W.-J., Conrad, L. and Janeway, C. A., Nature 1988.332: 378.
