Immunology Letters, 43 (1994) 39-43 0165-2478/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved

IMLET 2229

Dendritic cells from HIV-1 infected individuals show reduced capacity to stimulate autologous T-cell proliferation

Mary Roberts b, Mark Gompels c, Anthony J. Pinching d and Stella C. Knight a,*

a Antigen Presentation Research Group, St Mary's Hospital Medical School, at Northwick Park Institute for Medical Research, Watford Road, Harrow, UK; b Sandoz Clin. Dev. Centre, Frimley, Camberley, Surrey, UK; c Newcastle General Hospital, Dept. of Immunology, Newcastle, UK;

d St Bartholomew's Hospital Medical College, Dept. of Immunology, London, UK

(Received 10 August 1994; accepted 17 August 1994)

Keywords: Autologous mixed leukocyte reaction; Pathogenesis of HIV-1 infection; Antibody production in HIV infection; Cell-mediated immunity in HIV-1 infection; Dendritic cells in HIV-1 infection

1. Summary 2. Introduction

Dendritic cells (DC) exposed for a short period of time (1 day) in vitro to HIV infection caused stimula- tion of autologous T-cells, but those exposed for a longer period (3 days) before addition to T-cells stimu- late little or no proliferation [1]. Since a proportion of DC from patients who are HIV-1 seropositive is in- fected with HIV-1 [2] we used the same culture system to test the level of ongoing stimulation of autologous T-cells by DC. The DC from patients with HIV-1 infection showed no enhanced stimulation of autologous T-cells; they produced significantly lower levels of proliferation than those induced by DC from normal controls or from high risk seronegative individuals. However, DC stimulated production of antibodies to gpl20 and p24 in cultures containing autologous B plus T-cell in the absence of any added exogenous antigens as previously described [3]. DC in asymptomatic indi- viduals thus appear to be fuelling antibody production but not T-cell proliferation. The results suggest that a failure by DC to stimulate T-cell proliferation either to HIV-1 or to other environmental antigens may be in- volved in the failure of cell-mediated responses in HIV infection.

* Corresponding author: Dr. S.C. Knight, Antigen Presentation Re- search Group, St Mary's Hospital Medical School, at Northwick Park Institute for Medical Research, Watford Road, Harrow, HA1 3UJ, UK.

Immunological studies on the blood leucocytes of individuals infected with HIV-1 show progressive loss in the capacity to produce cell-mediated immune re- sponses [4]. However, concomitant with the loss of cell-mediated responses there is a persistent production of antibody to HIV itself [5,6]. This persistent antibody production in the face of declining cell-mediated immu- nity is accompanied by the production of cytokines which suggest that there is preferential stimulation of a TH2 pattern of stimulation rather than TH1 [7,8]. Our earlier studies indicated that the pattern of T-cell stimu- lation may revolve around the signalling via DC. Thus, the DC failed to stimulate T-cell proliferation to anti- gens and this effect was seen even when cells were from asymptomatic seropositive individuals where naive and memory T-cells and monocytes functioned nor- mally in lymphocyte proliferation assays [9]. However, DC taken from HIV infected asymptomatic patients and cultured without addition of exogenous antigen were able to stimulate antibody production to HIV-1 in autol- ogous lymphocytes [2]. Here we used the same 20 /.d culture system that allowed the development of anti- body production and tested the capacity of DC to stimulate autologous T-cell proliferation. We could also use this system to produce primary T-cell proliferation in response to DC pulsed with HIV-1 for 24 hours using cells from HIV seronegative individuals. However, us- ing cells from HIV-1 infected individuals we failed to see stimulation of autologous T-cells despite the persis-

SSDI 0165-2478(94)00147-2 39

tent stimulation of antibody production. This reflects the in vivo picture and suggests a role for DC signalling in the effects of HIV-1 on the immune system.

3. Materials and Methods

3.1. Patient material

Peripheral blood was obtained from homosexual men who were HIV seropositive and who were not undergo- ing treatment, seronegative with a high risk of part exposure to HIV or normal laboratory controls.

3.2. Cell separations

Human peripheral blood mononuclear cells, isolated using Ficoll gradients, were washed in medium (RPMI- 1640, Dutch modification with 100 IU penicillin, 100 mg/ml streptomycin and 10% fetal calf serum) and incubated overnight in tissue culture flasks before the adherent cells were thoroughly washed to remove non- adherent and lightly adherent cells. The firmly adherent cells were gently scraped off the flask to provide a population of MO. Enriched DC were isolated from the non-adherent fraction by centrifuging at 600 g for 10 min on metrizamide gradients (Nyegaard, Oslo 13.7% w / v in RPMI + 10% FCS). These cells in normals were around 30% DC, the rest being mainly monocytes with < 5% contamination by lymphocytes [9]. In some experiments, these cells were further purified, either by panning on human gammaglobulin coated plates to remove Fc receptor positive monocytes, or by panning on anti-mouse immunoglobulin coated plates (Applied Immune Sciences Inc., Menlo Park, CA) after labelling the cells with antibodies to CD14 (anti-MO), Becton- Dickinson, Mountain View, CA). CD19 (anti-B cell Dakopatts A / S , Glostrnp, Denmark) and CD3 (pan T-cell, Becton-Dickinson). The cells purified by the latter method are consistently > 95% pure as assessed by alkaline phosphatase/anti-alkaline phosphatase staining of antibody-labelled cells after the panning step. Routinely, DC were identified and counted using light microscopy of live cells at 37°C by morphological characteristics such as frequently moving veiled projec- tions.

The cells from the pellets of the metrizamide gradi- ents were diluted from the hypertonic metrizamide by drop-wise addition of medium and washed to provide a population of lymphocytes depleted of DC and MO.

3.3. Hanging drop culture

Cells isolated from peripheral blood were cultured in 20 /xl hanging drops in Terasaki plates [10]. Graded

40

numbers of lymphocytes (25000-100000) were cul- tured either alone or with 1000 autologous DC or MO. Antibody responses were measured after 9 days of culture by freezing and storing cultures at -70°C until the ELISA for HIV-specific antibodies could be per- formed. Some cultures received DC that were exposed for 1 or 3 days to 105 TCID of HIV-1 (Strain IIIB). Proliferative responses were measured after 5 days by uptake of 3H-thymidine (1 /xg/ml, 2 Ci/mmole, Amersham International, Amersham, Bucks) in a 2 hour pulse. These conditions gave low counts but since the thymidine remained in flooding conditions and was of low specific activity these reflected the DNA synthesis rather than availability of precursor or radiation damage of dividing cells [10].

3.4. ELISA

Immulon 2 96-well plates (Dynatech Laboratories Inc., Chantilly, VA) were coated overnight at 37°C with recombinant gpl20 or p24 (MRC, AIDS-Directed Pro- gramme, Baculovirus-derived) using 50 ~ l /we l l of a 2.5 /xg/ml solution in sodium carbonate/bicarbonate buffer pH 9.6. Plates were washed with PBS-Tween (0.05% Tween 20) then blocked with 2% gelatine in PBS at 37°C for 1 hour. Plates were washed twice with PBS-Tween before adding the samples. Cultures were thawed and 15 /.tl from triplicate wells added to 35 /.tl of Triton in PBS (final concentration 0.05% to inacti- vate virus) and incubated at 37°C for 2 hours. Plates were washed three times before addition of conjugate (1/2000 anti-human gammaglobulin conjugated to per- oxidase, Sigma Chemicals) for 2 hours at 37°C. Sub- strate (O-Phenylenediamine in citrate buffer) was added to the plates after five washes with PBS-Tween and left to develop for 30 mins at room temperature. The reac- tion was stopped with 4 N H2SO 4 and the absorbance read at 492 nm. Results were zeroed against medium controls.

3.5. Statistical analysis

Significant differences were assessed from analysis of variance and Student's t-tests [10].

4. Results

4.1. In vitro infection with HIV-1

DC isolated from normal non-infected individuals were cultured for 3 days in vitro with HIV present for the whole 3 days or for the last 24 hours. Although both infected and non-infected DC showed similar viability

80C

60C

40C

2OC

~ - - - - - _ _ . _ . _ . _ _ . . ~

[3

I

.L 25 50 180

C e l l s / w e l l x 10 .3

Fig. 1. DC from a normal individual were exposed to HIV-1 (IIIB, 105 TCID) for 3 days or for the last day of a 3 day culture. DC (1000) were added to different numbers of autologous lymphocytes in 20/.tl hanging drops for a further 4 days and the uptake of 3H-thymidine was measured. T-cells alone (Q); T-cells+DC (O); T-cells+DC exposed to HIV-1 for 1 day ( • ) and T-cells + DC exposed to HIV-1 for 3 days ([]).

at the end of the 3 day culture period, the DC exposed for 1 day to HIV caused high levels of stimulation of T-cell proliferation but the 3 day exposed cells caused significantly lower levels of proliferation (Fig. 1). As previously reported [1] the capacity of DC to stimulate responses to the virus in autologous T-cells was depen- dent on the time of exposure to the HIV-1 in vitro. Since a proportion of DC have been shown to be infected in HIV infected individuals, the capacity of DC from these infected individuals to stimulate autologous T-cell proliferation was assessed using this culture sys- tem which can support primary stimulation of T-cells.

4.2. In vivo exposure to HIV-1

As demonstrated in Fig. 2, DC taken from normal individuals or from individuals at high risk of infection with HIV-1 usually caused autologous T-cell stimula- tion. By comparison cells taken from HIV-1 infected individuals rarely showed stimulation of autologous T-cells. The enhancement of proliferation in lympho- cytes on addition of DC was statistically significant in the controls taken as a group but not in the HIV- seropositive individuals.

The lack of stimulation of T-cell proliferation by DC from seropositive individuals contrasted sharply with the capacity of these same DC to stimulate significant

1.5 11 I I I ~ " High r isk Low r isk se ronega t i ve

1.O g-

LY IX: ~ De kY ~ ~ I~ LY IX:

Fig. 2. Effect of autologous DC on lymphocyte proliferation. Donors were HIV seropositives separated according to Centre for Disease Control classification (Asymptomatic, II; persistent generalised lymphadenopathy, III; AIDS patients, IV), and seronegatives at high risk or low risk for HIV infection. 100000 lymphocytes were cultured alone (LY) or with 1000 autologous dendritic cells (DC). Proliferation was measured by uptake of 3H-thymidine in a 2 hour pulse after 5 days of culture.

levels of antibody production in the lymphocyte popula- tion [3]. Fig. 3 shows an example of stimulation of antibody production by DC from an asymptomatic HIV-seropositive individual. DC but not monocytes stimulated significant levels of antibody production both to gp120 and to p24.

1.C

0.8

~ O.6

.o

"( Q4

gp120

.__./

25 5O 100

p 24

j J

Z/ IO0

Cells / well

Fig. 3. Effect of autologous APC on HIV-specific antibody produc- tion by cultured lymphocytes from a seropositive asymptomatic indi- vidual. Graded numbers of lymphocytes (25000-100000) were cul- tured alone ( • ) or with 1000 autologous DC ( • ) or 1000 autologous MO (O). Antibodies to gp120 and p24 were measured (OD values) after 9 days in culture.

41

5. Discussion

DC isolated from HIV-seropositive individuals stim- ulated little or no proliferation in autologous T-cells causing significantly less stimulation than that seen using cells from either normal controls or from seroneg- ative individuals at high risk for infection with HIV-1. This work confirms and extends the observations made by Gupta and Safai [11], Puppo et al. [12] and Garbrecht et al. [13] indicating that the autologous mixed leuco- cyte reaction (AMLR) between T-cells and "non-T cells" was reduced in HIV-seropositive individuals. The nature of the AMLR has been a subject of consid- erable speculation. It is induced by DC and blocked by antibodies to MHC class II molecules [14], and thus has characteristics of presentation of antigen to T-cells. In mice that were specific pathogen free the autologous stimulation of T-cells by DC was negligible. Following either deliberate exposure to antigen [15] or environ- mental exposure to Sendai virus (unpublished observa- tions) the level of autologous stimulation increased suggesting that it reflects largely ongoing presentation of antigens. If this is the case, measurement of AMLR may be used to probe ongoing presentation of antigens to T-cells in different patient groups.

Exposure of DC to increasing doses of either human T-cell leukaemia virus type I (HTLV-1) or to HIV-1 in vitro and addition to autologous T-cells results in dose- dependent stimulation and then suppression of prolifera- tion [16,1]. This may be the basis for the data shown in Figure 1 where increased exposure time of DC to HIV-1 results in the loss of stimulation of T-cells in response to HIV-1. By contrast, although some DC are known to be infected with HIV-1 in vivo, there was no evidence of stimulation of T-cell proliferation but only suppression of responses, perhaps paralleling the high dose/longer exposure time effects seen with HIV-1 in vitro. However, in HTLV-1 infection in vivo, high levels of T-cell proliferation were observed. This was originally believed to reflect a "spontaneous" prolifera- tion caused by direct exposure of T-cells to the virus, but there is now evidence that this proliferation is stimulated by DC, some of which are infected with HTLV-1 [17]. This may reflect the lower dose of anti- gen which may also cause stimulation following in vitro exposure to this virus. More precise delineation of the doses of these retroviruses which result in either stimu- lation or suppression of T-cells via DC is currently being sought.

The low level of AMLR in patients with asymp- tomatic infection contrasts with the persistent stimula- tion of antibody production caused by DC from these patients. The immunoglobulin production in HIV-1 in- fection was originally believed to be a consequence of

42

"spontaneous" stimulation by HIV-1, but the antibody was later shown to be mainly specific for HIV-1 [18]. Our more recent studies indicate that this effect is fuelled via DC [3, Fig. 3]. As DC are believed to interact largely with T-cells the assumption is that this effect on antibody production is acting via T-cell help in the cultures. The lack of proliferation of T-cells does not necessarily argue against this involvement since cytokine production and proliferation do not necessarily occur in tandem. The requirement for DC to see suc- cessful production of antibody in vitro has already been reported [19]. However, the involvement of T-cells in this current system, the specificity of the signal pre- sented by the DC and the nature of the antibodies produced remain to be determined. There is evidence that low doses of antigen may favour the production of cell-mediated immunity to HIV and that higer doses may preferentially cause humoral immunity [8]. Such dosimetric effects would explain the pattern of re- sponses seen in our data. Whatever the precise mecha- nisms involved, our data support the view that sig- nalling through DC may be of fundamental importance to the development of the immune picture resulting from HIV-1 infection.

Acknowledgements

Mary Roberts was supported by the AIDS Directed Programme of the UK Medical Research Council.

References

[1] Knight, S.C., Macatonia, S.E. and Patterson, S. (1993) Res. Virol. 144, 75.

[2] Patterson, S., Roberts, M.S., English, N.R., Macatonia, S.E., Gompels, M.N., Pinching, A.J. and Knight, S.C. (1994) Res. Virol. (in press)

[3] Roberts, M., Gompels, M., Pinching, A.J. and Knight, S.C. (1994) AIDS (in press)

[4] Clerici, M., Landay, A.L., Kessler, H.A., Zajac, R.A., Boswell, R.N., Muluk, S.C. and Shearer, G.M. (1991) J. Immunol. 146, 2207.

[5] Rogers, L.A., Forster, S.M. and Pinching, A.J. (1989) Clin. Exp. Immunol. 75, 7.

[6] Amadori, A. and Chieco-Bianchi, L. (1991) Immunol. Today 12, 94.

[7] Clerici, M. and Shearer, G.M. (1993) Immunol. Today 14, 107. [8] Clerici, M., Lucey, D.R., Berzofsky, J.A., Pinto, L.A., Wynn,

T.A,, Blatt, S.P., Dolan, M.J., Hendrix, C.W., Wolf, S.F. and Shearer, G.M. (1993) Science 262, 1721.

[9] Macatonia, S., Gompels, M., Pinching, A.J., Patterson, S. and Knight, S. (1992) Immunology 75, 576.

[10] Knight, S.C. (1988) in: Lymphocytes: A Practical Approach - - Lymphocytes Proliferation Assays (G.G.B. Klaus, Ed.), Chap. 9, pp. 189-207, IRL Press Ltd, Oxford.

[11] Gupta, S. and Safai, B. (1983) J. Clin. Invest. 71, 296.

[12] Puppo, F., Pierri, I., Rogna, S., Pattarini, R., Piovano, P.L., Catellani, S., Varnier, O.E. and Indiveri, F. (1987) AIDS Res. Hum. Retrovir. 4, 423.

[13] Garbrecht, F.C., Sisldnd, G.W., Weksler, M.E. (1987) Clin. Exp. Immunol. 67, 245.

[14] Nussenzweig, M.C. and Steinman, R.M. (1980) J. Exp. Med. 151, 1196.

[15] Knight, S.C., Krejci, J., Malkovsky, M., Colizzi, V., Gautam, A. and Asherson, G.L. (1985) Cell. Immunol. 94, 427.

[16] Ali, A., Patterson, S., Cruickshank, K., Rudge, P., Dalgleish, A.G. and Knight, S.C. (1993) Clin. Exp. Immunol. 94, 32.

[17] Macatonia, S.E., Cruickshank, J.IC, Rudge, P. and Knight, S.C. (1992) AIDS Res. Hum. Retrovir. 8, 1699.

[18] Amadori, A., Zamarcki, R., Ciminale, V., Del Mistro, S., Siervo, A., Alberti, M., Colombatti, M. and Chieco-Bianchi, L. (1989) J. Immunol. 143, 2146.

[19] Inaba, K., Steinman, R.M., Van Voorhis, W.C. and Muramatsu, S. (1983) Proc. Natl..&cad. Sci. USA 80, 6041.

43