Cytomegalovirus (CMV) infection in AIDS patients is associated with a CD3 receptor-mediated T cell hyporesponsiveness

Abstract

HIV⁺ individuals with human CMV (HCMV) reactivation have a CD3 receptor-mediated T cell hyporesponsiveness when compared with CD4-matched HIV⁺ and HCMV⁻control groups. The impairment of proliferation was not reversed by exogenous IL-2. A typical increase in NFκB expression was observed following cross-linking of the CD3 receptor, but did not lead to increased CD25 cell surface expression or cell proliferation. The HCMV-induced non-responsiveness was not observed when cells were stimulated with phorbol esters. Lymphocytes cultured with media collected from cell cultures infected with HCMV showed a dose-dependent inhibition in the total T cell population even though cells staining dually for CD8/57 increased in number. The altered growth factor requirements of CD8/57⁺ cells may therefore account for their presence in AIDS and patients following bone marrow transplantation.

INTRODUCTION

Human CMV (HCMV) is an ubiquitous herpesvirus that causes severe morbidity and mortality in immunocompromised patients and congenitally infected newborns [1–3]. Acute infection results in depressed cellular function, as shown by decreased lymphoproliferative responses to mitogens, viral antigens [4] and diminished reactivity to skin tests [5]. Conversely, HCMV infection is associated with a series of cellular responses that resemble those observed in cells activated by growth factors or hormones, including hydrolysis of phosphatidylinositol biphosphate, Ca²⁺ influx with increased cytosol-free Ca²⁺, and increase in cellular levels of cyclic AMP [6]. The intracellular pathways which govern these interactions are poorly defined.

MoAbs against the CD3 structure, an essential transducer in T lymphocyte activation [7], are mitogenic for T cells [8] and, when bound to Fc receptors on accessory cells or a solid surface, support the growth of resting and antigen-primed memory T cells [9–11]. The activation through the T cell receptor (TCR) elicits a cascade of biochemical processes including stimulation of tyrosine kinases and inositol triphospate-mediated release of Ca²⁺ from intracellular stores. The result is increased expression of transcription factors, nuclear factor-activated T (NF-AT) and nuclear factor κ B (NFkB), leading to up-regulation of IL-2 production and CD25 cell surface expression [9].

In a study of 29 children with acute HCMV infection Timon and coworkers showed a selective impairment of T cell activation via the TCR/CD3 complex [12]. Of note, this impairment was corrected by addition of exogenous IL-2, although the precise mechanism of the underlying HCMV-induced anergy was not investigated. A recent case report of a young child with a T cell activation deficiency, who presented with a family history of serious HCMV infections, showed diminished IL-2 production and abnormal NF-AT cell binding activity. Again, T cell proliferation could be restored by addition of exogenous IL-2 [13].

Patients with HCMV mononucleosis have elevated levels of CD8⁺ cells and increased expression of class II major histocompatibility antigens, as well as reversed ratios of CD4 and CD8 cells compared with normal individuals [14]. Wang and coworkers have shown that these CD8 cells also co-express the CD57 antigen and they perform important immunoregulatory functions [15,16]. Interestingly, CD8/57 cells from healthy individuals mediate contact-dependent suppression which is distinct from the non-specific, soluble factor-mediated suppression exhibited by a phenotypically similar subset in HIV⁺ and bone marrow transplantation (BMT) patients. This suggests that HCMV potentiates functional differences within this CD8 cell population, which may result in HIV infection and immune reconstitution following BMT being associated with higher numbers of this subset [17]. The mechanism of this selective cell expansion and production of inhibitory soluble factors is unknown.

In this study we investigated HIV⁺ patients with acute HCMV infection in order to determine whether the selective impairment is consistent with that described in HCMV-infected children reported by Timon and coworkers [12], and to extend this work by studying the effect of HCMV infection on NFκB production during T cell activation. Peripheral blood mononuclear cells (PBMC) from HIV-infected patients with and without acute HCMV infection were stimulated with an IL-2-independent specific MoAb to CD3 immobilized on plastic.

PATIENTS AND METHODS

Patients

Seventy-two HIV⁺ subjects and 12 healthy HIV⁻ individuals were recruited to this study. The HIV⁺ patients were divided into three groups according to peripheral CD4 count and HCMV status. Patients in group 1 (n = 30) had acute active HCMV infection with either retinitis (n = 12) or gastrointestinal involvement (n = 18). HCMV retinitis was diagnosed by an experienced ophthalmologist. Gastrointestinal involvement was diagnosed by symptoms, endoscopic features and histology of four biopsies per patient with confirmation by staining for the immediate early gene product and positive DEAFF test. The remaining patients with no symptoms (evidence of inflammation to eyes or gastrointestinal tract) and seropositive for HCMV with no evidence of reactivation as determined by DEAFF test of urine samples and reverse transcriptase-polymerase chain reaction (RT-PCR) of peripheral blood cells for immediate gene expression were recruited into group 2 (n = 30) or group 3 (n = 12) according to peripheral CD4; < 60 cells/μl and > 200 cells/μl, respectively. Group 4 and 4a (n = 12) consisted of HCMV⁺ HIV⁻individuals with CD4 counts > 500 cells/μl.

Preparation of PBMC

Peripheral blood was collected into endotoxin-free lithium heparin tubes and the mononuclear cells separated on Histopaque (Sigma Chemical Co., Poole, UK) according to manufacturer's instructions. The cells recovered from the mononuclear layer were washed twice in Hanks' buffered salt solution (HBSS; Flow ICN, Thame, UK) and resuspended at a concentration of 1 × 10⁶ cells/ml in sterile filtered tissue culture media consisting of: RPMI (Sigma) supplemented with glutamine 2 mm, penicillin 100 U/ml, streptomicin 100 μg/ml, HEPES 25 mm (pH 7.6) buffer and 10% heat-inactivated fetal calf serum (FCS; Sigma).

Stimulation of PBMC

A cross-linking anti-CD3 antibody independent of exogenous IL-2 was obtained from Immunotech (Immunotech, USA). The optimal concentration for cell activation was established by stimulation of PBMC from normal healthy individuals (data not shown). A stock antibody solution (0.5 μg/ml) was prepared in sterile RPMI 1640. Anti-CD3 (50 μl; 25 ng) was coated onto each well of a 96-well microtitre plate by incubation at 37°C for 2 h immediately before use. Excess antibody was removed by washing in sterile PBS pH 7.5. Cells were seeded at 1 × 10⁵/well in triplicate and incubated at 37°C in 5% CO₂. Cell proliferation was measured at 48 h and 72 h post-stimulation. As a receptor-independent activation signal control, PBMC were also stimulated with a combination of phorbol-12-myristate-13-acetate (PMA; 10 μg/ml; Sigma) and calcium ionophore A127487 (61.5 ng/ml; Sigma).

Cell proliferation assay

Cell proliferation was monitored by tritiated ³H-thymidine incorporation (0.5μCi/well; Amersham, High Wycombe, UK). Cultures were pulsed 6 h before harvesting. Labelled cells were harvested onto glassfibre filter mats (Wallac LKB, Milton Keynes, UK) and counted by liquid scintillation using a 1205 Betaplate counter (Wallac LKB). Results were recorded as ct/min.

Cell proliferation of PBMC from normal healthy individuals following culture with HCMV conditioned media

PBMC isolated from HCMV-infected individuals (as determined by a positive DEAFF test) were cultured at 1 × 10⁶ cells/ml at 37°C in 5% CO₂. After 72 h, supernatants were pooled and 10-fold dilutions made to 1 × 10⁻⁴ in RPMI 1640 and filter sterilized. Volumes (50 μl) of each dilution were added to 200 μl of PBMC from normal healthy individuals seeded at 1 × 10⁶/ml, mixed and plated onto anti-CD3-coated plates. Cell proliferation was measured at 48 h and 72 h following the addition of HCMV conditioned media.

RT-PCR for NFκB gene expression

RNA was isolated from PBMC following lysis with 800 μl of RNAzol B according to the manufacturer's instructions (Biogenesis, Berks). RNA precipitates were washed in 75% ethanol, air-dried and rehydrated in 30 μl of diethylpyrocarbonate-treated water. RNA integrity was confirmed by agarose gel electrophoresis. cDNA was produced by incubating 5 μl of total RNA at 37°C for 60 min in a 30-μl reaction mix consisting of Tris–HCl 50 mm pH 8.3, KCl 40 mm, MgCl₂ 6 mm, DTT 1 mm, dNTPs (10 nm equimolar mix), oligo dT_12–18 and MMLV reverse transcriptase (200 U; Gibco BRL, Paisley, UK). Following incubation the reaction mix was heated to 70°C for 5 min. cDNA mix (3 μl) was used in each 50 μl PCR reaction mix consisting of MgCl₂ 1.5 mm, dNTPs 10 nm equimolar mix, AmpliTaq (5 U per reaction; Perkin Elmer, Warrington, UK), 10 × reaction buffer, 5 μl of each oligonucleotide primer (0.3 μm).

PCR primer sequences

β actin upstream primer: 5′ TTTAAGGGCCCCTAGC 3′, downstream 5′ ATCAGTACCGTTTGCATGCAT 3′; NFκB upstream primer: 5′ ATGGATGATGATGATATCGCCGCG 3′, downstream 5′ CGGGGAGGTAGCAGGTGGCGTTTACGAAGATC 3′. PCR cycle conditions were 94°C for 1 min, 55°C for 2 min, 72°C for 3 min for 35 cycles and a final extension of 72°C for 7 min. PCR products were separated on a 1.2% (w/v) agarose gel and viewed under ultra-violet light following ethidium bromide staining (10 μg/ml).

ELISA for IL-2

A standard sandwich ELISA technique was used. Briefly, 15 ng/ml of a primary antibody (BAF202; R&D Systems, Oxford, UK) diluted 1:100 in coating buffer (Whitley Scientific, Preston, UK) were added to each well of a Nunc Immuno II plate and incubated overnight at 4°C. The following morning the plates were washed and blocked for 3 h at 37°C with 1% (v/v) Tween 20–PBS pH 7.4, containing 1% (v/v) human serum albumin (HSA; Sigma). Plates were then washed, and 50 μl of the test sample or diluted standard were added to the coated well and incubated overnight at 4°C. Plates were washed and 100 μl of the second mouse MoAb (MAB 602; R&D Systems) were added and the plates incubated for a further 4 h at 4°C. Plates were washed and 100 μl of an alkaline phosphatase-conjugated rabbit anti-mouse antibody (Sigma) added and the plates incubated for 3 h at room temperature. At this point, 100 μl of the substrate buffer (Whitley Scientific) containing 1.5 mg/ml p-nitrophenyl phosphate disodium were added. The absorbances were read at 405 nm when the top standard read ≈ 1.6 optical density (OD) units. The assay was standardized against International Standard 86/504 obtained from the National Institute of Biological Standards and Controls (Potter's Bar, UK). The lower limit of detection was 10 pg/ml and the within batch coefficient of variation was 12.6%.

Immunophenotyping by flow cytometry

The following conjugated MoAbs were used for phenotypic analysis: CD3-FITC, CD8-PE-Cy5, CD25-FITC and CD57-FITC (Immunotech, Luton, UK). Cells were incubated with the appropriate concentration of MoAb in PBS supplemented with 20% (v/v) pooled normal human serum (Sigma) and 0.1% (w/v) sodium azide at 4°C for 30 min. Cells were washed in PBS supplemented with 1% (w/v) bovine serum albumin (BSA; Sigma) and analysed on an Ortho Cytron flow cytometer (Ortho Diagnostics, Amersham, UK).

Statistical analysis

Mann–Whitney U-test was used for comparisons between patient and normal control groups. P < 0.05 was considered significant.

RESULTS

Anti-CD3 activation of PBMC

Comparison of proliferative responses revealed that PBMC of patients in group 2 had significantly lower thymidine incorporation than those of patients in group 3 (P < 0.05). The degree of cell proliferation was dependent on absolute CD4 numbers (Table 1), patients in group 2 having CD4 counts < 60 cells/μl (mean 53 cells/μl, range 5–55 cells/μl). The poor proliferation in group 2 patients was restored by culturing cells with FCS instead of autologous plasma (Table 1). Proliferative responses of cells from patients in group 1 (positive HCMV infection) were < 100 ct/min and, unlike those of group 2 patients, could not be restored by culturing with FCS (Fig. 1). Proliferative responses presented are taken at 72 h post-stimulation and are consistent with the 48 h time point.

Table 1.

Proliferative responses to anti-CD3 and phorbol esters

Fig. 1.

A box and whiskers plot of the effect of human CMV (HCMV) infection on anti-CD3 stimulation on peripheral blood mononuclear cells. Details of the groups are given in Patients and Methods. Cells from normal HIV⁻ donors stimulated with anti-CD3 and cultured with 25% HCMV-conditioned media are shown in group 4a. The box represents the interquartile range and the whiskers are lines that extend from the box to the highest and lowest values excluding outliers. Outliers (○) are samples with values between 1.5 and 2.5 box lengths from the upper or lower edge of the box.

To exclude the possibility that the failure of cells to respond was simply due to the cytopathic effects of HCMV, cell viability was shown to be > 90% in all four patient groups by dye exclusion. However, the failure of HCMV-infected cells from group 1 patients to respond was overcome when CD3 receptor signalling was bypassed and cells activated by a combination of phorbol esters and calcium ionophore (Table 1).

The effects of exogenous IL-2 and HCMV-conditioned media on PBMC

The failure of T cell responses in group 1 patients was not recovered by the addition of exogenous IL-2 (50 IU/ml, Table 1). Flow cytometric analysis of cells from patients in group 1 (following stimulation with anti-CD3 and exogenous IL-2) revealed normal surface expression of CD3 with no up-regulation of CD25 (data not shown). Furthermore, soluble IL-2 was undetectable in the tissue culture media (limit of detection 10 pg/ml) at 4, 8, 16, 24 and 48 h post-stimulation, in contrast to samples from groups 2 and 3, where concentrations showed increased production until 48 h post-stimulation of 480 and 691 pg/ml, respectively.

As described above, the prolonged period of immunosuppression following HIV infection is often associated with the emergence of a dominant subpopulation of CD8⁺ cells (CD8/57) which are believed to exhibit suppressive activity through the release of soluble factors. Using filtered conditioned media from HCMV-infected individuals, we examined this suppressive activity on freshly isolated peripheral blood cells from normal healthy donors. The proliferative response of these cells following stimulation with anti-CD3 was unaffected by culturing in the presence of HCMV-conditioned media (Fig. 1). Immunophenotypic studies revealed a dose-dependent reduction in the numbers of CD8 cells when cultured in HCMV-conditioned media for 48 h (Fig. 2a). In contrast, dual staining for CD8/57⁺ cells showed a minor population which were expanded in HCMV-conditioned media. The absolute numbers of these cells were too small to affect greatly the overall CD8 count (Fig. 2b).

Fig. 2.

(a) Flow cytometric analysis of cells cultured with human CMV (HCMV)-conditioned media showed a dose-dependent inhibition in CD3 and CD8 lymphocytes compared with media alone (at 25% P < 0.05). (b) Dual staining of CD8 cells for CD57 expression revealed a significant increase in the number of CD8/57⁺ lymphocytes. At 25% HCMV-conditioned media a six-fold increase was detected over the media alone control (P < 0.05).

NFκB gene expression

The effect of HCMV on NFκB gene expression following anti-CD3 stimulation was measured in cells obtained from five patients in groups 1, 2 and 3 by RT-PCR. An increased level of NFκB expression was seen in all groups following stimulation with anti-CD3. Representative staining patterns of cells obtained from each group are shown in Fig. 3.

Fig. 3.

Agarose gel electrophoresis of polymerase chain reaction (PCR) products from groups 1, 2 and 3 are shown. Amplification of β-actin was used to confirm integrity of isolated RNA and efficiency of the PCR. The expected product sizes for β-actin and NFκB were 1000 bp and 500 bp, respectively. The molecular weight marker (M) was a Hind III digest of λ phage DNA.

DISCUSSION

The recent introduction of highly effective anti-retroviral treatments, consisting of reverse transcriptase inhibitors in association with proteinase inhibitors, suggests that despite considerable reductions in viral load and improvements in CD4 cell counts, patients are at continuing risk of recurrent HCMV disease, particularly the eye and gastrointestinal tract [18].

In this study we show that HIV⁺ patients co-infected with HCMV had reduced anti-CD3 proliferative responses. Unlike previous reports, the addition of exogenous IL-2 failed to restore proliferation, although normal responses were observed when cells were stimulated with phorbol esters. Using PBMC from HIV individuals, Borthwick and coworkers have shown a strong correlation between lack of CD28 expression and poor proliferation to mitogens. They concluded that the failure to respond was due to the absence of the CD28 signal transduction pathway [19]. Our results for HCMV co-infected cells, however, would suggest that the poor proliferation was due to a failure in IL-2 production and/or IL-2 receptor expression. Furthermore, the results show that the impaired proliferative response of HIV-infected patients could be restored by culturing PBMC in FCS. However, cells from HIV⁺ patients with HCMV infection could not be stimulated with anti-CD3 even when cultured with FCS and additional exogenous IL-2. The suppression could not be transferred to PBMC from normal healthy individuals using HCMV-conditioned media. Interestingly, there was no detectable difference in the level of NFκB expression between the HCMV-infected and non-infected group, suggesting that the interaction is independent of NFκB.

The current model for T cell activation requires two signals, and anergy induction in resting T cells occurs if primary proliferation is induced by high density triggering of the TCR in the absence of accessory signals [20,21]. The hyporesponsiveness observed in cells from group 1 patients could not, however, be recovered even when additional secondary signals such as exogenous IL-2 were present. No DNA fragmentation was detected by agarose gel electrophoresis (data not shown), indicating that the cells were not undergoing apoptosis. Our results show increased levels of NFκB following anti-CD3 cell stimulation that did not result in the production of IL-2 and CD25 expression. It is possible that the increase in NFκB demonstrated in the HCMV-infected group could result via an alternative activation pathway, which leads to a qualitative, not a quantitative, difference in NFκB being responsible for the lack in signal transduction and IL-2 production which will require further investigation.

Interestingly, unlike CD57⁻ cells, CD57⁺ cells do not contain message for the IL-2R α-chain, which may explain why Lewis and coworkers were unsuccessful when attempting to reduce the amount of DNA fragmentation in cultured CD28⁻CD57⁺ CD8⁺ cells by the addition of IL-2, which normally prevents apoptosis in T cells [22]. We showed that following culture of PBMC from normal individuals in HCMV-conditioned media a higher proportion of cells expressed CD57, a phenotype believed to indicate a maturational step and associated with functional changes. Data from D'Angeac and colleagues suggest that CD57 expression is a differentiation event which occurs on CD57⁻ T cells late in the immune response [23]. Our results show that normal cells cultured with HCMV-conditioned media had increased numbers of CD8 cells expressing CD57, that we believe represents a terminally differentiated non-proliferating population of T effector cells. A recent report by Wang and coworkers has shown that CD57 cells can be maintained by HCMV and not Epstein–Barr virus (EBV) or influenza proteins [24]. Further studies are required to demonstrate whether this increase in CD57 expression was driven by growth factor or viral antigen.

In conclusion, we have shown that peripheral blood cells isolated from patients co-infected with HIV and HCMV have an anti-CD3 receptor-mediated hyporesponsiveness. The synergistic effect of HIV and HCMV leads to a block in IL-2 production and not to an increased expression of NFκB. Furthermore, the failure of exogenous IL-2 to recover the anti-CD3 response suggests that the mechanism is different from that described by Timon and coworkers [12]. It is possible that the IL-2 requirement of a terminally differentiated cell population such as CD8/57 may be altered from a high to a low dependency, a process which may be influenced by the presence of HCMV-encoded proteins. Such an hypothesis may account for the prevalence of CD8/57 cell populations in cases of long-term chronic immunosuppression such as AIDS and post-BMT, and warrants further investigation.