Expression of perforin on HIV-1-specific CD8⁺ lymphocytes after immunization with a gp120-depleted, whole-killed HIV-1 immunogen

Abstract

We examined HIV-1 antigen specific intracellular expression of perforin on CD4⁺ and CD8⁺ lymphocytes in subjects with chronic HIV-1 infection on antiviral drug therapy after immunization with a gp120-depleted, whole killed HIV-1 immunogen (inactivated, gp120-depleted HIV-1 in IFA, Remune). Based upon previous results, we hypothesized that the restoration of adequate T helper immune responses by vaccination against HIV-1 could result in the augmentation of CD8⁺ lymphocyte immune responses measured as perforin expression. In the current study we observed an increase in the frequency of perforin in CD8⁺ lymphocytes in HIV infected individuals immunized with a gp120-depleted HIV-1 immunogen while on antiviral drug therapy. Furthermore, the frequency of HIV-specific CD8⁺ perforin expressing cells correlated with the T helper immune response as measured by the lymphocyte proliferative response (LPR). The induction of such responses with immunization may have direct antiviral consequences and is being studied inm ongoing clinical trials.

Introduction

CD8⁺ T cells are important effectors of antiviral activity against human immunodeficiency type 1 virus (HIV-1). Their function can be divided into cytolytic and non-cytolytic activities. The cytolytic component involves a calcium dependent exocytosis of perforin and granzyme proteases. Phenotypically, cells that express high levels of perforin have been shown to be either CD8 dim cells (activated) or NK cells [1]. It is these activated CD8⁺ cells that are thought to be involved in lysis of HIV-1 infected cells. Recent studies have suggested that in HIV infected subjects on highly active antiretroviral therapy (HAART), CD8⁺ T cells stimulated with HIV-1 peptides express lower levels of perforin compared to CMV specific CD8⁺ T cells [2]. In addition, this lower expression of perforin was associated with impaired cytolytic activity. We examined intracellular expression of HIV-1 antigen-specific expression of perforin in CD4⁺ and CD8⁺ lymphocytes in subjects receiving antiretroviral drug therapy after immunization with a gp120-depleted, whole-killed HIV-1 immunogen (inactivated, gp120-depleted HIV-1 in IFA, Remune). Based upon previous studies, we hypothesized that the restoration of adequate T helper immune responses by vaccination against HIV-1 could result in the augmentation of CD8⁺ T cell responses such as perforin expression.

Methods

Patient samples

Eleven HIV-1 seropositive subjects were enrolled in an open-label research study as part of an expanded access programme of Remune (HIV-1 Immunogen). Four non-randomized, unimmunized controls were used as a comparison. The study was approved by the San Diego Naval Medical Center's Institutional Review Board and informed consent was obtained from all subjects. Immunized subjects received one intramuscular injection of the HIV-1 Immunogen at day 1 and at weeks 12, 24 and 36. Each immunization consisted of 10 units native p24 (approximately 100 µg of total protein) gp120-depleted, inactivated HIV-1 antigen (HZ321) in incomplete Freund's adjuvant (IFA). HIV-1 plasma RNA was measured by the Amplicor assay (Hoffman La Roche, Nutley, NJ, USA).

Antigens used in T cell cytokine response assays

Both gp120-depleted HIV-1 (HZ321; immunizing antigen) and native p24 antigens were utilized for in vitro immune function assays. Gp120-depleted HIV-1 (HZ321) antigen is highly purified by ultrafiltration and ion exchange chromatography from the extracellular supernatant fluid of HIV-1 HZ321 Hut-78 cells [2]. HIV-1 antigen is of clade A envelope and clade G gag [3]. The outer envelope protein (gp120) is depleted at the ultrafiltration stage of the purification process. Antigen preparations were inactivated through a sequential application of beta-propiolactone (BPL) [4] and ⁶⁰Co irradiation [5,6]. Native p24 was preferentially lysed from purified gp120-depleted, inactivated HIV-1 with 2% Triton X-100 and then purified using Pharmacia Sepharose Fast Flow S resin. Chromatography was carried out at pH 5·0 and p24 was eluted using a linear salt gradient. Purity of the final product was estimated by both SDS-PAGE and reverse phase HPLC to be > 99%. Staphylococcal enterotoxin B (SEB) was obtained from Sigma Biochemicals (St. Louis, MO, USA).

Stimulation of whole blood cultures with antigens

Measurement of intracellular cytokine expression in cells responding to antigen was performed as described previously and as illustrated in Fig. 1 [7–10]. Briefly, sodium heparinized venous whole blood was aliquoted into 15-ml conical polypropylene tubes (Becton Dickinson Labware, Franklin Lakes, NJ, USA) at 1 ml per tube. The costimulatory MoAbs CD28 and CD49d (Phamingen, San Diego, CA, USA) were added to each of the whole blood samples at a final concentration of 3 µg/ml. Media alone, SEB (3 µg/ml), gp120-depleted HIV-1 (10 µg/ml) and np24 (10 µg/ml) antigens were subsequently added. The polypropylene culture tubes were incubated slanted in a humidified 37°C, 5% CO₂ incubator for a total of 5 h. The secretion inhibitor Brefeldin A (10 µg/ml, Pharmingen, San Diego, CA, USA) was included for the final 3 h of activation [8,9] to enable optimal antigen processing by antigen presenting cells (APC) [6]. At 5 h, 100 µl of 20 mm EDTA (for a final concentration of 2 mm EDTA) was added directly to the whole blood cultures. Sample tubes were vortexed and incubated for 15 min at 25°C (RT). Blood samples were then lysed and fixed in 12 ml of 1× FACS^TM Lysing Solution (Becton Dickinson Immunocytometry Systems, San Jose, CA, USA) for 10 min at RT. Cells were washed in PBS + 1% FBS + 0·09% NaN₃ and frozen at −70°C in a freezing medium containing 10% DMSO and 1% BSA in PBS.

Fig. 1.

A representative FACS plot of HIV antigen stimulated perforin expression (subject 7) on activated CD8⁺ lymphocytes before and after immunization. Arrows denote time of immunization.

Immunofluorescent staining

Frozen cells were thawed rapidly in a 37°C water bath, aliquoted into staining tubes and washed once with cold wash buffer (PBS, 1% FBS, 0·09% NaN₃). Cells were subsequently resuspended in 0·5 ml of FACS ™ permeabilizing solution (BDIS) for 10 min at RT in the dark. After permeabilization, cells were washed once and staining was performed for 30 min at RT in the dark using a titrated mixture of fluorescent conjugated MoAbs. For a typical three-colour analysis of perforin response, the staining antibody cocktail consisted of CD4 or CD8 PerCP, CD69 FITC and perforin PE. CD69FITC (mouse IgG1 kappa, cat. no. 31954X) and perforin PE (mouse IgG2b, cat. no. 6599kk) were obtained from Pharmingen (San Diego, CA, USA). Mouse IgG isotype matched control antibodies were included to detect non-specific staining. For a three-colour analysis of perforin and CD27 expression on CD8 cells, the following antibodies were used: CD27 FITC (cat. no. 30824X), perforin PE (cat. no. 6599KK) (Pharmingen, San Diego, CA, USA) and CD8 PerCP (cat. no. 347314) (Becton Dickinson, San Jose, CA, USA). After staining, samples were washed and fixed in 1% paraformaldehyde in PBS for at least 3 h and stored at 4°C in the dark until FACS analysis. All antibodies and isotype controls were obtained from Becton Dickinson Immunocytometry Systems (San Jose, CA, USA) and Pharmingen (San Diego, CA, USA).

Flow cytometric analysis

Three-colour flow cytometric analysis was performed on a FACSORT ™ flow cytometer (BDIS). Data were acquired using CELLQuest ™ software (BDIS), typically collecting 40 000–50 000 gated lymphocyte events. Data were displayed as two-colour dot plots (FL1 versus FL2) in Realist Software (Phoenix Flow Systems, San Diego, CA, USA) to measure the proportion of the double-positive (perforin⁺/CD69⁺) on CD4 or CD8 cells. Since all specific perforin expression occurs within the CD69⁺ (activated) cell subset, CD69 staining was included to enhance the identification of Ag-responsive T cells, as reported by Waldrop et al. [11]. Forward-scatter vs. side-scatter gating was employed in data analysis to exclude any CD4⁺ monocytes and doublets. For CD27/perforin analysis data were calculated and reported as percentage of CD27/perforin on CD8 (bright cells) in lymphocyte region and analysed using CellQuest™ software.

Lymphoproliferation assays

For the lymphocyte proliferation assays, fresh PBMCs from HIV-1 seropositive subjects were purified and cultured with medium alone or with inactivated HIV-1 antigens including whole gp120-depleted HIV-1 (5 µg/ml) and np24 (5 µg/ml). PBMCs were seeded in a round bottom 96 well plate (Falcon) at 2 × 10⁵ cells/well in complete RPMI medium (Hyclone) containing 10% heat-inactivated (30′, 56°C) human AB serum (Gemini), 100 U/ml penicillin, 100 µg/ml streptomycin (Gibco), and l-glutamine 1% (Hyclone). All assays were performed in triplicate. After 6 days of incubation, cells were labelled with 1 µCi/well of ³H-thymidine in complete RPMI for 16–18 h. On day 7, 20 µl of betapropiolactone (BPL) (1:1600 final concentration) was added to each well to neutralize any virus produced during the incubation period. Cells were harvested after a 2-h incubation in BPL at 37°C and incorporated label was determined by scintillation counting in a beta-counter. Geometric mean counts per minute (c.p.m.) were calculated from the triplicate wells with and without antigen. Results were calculated as a lymphocyte stimulation index (LSI), which is the geometric mean cpm of the cells incubated with antigen divided by the geometric mean cpm of the cells without antigen (cells incubated in media alone).

Statistical analysis

The Spearman rank sum test was utilized to examine relationships for immunological parameters. The Mann–Whitney U non-parametric test was used to compare post-immunization and pre-immunization perforin levels. All P-values presented are two-tailed.

Results

Table 1 shows the baseline demographics of this cohort including HIV-1 plasma RNA, CD4⁺ counts, and concomitant medications. The median CD4⁺ count at baseline for this cohort was 810 cells/mm³. The median plasma HIV-1 RNA copy number at baseline for this cohort was < 400 RNA copies/ml.

Table 1.

Baseline demographics, CD4, viral load, and antiviral drugs

Subject no.	CD4 cc/mm3 Absolute no. Day 1	RNA/copies/ml Day 1	Months prior to baseline on HIV medication
1	607	< 400	Lamivudine	32
			Stavudine	31
			Nelfinavir	16
2	1969	< 400	Lamivudine	21
			Stavudine	21
			Nelfinavir	21
3	1270	< 400	Zidovudine	30
			Lamivudine	30
			Indinavir	30
4	657	< 400	Zidovudine	88
			Zalcitabine	49
			Indinavir	60
5	954	< 400	Efavirenz	3
			Stavudine	15
			Lamivudine	15
6	597	< 400	Zidovudine	29
			Lamivudine	29
			Efavirenz	2
7	834	< 400	Zidovudine	37
			Lamivudine	37
			Indinavir	32
8	568	< 400	Zidovudine	17
			Lamivudine	17
			Nelfinavir	17
9	630	< 400	Stavudine	10
			Lamivudine	10
			Nevirapine	10
10	520	< 400	Stavudine	38
			Lamivudine	43
			Efavirenz	7
11	305	< 400	Zidovudine	30
			Lamivudine	30
			Efavirenz	5
Mean	810	< 400

Baseline demographics, CD4, viral load, and antiviral drugs.

Figures 1 and 2 reveal representative FACS plots of perforin expression on activated CD8⁺ cells after stimulation with inactivated gp120-depleted HIV-1 antigen. The subject (subject 7) displayed in Fig. 1, had a frequency of 1·9% of HIV-1 antigen activated CD8⁺ lymphocytes that expressed perforin at week 1, which increased to 5·7% after three immunizations (week 36). The subject displayed in Fig. 2a (subject 2) had a frequency of 2·0% of HIV-1 antigen-activated CD8⁺ lymphocytes that expressed perforin at week 1, which increased to 5·0% after three immunizations (week 36). Figure 2b reveals the non-HIV-1 antigen controls which included media alone, SEB and candida antigen stimulation for subject 2 at week 36. In addition, most of the perforin-positive staining HIV-specific CD8⁺ T cells observed after immunization were predominantly CD27⁺(Fig. 3).

Fig. 2.

(a) A representative FACS plot of perforin expression (subject 2) on activated CD8⁺ lymphocytes after immunization. Arrows denote time of immunization. (b) Non-HIV antigens for the same subject at week 36. (i) media control; (ii) SEB; (iii) HIV antigens.

Fig. 3.

Representative FACS plot (subject 9) of CD27 expression and perforin coexpression on CD8⁺ lymphocytes. (a) Control (media alone); (b) SEB, (c) HIV antigen stimulation.

As shown in Fig. 4, the median frequency of stimulated activated CD8⁺ lymphoctyes expressing perforin for the 11 subjects increased approximately two-fold from baseline at week 24 when stimulated with HIV (P = 0·001, Fig. 4a and p24 antigens (P = 0·009 Fig. 4b) after two immunizations (week 24). The frequency of CD8⁺ lymphocytes expressing perforin was maintained after the third immunization (at week 36) when stimulated with HIV (P < 0·0001) or p24 (P = 0·005) antigen. In contrast, the frequency of CD8⁺ lymphocytes expressing perforin did not significantly change in response to SEB antigen with immunization (P > 0·05). Furthermore, the frequency of HIV-1 or p24 antigen stimulated CD8⁺ lymphocytes which did not express perforin (CD8⁺/perforin) did not significantly change with immunization (P > 0·05).

Fig. 4.

Perforin expression on CD8⁺ lymphocytes in response to (a) HIV-1 antigen or (b) p24 antigen. Arrows denote time of immunization. Horizontal lines represent the medians ±s.e.

We next examined the expression of antigen-stimulated perforin in CD4⁺ lymphocytes. In contrast to the CD8⁺ lymphocytes there was no significant change after two immunizations in the median frequency of CD4⁺ lymphocytes expressing perforin at week 24 with HIV (P > 0·05) or p24 (P > 0·05) antigen stimulation (Fig. 5a,b). Similarly, there was no change in perforin expression in CD4⁺ lymphocytes after three immunizations (week 36) when stimulated with HIV (P > 0·05) or p24 (P > 0·05) antigen stimulation.

Fig. 5.

Perforin expression on CD4⁺ lymphocytes in response to (a) HIV-1 antigen or (b) p24 antigen. Arrows denote time of immunization. Horizontal lines represent the medians ±s.e.

For the same subjects, we also examined the lymphoproliferative responses to HIV-1 as shown in Fig. 6. Subjects increased their in vitro lymphoproliferative responses after immunization (P < 0·05). We also examined whether a relationship existed between the frequency of perforin in CD8⁺ lymphocytes with lymphocyte proliferation, a marker of CD4⁺ T helper cell function. An association was observed between HIV and p24 antigen stimulated LPR and HIV (r = 0·57, P = 0·008) and p24 antigen (r = 0·44 P = 0·05) stimulated frequency of perforin expressing CD8 + lymphocytes as shown in Fig. 7.

Fig. 6.

Median ± s.e. lymphocyte proliferative responses to HIV-1 antigen.

Fig. 7.

Correlation between perforin expression on CD8⁺ lymphocytes and lymphocyte proliferation to HIV-1 antigen. The fitted linear regression is indicated by the solid lines, 95% confidence interval is indicated by the dotted lines. r = 0.57; P = 0.008.

Four unimmunized patients were also examined for expression of perforin. A low frequency of CD8⁺ cells stimulated with HIV (median ±s.e. = 1·5 ± 0·25) and p24 antigens (median ±s.e. = 1·1 ± 0·24) expressing perforin was observed. Similarly, in these unimmunized controls, a low level of expression of perforin was noted on CD4⁺ lymphocytes when stimulated with HIV (median ±s.e. = 0·35 ± 0·13) or p24 (median ±s.e. = 0·25 ± 0·38).

Discussion

This report examined CD8⁺ lymphocyte responses in subjects on antiviral therapy after treatment with an HIV-1 immunogen. We utilized an intracellular flow cytometry based assay to examine the frequency of HIV-1 antigen stimulated cells expressing perforin, a cytotoxic factor associated with CD8⁺ T cells, after immunization with a gp120-depleted, whole-killed HIV-1 immunogen. We hypothesized that HIV-1-specific induction of CD4⁺ T helper cells via immunization would result in the activation of CD8⁺ lymphocytes, expressed as an increased frequency of intracellular perforin.

The expression of perforin in CD8⁺ lymphocytes was enhanced by immunization over the course of the observation period. These subjects displayed a low frequency of CD4⁺ or CD8⁺ lymphocytes expressing perforin prior to immunization. Furthermore, unimmunized control patients displayed a low frequency of CD8⁺ or CD4⁺ lymphocytes expressing perforin. The frequency of CD8⁺ but not CD4⁺ lymphocytes expressing perforin significantly increased after immunization. Interestingly, previous reports suggested a diminution of other markers of CD8⁺ lymphocyte activity in patients on potent antiviral drug therapy [12–16]. In contrast to CD8⁺ lymphocyte perforin expression, the frequency of perforin expression on CD4⁺ T cells did not increase post-immunization. Similarly, the frequency of expression of CD8⁺/perforin did significantly change with immunization. Thus, the increase in the frequency of perforin expression on CD8⁺ T cells is probably not a reflection of overall increase in CD8⁺ T cell number. We also observed, as have others [2], that HIV-specific perforin expressing CD8⁺ lymphocytes are predominantly CD27⁺, and thus early in their maturation.

Previously we had observed that immunization increased the frequency of IFN-γ on CD4 cells by a intracellular flow cytometry based assay [17]. Thus, immunization may be activating the ‘helper’ function of these lymphocytes via cytokine expression and proliferation, but not their cytotoxic function as measured by the perforin pathway. We cannot rule out, from this study, that some of the lymphocytes activated with immunization may have been natural killer cells, which can be phenotypically CD8⁺ and express perforin as a means to kill virus-infected cells.

We then examined the relationship between HIV-specific CD4 T cell helper responses and CD8⁺ lymphocyte expression of perforin. As observed in previous studies, an enhancement of lymphocyte proliferation to HIV-1 antigens after immunization was seen in this study. We also noted a correlation between T helper activity as measured by lymphocyte proliferation and perforin expression on CD8⁺ lymphocytes. This is consistent with observations of individuals with non-progressive HIV-1 infection [14], as well as in other chronic viral diseases [18,19], where an association has also been noted between CD4⁺ T helper and CD8⁺ T cell immune responses. Although this correlation does not imply cause or effect, one may speculate that the ability to enhance with immunization or maintain CD4 T helper immune function, as observed in non-progressors, may allow the generation of antiviral CD8 T cell responses, as observed in this study. Recently, enhanced cytotoxicity in a chromium release assay was observed in subjects receiving this therapeutic vaccination while on potent antiviral drugs, but not in subjects on antiviral drugs alone [20]. This and other ongoing, placebo-controlled studies will help determine whether the induction of such responses correlates with better control of plasma viremia.

In summary, in this study we have demonstrated an augmentation of the frequency of perforin on CD8⁺ lymphocytes in HIV infected individuals immunized with a gp120-depleted HIV-1 immunogen while on antiviral drug therapy. The induction of such responses with immunization may correlate with direct antiviral effects and is being studied in ongoing clinical trials.