Reversal of Human Immunodeficiency Virus Type 1 Protein-Induced Inhibition of Natural Killer Cell Activity by Alpha Interferon and Interleukin-2

Abstract

A recombinant fusion peptide, Env-Gag, derived from the human immunodeficiency virus type 1 (HIV-1) genome corresponding to a defined portion of the envelope (Env) and internal core (Gag) proteins was examined for immunoregulatory effects on the cytotoxic activity of natural killer (NK) cell-enriched, large granular lymphocytes (LGL) from healthy donors. Percoll-separated, NK cell-enriched LGL precultured for 24 h with Env-Gag at 10- and 50-ng/ml concentrations, which significantly stimulated lymphocyte proliferation, caused significant suppression of NK cell activity. Denatured Env-Gag did not cause any effect on the NK cell activity of LGL. Two other control peptides, one derived from the Escherichia coli vector used to clone the HIV Env-Gag fusion peptide and the other derived from a non-HIV-1 viral antigen (rubeola virus), did not produce any observable effect on the NK cell activity of LGL, demonstrating the specificity of the effect produced by Env-Gag. Subsequent treatment of LGL with alpha interferon (IFN-α) or interleukin 2 (IL-2) alone partially reversed the Env-Gag-induced suppression of NK cell activity. However, LGL treated with both IFN-α and IL-2 completely reversed the suppression of NK cell cytotoxicity by Env-Gag. The combined effect of IFN-α and IL-2 in enhancing NK cell activity may provide a novel therapeutic approach to the restoration of depressed NK cell activity observed in HIV-infected patients.

Natural killer (NK) cells are considered to be a first line of defense against virus infections and tumors and may be responsible for controlling occult metastases (for a review, see reference 42). NK cells also play a significant regulatory role in various immune reactions (9, 22). Defective NK cell activity is an early manifestation of human immunodeficiency virus type 1 (HIV-1) infection (1, 4, 40), and severe dysfunction of NK cells occurs in the later stages of the disease (10) despite their normal numbers in peripheral blood as identified by several monoclonal antibodies (36, 40). This finding suggests that either the virus or soluble factors derived from HIV-1 may be responsible for this inhibition of NK cell activity in HIV-1-infected patients. We previously demonstrated that the recombinant HIV-1 peptide Env-Gag, a fusion product of the env, gp41, and gag p24 genes, can induce de novo polyclonal immunoglobulin synthesis and can suppress pokeweed mitogen-stimulated immunoglobulin production by normal lymphocytes in vitro (26). We also reported differential effects of HIV-1 gp120 on interferon (IFN) production by mononuclear cells (25). Further, we have shown that certain HIV-1 peptides can inhibit the NK cell activity of normal lymphocytes and that NK cells from HIV-1-infected subjects are selectively sensitive to the inhibitory effects of Env-Gag (24). Our previous studies also showed that lymphocytes from intravenous drug abusers demonstrate lower NK cell and antibody-dependent cellular cytotoxic activities and that the suppressed NK cell activity can be partially reversed by in vitro treatment with interleukin 2 (IL-2). The present study was undertaken to examine the combined effects of IFN-α and IL-2 on Env-Gag-induced NK cell inhibition of large granular lymphocytes (LGL).

MATERIALS AND METHODS

Recombinant HIV-1 Env-Gag peptide.

Expression and purification of the Env-Gag recombinant HIV fusion protein expressed in Escherichia coli and used in this investigation have been described previously (6, 8, 18). Env-Gag has conserved and antigenic epitopes from the env and gag regions of the HIV-1 genome. The sequence of Env-Gag used in the present experiments is equivalent to amino acids 560 to 639 in Env and 87 to 276 in Gag, totaling 270 amino acids. Twenty-one extra amino acids represent coding from the vector region. The Env-Gag protein is expressed in E. coli as a single polypeptide without any evidence of premature termination or internal initiation. Env-Gag was purified on sodium dodecyl sulfate–10 to 20% polyacrylamide gradient gels under reducing conditions and migrated at approximately 33 kDa. The protein was further analyzed by high-performance liquid chromatography on a model TSK 4000 column (Tusohaas, Montgomeryville, Pa.) in an Na phosphate buffer, pH 6.8, in the presence of 10 mM dithiothreitol and produced a single peak. The protein content of our preparation was about 0.27 mg/ml based on the optical density at 280 nm, with excitation maxima being equal to 0.487. The protein is highly conserved and antigenic. The rationale for selecting Env-Gag in this investigation is that this compound has been a primary reagent in our hands, yielding 100% reactivity by immunoblot assays with HIV-positive sera from numerous donors (6). Further, Env-Gag produced significant proliferative responses and suppression of pokeweed mitogen-induced immunoglobulin synthesis by normal lymphocytes (18, 26). As controls, a peptide derived from the E. coli expression vector (HIV-1 vector 279–291) used to clone the Env-Gag recombinant fusion peptide and a non-HIV-1, rubeola virus antigen were used. All peptides were stored at −70°C in RPMI 1640 medium containing 0.1% bovine serum albumin. For each experiment a small aliquot was thawed and used. Stored Env-Gag was further tested on a sodium dodecyl sulfate–12.5% polyacrylamide gel under reducing conditions and was shown to produce only a single band at the expected region (∼33 kDa), demonstrating that Env-Gag was not degraded on storage.

Blood donors.

Peripheral blood from healthy, HIV-1-seronegative donors was drawn into a syringe containing heparin (20 U/ml). All donors were apprised of this study, and informed consents were obtained in a manner consistent with the policies of the National Institutes of Health and the State University of New York at Buffalo. Blood donors were not taking nonsteroidal anti-inflammatory agents, corticosteroids, opiates, or drugs of abuse at the time of the study.

Isolation of lymphocytes.

Peripheral blood mononuclear cells were isolated from heparinized venous blood by a modified method of Boyum (2). Blood was diluted with an equal volume of normal saline and was centrifuged at 400 × g for 30 min at 18°C. The mononuclear cell band was harvested, washed three times with saline, and resuspended in RPMI 1640 medium containing 25 mM HEPES buffer supplemented with 10% heat-inactivated fetal calf serum (GIBCO), 80 μg of gentamicin (Schering, Kenilworth, N.J.) per ml, and 300 μg of fresh glutamine per ml (complete medium).

Peripheral blood mononuclear cells in RPMI 1640 with 10% fetal calf serum were depleted of adherent cells by passage through a 7-ml column of Sephadex G-10 beads (Pharmacia Fine Chemicals, Piscataway, N.J.). After 45 min of incubation at 37°C, nonadherent, peripheral blood lymphocytes were eluted with 1 bed volume of medium (7 ml, equivalent to 1 Sephadex G-10 column volume) at 37°C. Cell recovery was ∼70% of the total input, and monocyte contamination of the resultant peripheral blood lymphocytes, as indicated by nonspecific esterase staining, was <2% (21).

Enrichment of NK cells.

Enrichment of LGL with a discontinuous gradient of Percoll (Pharmacia) was carried out as described previously (27). LGL preparations were cytocentrifuged and stained with Giemsa stain, the smear was examined with an oil immersion microscope, and at least 200 cells per slide were counted. The LGL were identified according to established cytological criteria such as azurophilic and cytophilic granules, kidney-shaped nuclei, and a low nucleus/cytoplasm ratio. Further, preparations were stained with anti-CD56 monoclonal antibodies specific for NK cells to confirm the identity of the LGL. In some experiments LGL-enriched populations were additionally purified by negative selection with monoclonal antibodies (Coulter Immunology, Hialeah, Fla.) and by complement-dependent lysis. The latter procedure involves treating 10⁷ LGL with anti-CD3 and anti-CD15 monoclonal antibodies for 30 min at 4°C and then incubating them with rabbit complement (Lowtox H; Cedarlane Laboratories Ltd., Hornby, Ontario, Canada) for 45 min at 37°C. The resulting population, depleted of T cells and monocytes, was analyzed by fluorescence-activated cell sorter analysis. About 80% of the cells of this fraction were positive by staining with anti-CD56 monoclonal antibody, and less than 0.1% were positive for anti-CD15 monoclonal antibody.

Treatment of LGL with Env-Gag peptide.

LGL were suspended in complete medium at a concentration of 10⁶ cells/ml, to which Env-Gag and control peptides were added at final concentrations of 0, 5, 10, and 50 ng/ml. LGL cultures also received human IFN-α (5 to 500 U/ml) or IL-2 (8 to 128 U/ml) alone or in combination. In experiments to reverse the Env-Gag-induced inhibition of NK cell activity by cytokines, LGL were treated with cytokines at the same time that Env-Gag was added. The cultures were incubated for 24 h in a humidified environment of 5% CO₂ in air at 37°C. Cells were washed and resuspended in complete medium. Cell viability was comparable to that of untreated control cultures as determined by trypan blue dye exclusion and was not affected by the peptide concentrations used in this study. Treated and control cultures were assayed for NK cell activity.

Tumor target cells.

The human erythroleukemia cell line K562 was used as the target for NK cells. To 0.8-ml aliquots of complete medium containing 5 × 10⁶ tumor cells, 200 μCi of ⁵¹Cr as sodium chromate (New England Nuclear, Boston, Mass.) was added. The cells were incubated at 37°C for 1 h in a humidified atmosphere of 5% CO₂ in air with intermittent shaking. After incubation, the cells were washed three times with complete medium and resuspended to a concentration of 2 × 10⁵ cells/ml.

Assay for NK cell activity.

NK cell activity was determined in a direct ⁵¹Cr-release assay as described previously (21). Briefly, various concentrations of viable effector cells in complete medium were added to triplicate cultures of ⁵¹Cr-labeled target cells in 0.2-ml volumes in V-bottom microtitration plates (Dynatech Laboratories, Alexandria, Va.). After centrifugation at 40 × g for 2 min, the cells were incubated at 30°C in a humidified atmosphere of 5% CO₂ in air for 4 h. Percent cytotoxicity was calculated as (experimental release − spontaneous release)/(total release − spontaneous release) × 100, where total release represents counts obtained in an aliquot of 10⁴ target cells and spontaneous release represents counts released into the supernates of control wells containing only 10⁴ target cells. The spontaneous release was always less than 5% of the total release. Because the maximum counts (total release) determined by either lysing an aliquot of labeled target cells or using unlysed target cells were the same, the latter method was employed to asses total release. Cytotoxicity was expressed as lytic units (LU) per 10⁷ effector cells and quantified as the number of effector cells needed to yield 30% cytotoxicity of 10⁴ target cells. LU were calculated from the cytotoxicity curve by using four different effector cell/target cell ratios for each test and linear regression analysis as described previously (14).

RESULTS

Data presented in Table 1 show the effect of HIV-1 Env-Gag peptide, denatured Env-Gag peptide, HIV-1 vector peptide, and a non-HIV control peptide on the NK cell activity of LGL from normal donors. LGL were precultured separately with different concentrations of control or HIV-1 Env-Gag peptides for 24 h and were washed and tested for their NK cell activity at 100:1, 50:1, 25:1, and 10:1 effector cell/target cell ratios, and LU were calculated. LGL precultured with 10 or 50 ng of denatured HIV-1 Env-Gag peptide, HIV-1 vector peptide, or rubeola virus antigen per ml demonstrated levels of NK all activity similar to that of LGL cultured in medium alone. However, LGL precultured with 5, 10, or 50 ng of Env-Gag per ml demonstrated decreased cytotoxicities of 171, 93, and 104 LU, respectively (11, 51, and 45% inhibition of cytotoxicity), compared to the 193 LU manifested by untreated control LGL.

TABLE 1.

Suppression of the NK cell activity of LGL by Env-Gag^a

Peptide or antigen	Conc (ng/ml)	Cytotoxicityb	% Suppressionc	P valued
Env-Gag	0	193.2 ± 15.0
	5	171.5 ± 11.7	11.2	<0.119
	10	93.4 ± 8.9	51.6	<0.032
	50	104.5 ± 12.7	45.9	<0.001
Denatured Env-Gag	10	190.2 ± 14.7	1.5	NS
	50	186.0 ± 12.5	3.7	NS
HIV-1 vector peptide	10	201.7 ± 18.0	−4.1	NS
	50	189.5 ± 11.1	1.8	NS
Rubeola virus antigen	10	190.3 ± 21.0	1.5	NS
	50	178.6 ± 13.9	7.5	NS

^aLGL (10⁶) from normal subjects were cultured in medium alone or with different concentrations of either Env-Gag, denatured Env-Gag (boiled for 10 min), the HIV-1 vector peptide, or rubeola virus antigen for 24 h, washed, and tested for NK cell activity against K562 target cells. 

^bNK cell cytotoxicity was calculated in LU. Values are means ± standard deviations of results of three experiments. 

^cPercent suppression was calculated in comparison with the level of NK cell activity in an untreated control culture. Both the controls and the Env-Gag peptides were also tested at the same molar concentrations and compared with protein concentrations, and the results showed that Env-Gag produced similar levels of NK cell inhibition but that control peptides did not produce any inhibition of NK cell activity (data not presented). 

^dStatistical significance of differences was calculated by Student's t test. NS, nonsignificant. 

Data presented in Table 2 show the dose-response effect of human IFN-α and IL-2 on the NK cell activity of LGL. IFN-α produced a dose-dependent enhancement of NK cell activity of LGL at 50, 100, and 250 U/ml, with the percent increases being 5, 17, and 41%, respectively, compared to the level of NK cell activity of untreated control LGL. IFN-α at a higher concentration (500 U/ml) produced only a 33% enhancement of NK cell activity. LGL precultured with IL-2 at 8, 32, and 64 U/ml produced a dose-dependent enhancement of NK cell activity (3, 13, and 37% increases, respectively) compared to the level of activity in untreated control LGL. LGL cultured with IL-2 at 128 U/ml produced only a 22% enhancement of NK cell activity compared to the level in untreated control LGL. IFN-α and IL-2 at 250- and 64-U/ml concentrations, respectively, were selected as maximum NK cell-activating doses in our subsequent experiments with Env-Gag.

TABLE 2.

Enhancement of the NK cell activity of LGL by IFN-α and IL-2^a

Treatment of LGL	Concn (U/ml)	Cytotoxicityb	% Enhancementc	P value
Medium		213.9 ± 17.4
IFN-α	50	225.2 ± 16.7	5.2	<0.463
	100	251.8 ± 15.3	17.7	<0.047
	250	302.5 ± 19.2	41.4	<0.004
	500	284.7 ± 18.7	33.1	<0.009
IL-2	8	221.6 ± 16.0	3.5	<0.603
	32	242.6 ± 22.6	13.4	<0.156
	64	294.3 ± 18.4	37.5	<0.005
	128	262.7 ± 19.2	22.8	<0.031

^aLGL (10⁶) from normal subjects were cultured in medium alone or with recombinant human IFN-α (Roferon; Hoffman LaRoche, Inc., Nutley, N.J.) or recombinant IL-2 alone (Collaborative Research, Inc., Boston, Mass.) at different concentrations for 24 h, washed, and tested for NK cell activity against K562 target cells. 

^bNK cell cytotoxicity was calculated in LU. Values are means ± standard deviations of results of three experiments. 

^cPercent enhancement was calculated in comparison with the level of NK cell activity in the untreated control culture. 

Data presented in Table 3 show the separate and combined effects of IFN-α and IL-2 on Env-Gag-induced suppression of the NK cell activity of LGL. LGL treated separately with 250 U of IFN-α per ml or 64 U of IL-2 per ml produced 290 or 310 LU, respectively, resulting in a significant enhancement of their NK cell activity (38 or 48% enhancement, respectively) compared to the level of activity in the untreated control culture (209 LU). LGL cultured with IFN-α plus IL-2 produced 340 LU (62% enhancement) compared to the 209 LU produced by the untreated control culture. LGL treated with Env-Gag alone produced significant inhibition of NK cell activity (98 LU; 53% suppression) compared to the activity in the untreated control LGL (290 LU). LGL treated with IFN-α or IL-2 in the presence of Env-Gag produced a moderate reversal of Env-Gag-induced inhibition of NK cell activity, with the cytotoxicities being 198 or 201 LU compared to the 290 or 310 LU, respectively, produced by LGL treated with IFN-α or IL-2 alone. Thus, IFN-α or IL-2 when used individually only partially reversed Env-Gag-induced suppression of the NK cell activity of LGL, i.e., by 31 and 35%, respectively, compared to the 53% suppression produced by Env-Gag alone. However, the combined addition of IFN-α and IL-2 to Env-Gag-treated culture completely reversed the suppression of NK cell activity, producing 337 LU compared to the 340 LU produced by cultures treated with IFN-α and IL-2 in the absence of Env-Gag (0.8% suppression). These results demonstrate that IFN-α and IL-2 in synergy can completely reverse the Env-Gag-induced NK cell inhibition.

TABLE 3.

Combined effect of IFN-α and IL-2 on Env-Gag peptide-induced suppression of NK cell activity by LGL^a

Treatment of LGL	Cytotoxicityb	% Changec	P valued
None (control)	209.8 ± 15.7
IFN-α (250 U/ml)	290.4 ± 18.8	+38	<0.01
IL-2 (64 U/ml)	310.3 ± 21.0	+48	<0.01
IFN-α + IL-2	340.6 ± 22.4	+62	<0.001
Env-Gag (10 ng/ml)	98.2 ± 10.3	−53	<0.001
Env-Gag + IFN-α	198.6 ± 12.5	−31	<0.002
Env-Gag + IL-2	201.4 ± 15.7	−35	<0.002
Env-Gag + IFN-α + IL-2	337.7 ± 19.3	−0.8	<0.9

^aLGL (10⁶) from normal subjects were cultured in medium alone or with IFN-α alone, IL-2 alone, IFN-α plus IL-2, Env-Gag alone, Env-Gag plus IFN-α, Env-Gag plus IL-2, and Env-Gag plus IFN-α plus IL-2 for 24 h, washed, and tested for NK cell activity against K562 target cells. 

^bNK cell cytotoxicity was calculated in LU. Values are means ± standard deviations of results from five experiments. 

^cA + sign indicates enhancement relative to the level of NK cell activity in the control, and a − sign refers to suppression relative to the level of activity in the control (−53%) or in LGL treated with IFN-α (−31%), IL-2 (−35%), or IFN-α plus IL-2 (−0.8%). 

^dStatistical significance of differences was calculated by Student's t test. Values are versus the control value except for the probability for Env-Gag plus IFN-α (versus the IFN-α value), Env-Gag plus IL-2 (versus the IL-2 value), and Env-Gag plus IFN-α plus IL-2 (versus the IL-2 plus IFN-α value). 

DISCUSSION

Numerous studies of the immunomodulatory effects of various HIV-1-derived proteins on lymphocytes from healthy and HIV-1-infected subjects have been reported (11, 13, 16, 24). We have recently reviewed the biological effects of different HIV-1 peptides (33). Earlier we demonstrated that lymphocytes from HIV-1-infected patients have reduced NK cell activity compared to those of normal subjects (24). In addition to Env-Gag, the HIV-1 synthetic peptides Env-487–511 and Env-647–659, where the numbers correspond to different conserved amino acids of gp41, also produced significant suppression of the NK cell activity of normal lymphocytes. In the present investigation, we demonstrate that the cytotoxic activities of NK cell-enriched LGL can be significantly suppressed by the HIV-1 Env-Gag peptide and that this effect can be reversed by biological response modifiers such as IFN-α and IL-2. The Env-Gag-induced suppression of NK cell activity is not merely due to toxic effects on effector cells, as Env-Gag-treated cells were as viable as untreated control cultures (data not shown). Further, the suppression of cytotoxicity was also not due to effects on target cells, as targets treated with Env-Gag for 4 h were as sensitive to lysis by effector cells as untreated, control target cells (data not shown). We reported earlier that lymphocytes from HIV-1-infected subjects in comparison with lymphocytes from healthy controls showed increased sensitivity to the inhibitory effects of HIV-1 peptides (24). We also showed that lymphocytes from intravenous drug users manifested significantly lower NK cell and antibody-dependent cellular cytotoxic activities than those of age- and sex-matched normal controls and that the suppressed NK cell activities could be partly reversed by treating their lymphocytes with IL-2 (23). Others have shown that NK cell activity could be significantly modulated by IFN and IL-2 (29, 39). In the present investigation, we demonstrate that Env-Gag-induced suppression of the NK cell activity of LGL can be partly reversed by treatment with either IFN-α or IL-2 and completely reversed to normal levels by a treatment with a combination of IFN-α plus IL-2.

It is now clear that an important consequence of HIV infection is dysregulation of the production of various cytokines that otherwise help to prolong the asymptomatic period of infection or assist in the clearance of the virus (7, 15, 34, 38, 41). Deficiency in the production of IL-2 or its receptor (12, 31) and of IFN-α and IFN-γ (20) have been reported. It is generally accepted that Th1 cytokines such as IL-2, IFN-γ, and IL-12 promote cellular immunity and cytotoxic activities in T cells but that Th2-derived cytokines such as IL-4, IL-5, and IL-6 exert negative immunoregulatory effects on cellular immunity in HIV-1 disease (7, 15). Further, decreases in NK cell activities during HIV-1 infections were reported to be associated with dysregulation of Th1-derived cytokines and/or defects in the cytolytic machinery involved in the killing function (for a review, see reference 3). This finding suggests that an imbalance or lack of IFN and/or IL-2 may be involved in the suppressed NK cell activity observed in patients with HIV-1 infection and that treatment with IL-2 plus IFN may significantly restore the cytotoxic functions of NK cells. Thus, as shown in our experiments, Env-Gag-induced NK cell suppression could be completely reversed by the Th1-derived cytokine IL-2 plus IFN-α.

Reports on the effects of cytokine on the suppressed NK cell activity in HIV-1-infected patients have been conflicting (18, 32, 43). Zaizov et al. (43) showed that IL-2 can partially restore the in vitro NK cell activity of lymphocytes from HIV-infected patients. IL-2 has been shown to partially overcome the suppression of NK cell activity caused by an HIV gp41 synthetic peptide corresponding to amino acid sequences 735 to 752 and 846 to 860 (5). Lew et al. (18), however, reported the absence of responsiveness to IL-2 and IFN by NK cells from HIV-infected subjects. The synergistic effects of cytokines on the NK cell activity of lymphocytes have been reported earlier. Zier and Gransbacher (44) showed a synergy between IL-2 and IFN-γ, leading to the rejection of tumor cells. IL-2 has been shown to down regulate expression of the CD4 receptor and the HIV-1 entry coreceptor CCR5 (17), demonstrating that IL-2 may also be effective in inhibiting HIV-1 infection. Further, the NK cell activity of patients with HIV infection was also partially restored with IFN-β and -γ (30). Lin et al. (19) demonstrated an NK cell-enhancing effect when cells were treated individually with either IL-15 or IL-12; when used together, these cytokines also produced a synergistic effect. Stine et al. (37), however, reported that gp120-specific cell-mediated cytotoxicity was not enhanced by overnight treatment of lymphocytes with IFN-γ alone; however, IFN-γ plus IL-2 produced significant enhancement of cytotoxicity. In our experiments, the Env-Gag-induced NK cell suppression could be completely reversed by treatment with a combination of IFN-α and IL-2. Combining biological response modifiers in HIV therapy not only might increase their therapeutic effectiveness but also might diminish their side effects by reducing the dose of each cytokine needed to produce a therapeutic effect. A better understanding of the mechanisms underlying the NK cell response to biological immune modifiers such as IFN and IL-2 may help us to design novel therapeutic strategies to stimulate the depressed NK cell activity often observed in HIV-1-infected patients. Although the maximal potential for the reversal of depressed immune responses in HIV- infected patients who have ongoing CD4 attrition is not clear, it is possible that intervention strategies with a combination of IFN-α and IL-2 will reverse disease progression at least in patients with asymptomatic infection and whose CD4 numbers are ≥400/mm³.