Control of viremia and maintenance of intestinal CD4⁺ memory T cells in SHIV_162P3 infected macaques after pathogenic SIV_MAC251 challenge

Abstract

Recent HIV vaccine failures have prompted calls for more preclinical vaccine testing in nonhuman primates. However, similar to HIV infection of humans, developing a vaccine that protects macaques from infection following pathogenic SIV_MAC251 challenge has proven difficult, and current vaccine candidates at best, only reduce viral loads after infection. Here we demonstrate that prior infection with a chimeric simian-human immunodeficiency virus (SHIV) containing an HIV envelope gene confers protection against intravenous infection with the heterologous, highly pathogenic SIV_MAC251 in rhesus macaques. Although definitive immune correlates of protection were not identified, preservation and/or restoration of intestinal CD4⁺ memory T cells were associated with protection from challenge and control of viremia. These results suggest that protection against pathogenic lentiviral infection or disease progression is indeed possible, and may correlate with preservation of mucosal CD4⁺ T cells.

INTRODUCTION

Although it has been over 25 years since the beginning of the AIDS epidemic, an effective vaccine remains elusive. To date, all HIV-1 vaccine candidates tested in humans have failed to prevent infection or meaningfully impact viral replication (Koff et al., 2006). Furthermore, most vaccine approaches evaluated in non-human primate models, including protein subunits, inactivated particles, DNA, and recombinant live vectors, alone or in various combinations, have shown at best only limited, partial efficacy, particularly in studies involving stringent challenges with pathogenic SIV strains such as SIV_MAC239, SIV_MAC251 or SIV_SME660 (Casimiro et al., 2005; Egan et al., 2000; Horton et al., 2002; Koff et al., 2006; Larke et al., 2005). The recent disappointing failure of the STEP clinical trial of the Merck Ad5 vectored candidate vaccine, despite induction of readily measurable virus specific cellular immune responses, emphasizes we still do not understand the basis of protective immunity against HIV infection (Cohen, 2007; Watkins, 2008).

In contrast, use of live attenuated viruses as vaccines has demonstrated more robust protection from rigorous homologous and heterologous viral challenges in macaques, thus providing critical proof-of-concept for the feasibility of an HIV vaccine to prevent or limit HIV infection (Abel et al., 2003; Connor et al., 1998; Daniel et al., 1992; Koff et al., 2006; Kumar et al., 2002; Marthas et al., 1992; Mori et al., 2001; Wyand et al., 1996). However, most of these studies utilized relatively insensitive methods for viral detection, or protection was defined as a reduction of ~1 log plasma viral load compared to controls. Moreover, relatively few studies have investigated the potential mechanisms of this protection with contemporary assays for virus-specific cellular immune assays or antiviral antibody. Abel et al have demonstrated that “immunization” with the attenuated virus SHIV_89.6 can confer protection in a significant fraction of animals against subsequent intravaginal (IVAG) challenge with the highly pathogenic SIV_MAC239 (Abel et al., 2003), demonstrating that protection can be achieved against mucosal transmission of a highly pathogenic partially heterologous virus that is sequence divergent in env and regulatory genes. However, since the challenge was IVAG, it is not clear whether this protection was attributable to mucosal or systemic immunity. A model that demonstrated protection against intravenous (IV) challenge would provide even more evidence that protection against highly pathogenic SIV challenge is achievable, and would allow assessment of the role of systemic immune responses in such control. Our recent study showed that rhesus macaques were highly susceptible to vaginal transmission of SHIV_162P3, yet ~90% had low to undetectable plasma viremia within 70–90 days of infection (Pahar et al., 2007). We hypothesized that these animals may thus be “immunized” against subsequent challenge with SIV_MAC251, which could provide a model for assessing correlates of immunity to this highly pathogenic challenge. Alternatively, we hypothesized that prior infection could immunocompromise animals, and result in an exacerbated course of infection in response to SIV_MAC251 challenge, thus providing a model for studying HIV “superinfection”, defined as when a currently HIV-infected person becomes infected with a second strain of HIV from a subsequent exposure.

While superinfection of HIV-1 infected patients with a second strain of HIV-1 has implications for vaccine development (Gottlieb et al., 2007), superinfection can be difficult to prove, as it is often difficult to distinguish whether viral variants are the result of subsequent infection or simply represent mutations of the initial infecting virus (Gottlieb et al., 2007). Furthermore, the consequences of superinfection remain unclear (Gottlieb et al., 2007). Nonhuman primate models provide an opportunity to address these questions in an experimentally tractable fashion. Here we demonstrate that macaques primarily infected with the minimally pathogenic virus SHIV_162P3 conferred partial to complete protection against subsequent IV challenge with the highly pathogenic SIV_MAC251. Furthermore. We demonstrate that mucosal CD4⁺ memory T cells were maintained following pathogenic challenge, suggesting that mucosal immune responses play a role in this protection. Finally, these results indicate that immunity to infection with SIV/HIV may indeed be possible, and additional studies with these and similar models of protection may provide insights into the immune correlates of protection from HIV infection.

RESULTS

“Immunization” with SHIV_162P3

Rhesus macaques of Indian (InRh) or Chinese (ChRh) were atraumatically inoculated intravaginally with 300 TCID₅₀ SHIV_162P3 and confirmed to be infected, as evidenced by high peak plasma viremia between 14 and 21 days post inoculation ranging from 0.45 to 6.3 x 10⁶ viral RNA (vRNA) copies/ml plasma, proving that none of the animals were genetically “resistant” to SHIV/SIV infection (Figs. 1 and 2). Viral loads were undetectable in 6 out of 14 SHIV_162P3 inoculated animals by 70 days post inoculation (Fig. 2). All macaques were followed for 89 to 345 days post inoculation, until plasma viral load (VL) declined to levels from undetectable to 7,600 vRNA copies/ml of plasma. This initial infection is subsequently referred to as the “immunization” phase.

Figure 1.

Experimental design of the SHIV_162P3 inoculation and SIV_MAC251 challenge. SHIV infected macaques were subsequently intravenously challenged with different dosages of pathogenic SIV_MAC251 on the days post inoculation indicated in three separate experiments. For controls, four naïve Chinese origin macaques were concurrently inoculated with SIV_MAC251 (highlighted in red).

Figure 2.

Plasma viral RNA levels in macaques after primary SHIV_162P3 infection, prior to pathogenic SIV_MAC251 challenge. All animals had high primary viremia 14–21 days after vaginal inoculation, which declined to low or undetectable levels by 70 days of infection. Viral loads were undetectable in 6 out of the 14 animals by 70 days post inoculation.

In a series of three sequential experiments, all “immunized” macaques were then intravenously challenged with SIV_MAC251. The three experiments were designed to determine whether the dose, source, or passage of the inoculum contributed to protection from rechallenge. In the first experiment (Group A), seven previously “immunized” macaques (four ChRh and three InRh) were IV challenged with a high dose (100 TCID₅₀) of SIV_MAC251. On the day of challenge, the InRh macaques (AK46, R191 and T552) had higher residual plasma VLs from the prior SHIV_162P3 immunization (0.03–0.76 x 10⁴ vRNA copies/ml of plasma) compared to the four ChRh macaques, two of which had undetectable (AJ61 and DT93) or very low levels (340–650 vRNA copies/ml plasma) on the day of challenge (Fig. 3A). As a concurrent control, a naïve (non-immunized) ChRh macaque (V780) was IV inoculated with 100 TCID₅₀ of SIV_MAC251.

Figure 3.

Plasma viral RNA levels in macaques following SIV_MAC251 challenge. Plasma viremia is reported as log₁₀ vRNA copies per ml of plasma (detection limit 30 vRNA copies/ml). Plasma viral loads from individual experiments (A) and mean +/− SE of the mean viral loads for each of the different groups (B) are shown. Note there was significantly higher viremia in non-immunized, control macaques (red lines) compared to immunized macaques during early (day 14), steady state (day 42) and late (day 282) stages of infection following SIV_MAC251 challenge. Plasma viral load from the 3 InRh in Group A is presented as a dashed line. (B) Mean plasma viral loads from 4 historical naive ChRh intravenously infected with 100TCID₅₀ SIV_MAC251 are shown as dashed purple line.

Following challenge, one ChRh macaque (DT93) showed no measurable viremia (<30 vRNA copies/ml) through >640 days of follow-up. The other 3 ChRh macaques showed excellent control of viral replication, with plasma viremia 2–3 logs lower than the naïve control throughout primary infection. All seven macaques immunized by prior infection with SHIV_162P3 showed greater levels of resistance or control of the high dose IV SIV_MAC251 challenge, at least through 154 days of challenge (Fig. 3A). In addition, five of the seven macaques did not demonstrate a “peak” of viremia in the first weeks following challenge. Only two macaques (InRh macaque T552 and ChRh macaque AJ61) showed any increased viremia after SIV_MAC251 challenge. Plasma viremia declined or remained low in one macaque (DT55) throughout 154 days after challenge and increased thereafter (Fig. 3A). These results suggest that substantial control of viremia following a high dose IV challenge with a pathogenic, heterologous virus can be achieved through prior immunization with an attenuated/weakly pathogenic SHIV.

The second experiment (Group B) was designed to determine if greater control could be achieved following a lower IV challenge. Four ChRh macaques previously “immunized” by SHIV_162P3 infection, and one SIV-naïve ChRh control were IV inoculated with 10 TCID₅₀ of the same SIV_MAC251 challenge stock used in Group A. On the day of challenge, all four “immunized” macaques had undetectable (FA10) or low levels of plasma viremia (40–340 vRNA copies/ml). Following IV SIV_MAC251 challenge, three out of four immunized macaques had no significant increase in plasma viremia over pre-challenge levels (i.e., no peak viremia), and their vRNA was intermittently undetectable in plasma (<30 vRNA copies/ml) on multiple occasions (Figs. 3A and B). Plasma VLs in CF58 remained less than 1000 copies/ml after challenge through day 42, but then gradually climbed to approximately 10,000 copies/ml by day 285 and this animal developed AIDS 2.5 years after SIV challenge (Fig. 3A). The three remaining SHIV immunized and SIV_MAC251 challenged macaques in this group have remained healthy through 3 years of follow-up, and have had undetectable plasma viremia. In contrast, the naïve ChRh control macaque had high peak and persistent set point plasma viremia indistinguishable from the control in Group A, and both controls were subsequently euthanized with signs and lesions of AIDS. Combined, the results of studies in Group A (high IV dose) and B (low IV dose) demonstrated that resistance to IV superinfection (1 macaque in group A) and/or remarkable control of plasma viremia (3 macaques in Group B) could be conferred by prior infection with a virus having a divergent envelope. Moreover, even in animals that were persistently infected, plasma viremia remained 1–3 logs lower during the early phase of infection than in control macaques inoculated with the same lot, route, and dose of virus.

To rule out potential artifacts of immune control resulting from using ChRh vs. InRh hosts, the third experiment (Group C) utilized an inoculum passaged in ChRh. Plasma VLs have been shown to be lower in ChRh than InRh macaques inoculated with SIV grown in InRh peripheral blood mononuclear cells (PBMC) (Ling et al., 2002), and thus the differences we observed could be due to adaptive “fitness” of the particular stocks for InRh macaque cells. Thus, we challenged five ChRh macaques intravenously with 1 ml plasma containing ~2.0 x10⁶ vRNA copies/ml of SIV_MAC251 which was obtained from a ChRh macaque on a serial passage study. This experiment was also important to compare the results of virus obtained from plasma compared to in vitro expanded viral stocks in generating infection and potential differences in protection from this challenge.

Of the 5 ChRh macaque animals inoculated with ChRh macaque-derived plasma virus, three were previously “immunized” with SHIV_162P3 and two were naïve controls. Of the three “immunized” macaques, two had undetectable plasma viremia (< 30 vRNA copies/ml), while one (L375) had 1600 copies/ml remaining at the time of challenge. Following challenge, both animals with undetectable viremia on the day of challenge continued to control viremia to less than 1000 copies/ml after challenge. The third (L375) also demonstrated VLs 1–2 logs lower than the naïve controls until day 190 post challenge (p.c.) (Fig. 3A) but then plasma VL increased to levels similar to naïve challenged macaques. In contrast, both naïve controls demonstrated uncontrolled viral replication and 1 log higher vRNA plasma copies compared to that of the controls in Groups A and B (Fig. 3A).

Highly significant differences in peak and post peak VLs were detected between controls and immunized challenged macaques (p = 0.015 to 0.003, Fig. 3B). When the InRh were excluded from the analysis, these differences were even more significant. Prior immunization with SHIV_162P3 resulted in a 1.8 to 3.7-log reduction in plasma VLs in all immunized macaques compared to naive controls through 156 days p.c. regardless of whether animals were challenged with high or low IV doses, or SIV_MAC infected plasma (Fig. 3B).

We also examined cell-associated viral loads in intestinal CD4⁺ T cells on a subset of the animals where sufficient cells were available (Fig. 4A). Copies of SIVgag RNA were low (FA14, 50 copies of SIVgag RNA) to undetectable in both SHIV immunized and naïve challenged macaques (Fig. 4A) on the day of SIV_MAC251 challenge. By 21 days post challenge all 3 immunized macaques had low (FA14, 90 copies of SIVgag RNA) to undetectable levels of virus whereas, naïve challenged macaques (FG46 and FA06) had much higher levels in intestinal tissues (Fig. 4A). Similar differences in intestinal VLs were detected in both SHIV immunized and naïve challenged macaques after 56 and 104 days of SIV_MAC251 challenge (Fig. 4A).

Figure 4.

(A) Cell-associated viral RNA levels in jejunum LPL in macaques following SIV_MAC251 challenge (detection limit 30 vRNA copies). There was low to undetectable viremia in intestinal tissues at the day of SIV_MAC251 challenge in SHIV immunized macaques. By 21, 56 and 104 days post SIV_MAC251 challenge increased viral replication was noticed in naïve challenged macaques, whereas SHIV_162P3 immunized macaques had low to undetectable tissue viral load. Mean tissue VLs are shown with a line in each plot. (B) Kaplan-Meier survival curves for the immunized and non-immunized macaques in all three groups after SIV_MAC251 challenge demonstrating significantly increased survival of immunized macaques compared to controls. Survival comparisons were performed using a log-rank test.

There was also a highly significant difference in survival between immunized and naïve macaques (p=0.009) (Fig. 4B). Of the 14 “immunized” macaques challenged with SIV, 5 remain alive and healthy (3 years since challenge); five were euthanized for unrelated reasons 2 years after infection (with no detectable plasma viremia or clinical signs related to SIV infection). Only 4 immunized animals were euthanized with AIDS defining illnesses (DT55, T552, CF58 and L375). In contrast, all 4 of the SIV naïve control ChRh macaques were euthanized due to AIDS-defining illnesses (opportunistic infection, lymphoma, and/or SIV encephalitis) within 14 months of challenge.

Maintenance of systemic and intestinal CD4⁺ T cells in immunized macaques

Although animals in groups A–C were challenged with SIV_MAC251 at very different time points after SHIV_162P3 immunization, all were challenged at time points when plasma VLs had declined to low or undetectable levels. To assess whether the limitation of viral replication observed in SHIV_162P3 immunized animals following SIV challenge reflected a paucity of viral target cells induced by the initial SHIV_162P3 infection, percentages of total CD4⁺ T cells were compared in peripheral blood, intestine (jejunum) and axillary lymph node (LN) in naïve and SHIV immunized macaques immediately prior to, and subsequent to SIV challenge. Prior to challenge, percentages of total CD4⁺ cells in blood were stable in all animals (mean CD4⁺ T cell % in group A, B and C was 51.7% (range 28.7–63.6), 56.4% (range 49.5–67.2) and 54.7% (range 49.3–62.2) respectively), reflecting the low level of plasma viremia at the time of challenge. Total CD4⁺ T cell populations in LN also were relatively normal in immunized macaques at the time of challenge. Following SIV_MAC251 challenge, percentages of CD4⁺ T cell counts dropped significantly (P<0.01) in the peripheral blood of all naïve controls, yet remained stable in all immunized animals through at least 119 days p.c. (Fig. 5A). However, six macaques (AK46, CF58, DT55, L375, R191 and T552) had substantial drops in CD4⁺ T cells 250–402 days after challenge, which inversely correlated with increasing VLs at these timepoints. Similarly, there were decreased percentages of CD4⁺ T cells in the LN of naïve macaques, yet CD4⁺ T cells remained fairly stable in immunized macaques (Fig. 5B).

Figure 5.

Dynamics of CD4⁺ T cell percentages in peripheral blood (A) and axillary lymph node (B) of immunized and non-immunized macaques following SIV_MAC251 challenge. Note all non-immunized (naïve) challenged macaques had a rapid decrease in CD4⁺ T cells by day 14 after challenge in both peripheral blood and LN (red lines) whereas CD4⁺ T cells were preserved in immunized macaques. InRh macaques are delineated by dashed lines.

Surprisingly, percentages of CD4⁺ T cells in the intestine at the time of challenge were highly variable in immunized animals, yet they did not appear to correlate with levels of plasma viremia. Several SHIV immunized macaques had lower levels of CD4⁺ T cells in intestinal lamina propria lymphocytes (LPL) compared to naïve controls at the time of challenge. Selective depletion of intestinal CD4⁺ T lymphocytes following SHIV_162P3 infection has been previously reported (Hsu et al., 2003). Intestinal CD4⁺ T cell percentages were similar in groups A (mean 25%, ranged from 10.0–42.1%) and B (mean 26.7%, ranged from 12.3–39.7%) whereas naive macaques V780 and V302 had around 33% intestinal CD4⁺ T cells at the time of SIV_MAC251 challenge (Figs. 6A and B). However, intestinal CD4⁺ T cells in group C were lower compared to animals in both A and B but not significantly different (p=0.62) (Fig. 6B). The two naïve controls in group C had higher (~43%) intestinal CD4⁺ T cells at the day of challenge (Fig. 6B). Following SIV_MAC251 challenge, essentially all SHIV-immunized macaques maintained near baseline levels of intestinal CD4⁺ T cells through at least 42 days of infection, whereas all naïve macaques demonstrated marked and persistent reductions in intestinal CD4⁺ T cells (Figs. 6A and B). In fact, percentages of intestinal CD4⁺ T cells increased by day 7 and 14 in several SHIV-immunized macaques (Fig. 6B). The main exceptions were in immunized macaques DT93 and R191, which showed moderate drops in intestinal CD4⁺ T cells by 14 days post challenge, yet these returned to baseline levels in the following weeks (Fig 6B). Similarly, macaques CF58 and FA05 had a transient drop in percentages of intestinal CD4⁺ T cells 7 days after challenge, yet these animals also recovered to levels higher than at the time of SIV challenge, indicating eventual control of replication in mucosal sites (Fig. 6B). Memory (CD45RA⁻) CD4⁺CCR5⁺ T cells in the intestine, the major early targets for SIV/HIV infection and replication (Brenchley et al., 2004; Mattapallil et al., 2005; Veazey et al., 2000a; Veazey et al., 2000b; Wang et al., 2007) were also relatively maintained in SHIV-infected macaques at the time of challenge (Figs. 7A and B) suggesting that the limitation of viral replication observed following pathogenic challenge was not attributable to a paucity of viral target cells. Overall, intestinal CD4⁺ T cell percentages from all naive control macaques (groups AC) demonstrated marked and significant losses by 14 days after challenge with SIV_MAC251 (p=0.007), whereas immunized macaques did not change significantly. Similarly, a highly significant difference (p=0.009) in the availability of CD4⁺CCR5⁺ target cells was detected between SHIV immunized and naïve challenged macaques by 14 days after SIV challenge indicating that immunized animals had a bona fide “resistance” to infection, associated with preservation/restoration of mucosal viral target cells (Figs. 6 and 7).

Figure 6.

(A) A representative dot plots showing CD4⁺ T cells as a percentage of total T cells (gated through CD3⁺ lymphocytes) in macaques (Group B) after SIV_MAC251 challenge. Note that restoration and/or preservation of intestinal CD4⁺ T cells was observed in a SHIV immunized macaque (FA08) compared to the non-immunized macaque (V302), in which there was rapid and sustained loss of intestinal CD4⁺ T cells within 14 days of challenge. (B) Changes in intestinal CD4⁺ T cells of immunized and non-immunized macaques following challenge with SIV_MAC251. Note that in all groups, there was a rapid decrease in intestinal CD4⁺ T cells in all non-immunized (highlighted in red) compared to immunized macaques by day 14-post challenge. Intestinal CD4⁺ T cells of InRh are shown as dashed lines.

Figure 7.

(A) Percentages of memory (CCR5⁺CD45RA^neg) CD4⁺ T cells in the intestine of a representative SHIV immunized (FA08) and non-immunized, naive (V302) macaque before (day 0) and 14 days after SIV_MAC251 challenge. Note the marked and rapid decrease in intestinal target cells (CD3⁺CD4⁺CCR5⁺CD45RA^neg) typical of naïve macaques challenged with SIV_MAC251, whereas these cells are maintained in the immunized macaque. (B) Percentages of intestinal CD4⁺CCR5⁺ cells (gated through CD3⁺ lymphocytes) are shown for groups A–C. Note a rapid decrease in CD4⁺CCR5⁺ T cells occurs in all non-immunized challenged macaques (highlighted in red) within 14 days of SIV_MAC251 challenge, whereas target cells are preserved in immunized macaques.

Investigations into potential “immune” control of SIV challenge

Several studies have suggested that neutralizing antibodies (Nabs) and/or T cell responses can play a role in controlling viral infection (Ferrantelli et al., 2004a; Ferrantelli et al., 2004b; Koup et al., 1994; Mascola et al., 2005; Schmitz et al., 1999). Nabs against the immunizing SHIV_162P3, the SIV_MAC251 challenge, as well as to laboratory-adapted SIV_MAC251 were measured for all macaques from groups A, B and C. However, neither SHIV-immunized or naïve challenged macaques produced significant levels of Nab to SHIV or SIV_MAC251 in any of the groups examined. Only one macaque (T552) had a transient low level of Nabs to SIV_MAC251 4 weeks after infection, but Nab titers ranged from <20 to 67 in all other SHIV vaccinated macaques. Similarly, low to undetectable levels of Nabs were detected to the SIV_MAC251 infection in naive challenged macaques.

On the day of SIV_MAC251 challenge, antigen specific IFN-γ responses to tat, gag, nef, and env were examined in peripheral blood, LN and intestine of all SHIV_162P3 immunized macaques in groups A & C. Overall, 90% of SHIV immunized animals demonstrated positive responses to SHIV-gag followed by HIV-env (70%) and HIV-tat (20%) antigens (Table 1). Despite these responses, neither the magnitude nor the presence of HIV-env and/or SIV-gag IFN-γ responses present at the time of challenge correlated with the level of viral control following SIV_MAC251 challenge.

Table 1.

Antigen specific IFN-γ responses on the day of SIV_MAC251 challenge from groups A and C macaques^a

Group	Macaques	Blood	Lymph Node	Jejunum LPL
A	AJ61	negative	568➄	negative
	AK46	negative	395➄	115➄
			55➉	45➉
	DT55	negative	negative	95
				115➄
				300➉
	DT91	negative	98➄	negative
	DT93	253➄	negative	70➉
		48❷
	T552	138	negative	negative
		300➄
		110➉
	R191	448➄	negative	negative
		210➉
	V780	negative	negative	negative
C	FA14	negative	302➄	608➄
			100➉	90➉
			115❷
	L375	207➄	122➂	720➄
		45❷	715➄
			57❷
	P015	negative	negative	negative
	FG46	negative	negative	negative
	FA06	negative	negative	negative

^aMacaques in red were unvaccinated (naïve) followed by SIV_MAC251 challenge. Other macaques were infected with SHIV followed by SIV_MAC251 challenge. Values represent responses to specific peptide pools; = Tat, ➂ = Nef, ➄ = Gag, ➉ =gp120, and ❷ = gp41. Numbers indicate spot forming cells per one million PBMCs after medium-only control values were subtracted.

Following challenge, IFN-γ responses in blood were monitored by ELISPOT for >118 days in groups A and C, and in selected animals, through 84 days of infection in lymph nodes (Tables 2 and 3). In general, gag-specific responses were higher in immunized macaques followed by tat, nef, gp120 and gp41 responses (Tables 2 and 3). SIV-env specific IFN-γ responses were detected in PBMC as early as 21 days after challenge, but only in one macaque (AK46). Similarly, 7 out of 10 immunized macaques had detectable antigen specific IFN-γ responses by day 14 after challenge in LN. However, there was no correlation between control of plasma vRNA and IFN-γ responses in any of the immunized or control macaques (r² ranged from 0.006 to 0.86 and p=0.06 to 0.89). We have also performed one way analysis of variance of the number of IFN-γ secreting cells in both vaccinated and unvaccinated challenged macaques in all the time point of this study but no significance difference in the number of IFN-γ positive cells at any time point (P>0.05) has been established.

Table 2.

Antigen specific IFN-γ responses in SIV_MAC251 challenged macaques in peripheral blood^a

Groups	Monkey	Peptide Pools	Days post challenge
			7	14	21	42	56	118	352
A	AJ61	Tat	48						ND
		Nef	ND	ND	ND	ND	ND	ND	ND
		Gag	195	180		350	420	118	ND
	AK46	Tat	153	58	148	50	112
		Nef	ND	ND	ND	ND	ND	ND
		Gag	540	170	625	130	362	278	230
		gp120			513		211
	DT55	Nef	ND	ND	ND	ND	ND	ND	210
		Gag	50	326		180	255	310	360
		gp120					52	62	98
	DT91	Nef	ND	ND	ND	ND	ND	ND	ND
		Gag		90		90	60		ND
		gp120							ND
		gp41							ND
	DT93	Nef	ND	ND	ND	ND	ND	ND
		Gag	103			118	82
	T552	Nef	ND	ND	ND	ND	ND	ND	93
		Gag				90	196	275	113
		gp120						83	53
		gp41							70
	R191	Nef	ND	ND	ND	ND	ND	ND
		Gag				53	222	428	338
		gp120				135	96	175	178
	V780	Tat						50
		Nef	ND	ND	ND	ND	ND	ND	73
		Gag						88
C	FA14	Nef			140	117	ND	135	ND
		Gag	102	65	357	420	ND	347	ND
	L375	Tat					ND	238	ND
		Nef		67			ND	217	ND
		Gag	152	247	182	450	ND	612	ND
		gp120					ND	105	ND
	FG46	gp120					ND	62	ND

^aPeptide–specific responses are depicted for PBMCs in one million cells, medium control subtracted. Although all peptide pools (tat, nef, gag and env) were tested for groups A and C macaques as indicated, only positive responses are shown. Macaques with red fonts depicted unvaccinated challenged macaques. ND= not determined.

Table 3.

Antigen specific IFN-γ responses in SIV_MAC251 challenged macaques in lymphocytes isolated from lymph node^a

Groups	Monkey	Peptide Pools	Days post challenge
			14	28	35	84
A	AJ61	Nef	ND	ND	ND	ND
		Gag	252	500	ND	424
		gp120		650	ND	168
	AK46	Nef	ND	ND	ND	ND
		Gag	1027		ND	366
		gp120	100		ND
	DT55	Nef	ND	ND	ND	ND
		Gag	40	50	ND	110
	DT93	Nef	ND	ND	ND	ND
		Gag	105		ND	162
	T552	Nef	ND	ND	ND	ND
		Gag	220		ND
	V780	Nef	ND	ND	ND
		Gag	120		ND
C	FA14	Nef	405	ND	300	ND
		Gag	372	ND	767	ND
	L375	Nef	55	ND	117	ND
		Gag	602	ND	1265	ND
	P015	Nef		ND	160	ND
		Gag		ND	260	ND
	FG46	Gag		ND	175	ND

^aPeptide–specific responses are reported for lymphocytes isolated from lymph nodes in one million cells, medium control subtracted. Although all peptide pools (tat, nef, gag and env) were tested for groups A and C macaques as indicated, only positive responses are shown. Macaques with red fonts are depicted as unvaccinated challenged. ND = not determined.

DISCUSSION

Correlates of immunity or resistance to HIV infection remain unknown. This has hampered efforts to produce an effective vaccine. The aim of a vaccine would be to prevent infection, ameliorate disease course, and/or reduce virus transmission. The risk of transmission is greatest when viremia is high, especially during acute, and uncontrolled chronic infection. Infected individuals with plasma VLs of <1,700 copies/ml fail to transmit the virus to their HIV-negative partners (Gray et al., 2003; Gray et al., 2001). An effective HIV vaccine should at least be able to limit viral replication to low levels in which transmission is unlikely (Wilson et al., 2006). Although some vaccine candidates have shown modest reductions in VLs in macaques after SIV challenge, these were usually performed under ideal conditions using homologous viral challenges or against less than optimal SHIV’s (Casimiro et al., 2005; Koff et al., 2006; Letvin et al., 2006; Mattapallil et al., 2006; Watkins et al., 2008; Wilson et al., 2006). However, a more recent study has shown partially effective control of viremia by a live attenuated SIV followed by IV challenge with a heterologous SIV (Reynolds et al., 2008). Moreover, the studies here also demonstrate that significant levels of protection from an IV heterologous challenge with a highly pathogenic strain of SIV in macaques is indeed possible, at least in a proportion of macaques. Although the immunogen used in this study was a replication-competent SHIV known to cause AIDS in a proportion of macaques, the results demonstrate that the SHIV immunization prolonged the survival and/or delayed disease progression after challenge with SIV_MAC251. Furthermore, the fact that control of viremia persisted for months after levels of the immunizing virus had subsided, combined with the observation that mucosal viral target cells are at least partially restored prior to challenge, and more importantly, maintained after pathogenic challenge, suggests this control is associated with immune responses rather than a phenomenon of target cell depletion. Low to undetectable viral replication in intestinal tissues also suggested that maintenance/restoration of CD4⁺CCR5⁺ memory T cell targets was associated with control of the SIV challenge.

Unfortunately, we could not determine the correlates of immunity in these animals. Neither levels of plasma Nabs, nor virus specific systemic or mucosal IFN-γ responses correlated with protection. However, the fact that the envelope of the immunizing virus markedly differs from the challenge virus indicated that protection was not associated with immune responses directed against envelope. This may be an additional source of optimism, as it is unlikely that a vaccine can be engineered to generate responses to all of the diverse HIV envelope sequences encountered in endemic areas. Thus, vaccines might not have to be matched for all possible envelope sequences in order to confer protection.

One caveat that must be considered is whether this “protection” is indeed due to inducible immune responses or instead, viral competition for viral binding or integration sites. In this and prior studies, “protection” appears to be associated with persistence of low levels of virus or antigen in tissues. Conceivably, integrated virus could somehow prevent “superinfection” of target cells by either competition for integration sites, or by mediating cellular machinery to prevent subsequent infection or replication within infected cells. In fact, the lower levels of memory CD4⁺CCR5⁺ target cells in intestinal tissues of SHIV_162P3 immunized animals at the time of SIV challenge suggested that there may be persistent viral replication in the gut, even though our analysis of cell-associated virus did not demonstrate significant levels of viral RNA in these animals (Fig. 4A).

However, there is some evidence that this protection is mediated by adaptive immune responses. The fact that “immunized” animals maintained memory CD4⁺CCR5⁺ T cells in intestinal tissues following SIV challenge, in contrast to naïve controls who showed a significantly greater degree of intestinal CD4 depletion, suggests that the protection observed is not due to a limitation of target cells necessary to establish productive superinfection. Furthermore, all controls became infected, maintained high VLs, and progressed to AIDS, proving that this is not an artifact of macaques of Chinese origin, which have a higher incidence of long term nonprogression than macaques of Indian origin (Ling et al., 2007). In addition, this protection is not attributed to immune responses directed against foreign host major histocompatibility complex (MHC) proteins incorporated into the viruses, as the effects were the same whether the viral inoculum was grown in InRh macaque cells or passaged in ChRh. Finally, four immunized macaques did demonstrate transient slight (FA05, CF58) to moderate (DT93, R191) drops in mucosal CD4⁺ T cells immediately following SIV challenge, suggesting virus was reaching intestinal CD4⁺ T cells, and that there was an initial struggle in mucosal sites for control of local viral replication (Fig. 6). Unlike infected naïve controls, all four of these animals eventually restored mucosal CD4⁺ T cells to levels higher than at the time of challenge (Fig. 6). In summary, these studies provide evidence that protection against a highly pathogenic strain of SIV may be achieved, and suggest that this protection may be associated with adaptive mucosal immune responses (Casimiro et al., 2005; Kawada et al., 2007; Wilson et al., 2006).

These studies also argue against the hypothesis that superinfection results in a more severe disease progression (Altfeld et al., 2002; Gottlieb et al., 2007; Jost et al., 2002; Smith et al., 2004; van der Kuyl et al., 2005). The challenge virus used in these studies (SIV_MAC251) is among the most pathogenic of viruses in macaques, and if infection with one virus impairs immune responses against a second virus, we should have observed an enhancement of disease progression following SIV challenge. However, VLs were clearly diminished in the superinfected animals compared to controls, indicating that the primary infection confers some type of protection against superinfection and disease progression. In summary, this model may prove useful for understanding superinfection as well as correlates of immunity to HIV infection in humans. Although we are not promoting the use of attenuated viruses as vaccine candidates, we propose that such models may eventually help define correlates of immunity to infection, which could eventually guide HIV vaccine development.

MATERIALS AND METHODS

Macaques

Fifteen adult female Chinese rhesus (ChRh) and three Indian rhesus (InRh) (AK46, R191 and T552) macaques (Macaca mulatta) were housed at the Tulane National Primate Research Center in accordance with the regulations of the American Association for Accreditation of Laboratory Animal Care International. All animals were healthy and free from SIV, STLV and Type D retrovirus at the time the study was initiated.

Immunization

A series of three experiments was performed to determine whether macaques previously infected with SHIV_162P3 were susceptible to subsequent “superinfection” with pathogenic SIV (Fig. 1). SHIV_162P3 is a chimeric virus constructed from a SIV_MAC239 template, yet it contains functional HIV-1 tat, rev and env genes (Harouse et al., 2001; Harouse et al., 1999). Although IV inoculation of macaques with high doses of this virus occasionally results in persistent viremia and AIDS (Harouse et al., 2001; Harouse et al., 1999), in our experience, vaginal inoculation with SHIV_162P3 usually results in transient viremia and apparent “clearance” of virus within 70–90 days of infection (Pahar et al., 2007). To determine whether macaques previously infected with SHIV_162P3 were susceptible to IV “superinfection” with pathogenic SIV, a total of fourteen infected macaques (11 of ChRh and 3 InRh) and 4 naïve controls (all ChRh) were IV challenged with SIV_MAC251. For the IVAG primary infections (“immunization” phase) macaques used as placebo controls in microbicide experiments were first treated with Depo Provera (depomedroxyprogesterone acetate) and 30–33 days later, sedated and intravaginally inoculated with 300 TCID₅₀ SHIV_162P3 obtained from the NIH AIDS Research & Reagent Program (Pahar et al., 2007; Veazey et al., 2003). All 14 animals became infected as evidenced by high peak viremia 14–21 days after inoculation (Fig. 2). Macaques were then followed for 89–345 days post inoculation, until plasma viral loads declined to low (<7,600 vRNA copies/ml) or undetectable levels. In three separate experiments, infected animals and controls were IV challenged with a high (group A) and low (groupB) dose of SIV_MAC251 and plasma from a SIV-infected macaque of ChRh origin (group C).

In the first experiment (group A), seven SHIV_162P3 infected (4 ChRh and 3 InRh) and one-naïve ChRh macaques were IV challenged with 100 TCID₅₀ SIV_MAC251. In subsequent experiments all macaques (immunized and controls) were of Chinese origin. In the second experiment (group B), 4 SHIV_162P3 infected and 1-naïve macaques were IV challenged with a lower dose (10 TCID₅₀) of SIV_MAC251. The SIV_MAC251 challenge stock used in both A and B was grown in Indian origin macaque PBMC, and thus to rule out the possibility that the results observed were an artifact of using virus grown in MHC-mismatched hosts, animals in the third experiment were challenged with plasma obtained in primary infection from a ChRh macaque infected with SIV_MAC251. In this experiment (group C), three SHIV infected and two naive macaques were IV challenged with 1ml of plasma containing 2 x 10⁶ vRNA copies/ml.

Quantitation of plasma and cell associated viral RNA

Plasma VL during prechallenge time points was quantified by a bDNA signal amplification assay, version 4.0, specific for SIV, from Bayer Inc., which has a threshold detection limit of 125 copies/ml of plasma (Pahar et al., 2007). After SIV_MAC251 challenge, plasma VL and cell-associated SIVgag RNA copies were determined by a sensitive real-time RT-PCR method as previously described (Cline et al., 2005) (limit of detection as used in the present study, 30 RNA copies/ml).

Isolation of lymphocytes from blood, intestine and lymph nodes

Lymphocytes from the peripheral blood, jejunum and axillary LN were isolated as described earlier (Veazey et al., 2000a; Veazey et al., 2000b; Wang et al., 2008). In brief, intestinal biopsies were obtained by endoscopy and digested with serial incubations with EDTA and collagenase with rapid agitation. To enrich for lymphocytes, cells were centrifuged over discontinuous Percoll (Sigma) density gradients, washed, and resuspended in complete RPMI media containing 5% FCS (RPMI-5). Intestinal lymphocytes were >90% viable as assessed by Trypan blue dye exclusion.

Lymphocytes were isolated from axillary LN by mincing with paired scalpel blades and pressing them through nylon cell mesh. Cells were washed twice and resuspended in RPMI-5.

T cell immunophenotyping

T cell immunophenotyping was performed using CD3-FITC/PerCP (SP34–2), CD4-APC (L200), CD8-PE/PerCP (RPA-T8), CCR5 PE (3A9) and CD45RA FITC (5A9), all from BD Biosciences. For surface staining, 100ul of EDTA whole blood was incubated with titrated mAb for 30 min on ice. Red blood cells were lysed with 1x FACS lysing buffer (BD Biosciences) using a whole blood lysis technique as previously described (Pahar, Lackner, and Veazey, 2006).

Intestinal and LN lymphocytes were adjusted to 10⁷ cells/ml and 100ul aliquots were incubated with appropriately diluted antibodies for 30 min on ice, washed, and fixed in 2% paraformaldehyde. Cells were protected from light at 4°C and acquisition was completed within 24 h of staining. Data were acquired on a FACS Calibur flow cytometer (BD Biosciences, SanJose, CA) using Cell Quest software. At least 20,000 events were collected by gating on lymphocytes, and data were analyzed using FlowJo software (TreeStar Inc.) version 8.5.3.

ELISPOT assay

The number of antigen-specific IFN-γ spot forming cells (SFC) was measured with a commercial ELISPOT assay specific for rhesus IFN-γ (U-Cytech, Utrecht, The Netherlands), using fresh heparinized PBMC or lymphocytes isolated from LNs or intestine as previously described (Pahar et al., 2003; Pahar et al., 2007; Van Rompay et al., 2004). Peptides (15-mers with 11 amino acids overlap) derived from the HIV-tat (cat. 5138), SIV-tat (cat. 6207), SIV-nef (cat. 8762), SIV-gag (cat. 6204), SIV-env (cat. 6883) and SHIV_162P3 env (cat. 7619) were obtained from the NIH AIDS Research & Reference Reagent Program. Peptide pools were prepared so that each protein was equally represented. One pool was made for HIV-tat (peptide 1–23), SIV-tat (peptide 1–30) and SIV-nef (peptide 1–64) peptides and three peptide pools (gagA: peptide 1–33; gagB: peptide 34–89; gagC: peptide 90–125) were made to encompass the entire gag region of SIV. Similarly four (envA: peptide 1-60; envB; peptide 61–124; envC: peptide 125–168; envD: peptide 169–211) and three (envA: peptide 1–60; envB; peptide 61–124; envC: peptide 125–218) peptide pools were prepared to encompass the entire env region of SHIV_162P3 and SIV-env regions, respectively. For all positive control wells PMA (50ng/ml) and Ionomycin (1ug/ml) were used. Negative controls had no antigen/mitogen stimulation. Pooled peptides were added to cells at a final concentration of 1ug/ml for each individual peptide.

The numbers of spot forming cells were counted using the KS Elispot system (Zellnet Consulting Inc.). The mean number of spots per duplicate or triplicate well was calculated and the results were considered positive if the number of SFC for 2 x 10⁵ cells was ≥ 10 per well and greater than two-fold that of the negative control (medium only) wells plus 2 standard deviations.

Virus neutralization assay

Neutralization of SHIV_162P3 and primary SIV_MAC251 was measured as a function of reduction in luciferase reporter gene expression after a single round of infection in TZM-bl cells as described (Montefiori, 2004). TZM-bl cells were obtained from the NIH AIDS Research and Reference Reagent Program. Briefly, 200 TCID₅₀ of virus was incubated with serial three-fold dilutions of serum in triplicate in a total volume of 150ul for 1 h at 37°C in 96-well flat-bottom cell culture plates (Costar). Freshly trypsinized cells (10,000 cells in 100ul of growth medium containing 75ug/ml DEAE dextran) were added to each well. Positive and negative controls were kept with this assay. After incubation, 100ul of cells was transferred to 96-well black solid plates (Costar) for measurements of luminescence using Bright Glo substrate solution as described by the supplier (Promega). Neutralization titers were defined as the serum dilution at which relative luminescence units (RLU) were reduced 50% compared to virus control wells after subtraction of background RLUs. Assay stocks of primary SIV_MAC251 and SHIV_162P3 were prepared in human PBMC.

Neutralizing antibodies against TCLA SIV_MAC251 were assessed in CEMx174 cells as described previously (Montefiori et al., 1996). Briefly 50 μl of cell-free virus (5,000 TCID₅₀) was added to multiple dilutions of test serum in 100 μl of growth medium in triplicate wells of 96-well microtiter plates and incubated for 1 hr at 37°C. Cells (7.5 x 10⁴) in 100 μl of growth medium were added and incubated until extensive syncytium formation and nearly complete cell-killing were evident microscopically in virus control wells. Cell densities were reduced and medium replaced after 3 days incubation in cases where it took longer than 3 days to reach the assay end-point. Viable cells were stained with Finter’s neutral red in poly-L-lysine coated plates. Percent protection from virus-induced cell-killing was determined by calculating the difference in absorption (A₅₄₀) between test wells (cells + serum sample + virus) and virus control wells (cells + virus), dividing this result by the difference in absorption between cell control wells (cells only) and virus control wells and multiplying by 100. Nab titers are expressed as the reciprocal of the serum dilution required to protect 50% of cells from virus-induced killing. This cut-off corresponds to an approximate 90% reduction in p24 antigen synthesis. An assay stock of TCLA-SIV_MAC251 was prepared in H9 cells.

Statistical analysis

Graphical presentation and statistical analysis of the results were performed by one-way analysis of variance or the Student t test using the GraphPad Prism 4.0 (GraphPad Software Inc., SanDiego, CA). Plasma VLs comparison between immunized and control challenged macaque was determined by nonparametric Mann-Whitney t test. Correlation between plasma VL and antigen specific cytokine responses was determined using Spearman rank correlation test using Prism software. For all statistical analysis, results were considered significant if p<0.05.