|Home | About | Journals | Submit | Contact Us | Français|
A vaccine to protect human immunodeficiency virus (HIV)-exposed infants is an important goal in the global fight against the HIV pandemic. Two major challenges in pediatric HIV vaccine design are the competence of the neonatal/infant immune system in comparison to the adult immune system and the frequent exposure to HIV via breast-feeding. Based on the hypothesis that an effective vaccine needs to elicit antiviral immune responses directly at the site of virus entry, the pattern of virus dissemination in relation to host immune responses was determined in mucosal and lymphoid tissues of infant macaques at 1 week after multiple oral exposures to simian immunodeficiency virus (SIV). The results show that SIV disseminates systemically by 1 week. Infant macaques can respond rapidly to virus challenge and mount strong innate immune responses. However, despite systemic infection, these responses are most pronounced in tissues close to the viral entry site, with the tonsil being the primary site of virus replication and induction of immune responses. Thus, distinct anatomic compartments are characterized by unique cytokine gene expression patterns. Importantly, the early response at mucosal entry sites is dominated by the induction of proinflammatory cytokines, while cytokines with direct antiviral activity, alpha/beta interferons, are only minimally induced. In contrast, both antiviral and proinflammatory cytokines are induced in lymphoid tissues. Thus, although infant macaques can respond quickly to oral viral challenge, the locally elicited immune responses at mucosal entry sites are likely to favor immune activation and thereby virus replication and are insufficient to limit virus replication and dissemination.
The latest Joint United Nations Programme on HIV/AIDS statistics show no significant reduction in the number of new human immunodeficiency virus type 1 (HIV-1) infections per year (www.unaids.org). The number of HIV-1-infected women of childbearing age has been on the rise in recent years; thus, the potential for HIV-1 transmission to infants is increasing. In fact, mother-to-child transmission of HIV-1 accounts for about 700,000 new HIV-1 infections in children each year (12, 39), and one-third to one-half of those transmissions occur orally via breast-feeding.
Pediatric HIV infections are often characterized by persistent high levels of viremia and a more rapid disease course than in HIV-1-infected adults (35, 41). Assuming that the viral set-point is determined early in infection (15, 42), a better understanding of the mechanisms of oral HIV-1 transmission and early virus-host interactions at mucosal entry sites is needed. The infant rhesus macaque model of simian immunodeficiency virus (SIV) infection is extremely valuable in addressing this question and testing intervention strategies (9, 14, 16, 24, 37, 38, 43, 44). We have recently developed a physiologically more-relevant SIV infection model in infant macaques that consists of repeated oral SIV exposures by feeding multiple times daily for a fixed time period to mimic HIV-1 infection via breast-feeding in humans (36, 40, 43, 45).
Based on the hypothesis that an effective pediatric HIV vaccine must act very rapidly and directly at the site of virus exposure, the goal of the current study was to determine the pattern of virus dissemination in relation to the early host immune responses in tissues of the oral mucosa and draining lymph nodes, in gut-associated tissues, and in distal systemic lymphoid tissues in infant macaques at 1 week after multiple-low-dose oral exposures to SIV. This time point was chosen because it has been shown previously that systemic infection is established at 1 week in the majority of infant macaques (28, 43, 46).
The results of the study show that SIV rapidly disseminates after multiple oral exposures in infant macaques and that systemic infection is established within the first week. Despite systemic infection at 1 week after the first virus exposure, immune responses are most pronounced in tissues close to the site of virus entry, suggesting a lag period between systemic virus replication and the induction of systemic antiviral host immune responses. Further, the early immune response in both mucosal and lymphoid tissues is characterized by the induction of a strong inflammatory response. In contrast, mucosal tissues seem to be less prone to the induction of effective antiviral type I interferon responses than are lymphoid tissues. Thus, although infant macaques can respond quickly to oral virus challenge, the locally elicited immune responses at the site of virus entry are insufficient to control early virus replication and to prevent virus spread; instead, they are likely to promote virus replication. It is important to note that the rapid inflammatory response at mucosal sites of virus exposure and lagging systemic antiviral immunity are also characteristic of adult immune responses to genital mucosal SIV exposure (3, 34). Thus, the overall pattern of infant immune responses to oral SIV exposure is not due solely to immunologic immaturity.
Newborn rhesus macaques (Macaca mulatta) were housed and hand reared in a primate nursery in accordance with the regulations of the American Association for Accreditation of Laboratory Animal Care Standards at the California National Primate Research Center (CNPRC). All animals were born to female rhesus macaques of the CNPRC colony that is negative for HIV-2, SIV, type D retrovirus, and simian T-cell lymphotropic virus type 1.
Twelve infant macaques were orally inoculated with SIVmac251 by a previously described repeated-exposure model (45). Briefly, at 4 weeks of age, infant macaques were handheld and bottle-fed SIVmac251, diluted in a 1:1 mixture of RPMI 1640 medium and isotonic sucrose, for a total of 15 times (3 times per day for 5 consecutive days). The sucrose in this mixture did not affect virus infectivity titers (45). It is important to emphasize that the animals were awake and that no anesthesia was used. This allowed the animals to actively drink the virus inoculum, resulting in more-rapid exposure of the oral cavity and the esophagus. In addition, compared to exposure under anesthesia, the virus inoculum did not have contact for as long in the oral mucosa alone.
The uncloned SIVmac251 stock (internal reference number 2/02) was propagated on rhesus peripheral blood mononuclear cells (PBMC) in interleukin-2 (IL-2)-containing culture medium and had a titer of 105 50% tissue culture infective doses (TCID50) and 0.86 ×109 SIV RNA copies per ml (determined by branched-DNA [bDNA] assay) (P. J. Dailey, M. Zamroud, R. Kelso, J. Kolberg, and M. Urdea, Abstr. 13th Annu. Symp. Nonhum. Primate Models AIDS, abstr. 99, 1995; C. Wingfield, J. Booth, P. Sheridan, J. Detmer, and J. Turczyn, Abstr. 20th Annu. Symp. Nonhum. Primate Models AIDS, abstr. 135, 2002). Each infant inoculation consisted of 2 ml of a 1:20 dilution of the SIVmac251 stock (about 104 TCID50 or 8.6 ×107 viral RNA [vRNA] copies per dose). Each dose of this repeated-dosing regimen is lower than the two higher doses given in the high-dose inoculation model (44). Although this dose is still relatively high, the main goal was to apply a physiologically relevant SIV infection model in infant macaques that consists of repeated oral SIV exposures by feeding multiple times daily for a fixed time period to mimic HIV-1 infection via breast-feeding in humans. HIV exposure through breast-feeding is characterized by frequent repeated daily exposures that start early after birth and are likely to continue for several months to years. Thus, despite relatively low vRNA copy numbers per milliliter of breast milk (5, 43), the absolute amount of expected virus exposure per day is rather high. In fact, it has been estimated that an infant could be exposed to more than a million copies per day (40). The actual dose(s) of HIV in breast milk leading to infection is unknown and for ethical reasons cannot be rigorously defined. There is consensus that higher levels of HIV in breast milk are associated with a higher risk of HIV transmission. We are modeling this “high-risk” situation in humans, not the entire range of HIV transmission efficiency.
A total of eight age-matched SIV-uninfected macaques were included as control animals in the study. Control animals were inoculated with IL-2-containing tissue culture medium from SIV-uninfected rhesus PBMC cultures by the same regimen. This is an essential control in studies assessing innate immune responses, and prior studies of oral SIV exposure have not included this control. All animals were euthanized 4 days after the last SIVmac251 exposure, which corresponded to 8 days after the first virus exposure. The time of euthanasia is subsequently referred to as 1 week postinoculation (p.i.).
At the time of euthanasia, blood, gingiva, tonsil, submandibular lymph nodes (LN), esophagus, jejunum, colon, mesenteric LN, axillary LN, and spleen samples were collected. Tissue samples were stored in RNAlater (Ambion, Austin, TX) at −20°C until RNA preparation.
In addition, cell suspensions were prepared from the mesenteric LN, spleen, and blood. Tissue cell suspensions from LN were prepared by gently dissecting LN with scalpels in RPMI 1640 (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (Gemini BioProducts, Calabasas, CA) (complete RPMI) and passing the cell homogenate through a cell strainer (Fisher, Pittsburgh, PA). The cells were washed twice by centrifugation for 10 min at 1,600 rpm. Spleen tissue samples were cut into small pieces and homogenized with a syringe plunger. The homogenate was passed through a cell strainer. Mononuclear cell suspensions from the spleen and PBMC were isolated by gradient centrifugation with lymphocyte separation medium from MP Biomedicals (Aurora, OH), followed by two washes with RPMI 1640. Cell suspensions were stored in liquid nitrogen in 10% dimethyl sulfoxide-90% bovine calf serum (Gemini BioProducts) and/or used immediately for functional assays.
Prior to RNA isolation, the tissue samples were homogenized with a Power Homogenizer (PowerGen [7 mm by 195 mm]; Fisher Scientific). Total RNA was isolated with TRIzol (Invitrogen, Grand Island, NY) according to the manufacturer's protocol. The purified RNA was used to determine tissue vRNA levels and for gene expression analysis. RNA samples were DNase treated with DNA-free (Ambion) for 1 h at 37°C. cDNA was prepared with random hexamer primers (Amersham-Pharmacia Biotech, Inc., Piscataway, NJ) and MMLV reverse transcriptase (Invitrogen).
Real-time PCR was performed as previously described (1-3). Briefly, samples were tested in duplicate, and the PCRs for the housekeeping gene GAPDH and the target gene from each sample were run in parallel on the same plate. The reaction was carried out on a 96-well optical plate (Applied Biosystems, Foster City, CA) in a 25-μl reaction volume containing 5 μl cDNA plus 20 μl Mastermix (Applied Biosystems). All sequences were amplified with the SDS 7900 default amplification program: 2 min at 50°C and 10 min at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C. Results were analyzed with the SDS 7900 system software, version 2.1 (Applied Biosystems).
Relative cytokine mRNA expression levels were calculated from normalized ΔCT (cycle threshold) values and are reported as the increase in cytokine mRNA levels in tissues of SIV-infected monkeys compared to average cytokine mRNA levels in the same tissues of uninfected (age-matched) animals. CT values correspond to the cycle number at which the fluorescence due to enrichment of the PCR product reaches significant levels above the background fluorescence (threshold). In this analysis, the CT value for the housekeeping gene (GAPDH) is subtracted from the CT value of the target (cytokine) gene. The ΔCT value for the tissue sample from the uninfected animals is then subtracted from the ΔCT value of the corresponding tissue sample from the SIV-infected monkey (ΔΔCT). Assuming that the target gene (cytokine) and the reference gene (GAPDH) are amplified with the same efficiency (data not shown), the increase in cytokine mRNA levels in tissue samples of SIV-infected monkeys compared to tissue samples of uninfected animals is then calculated as follows: increase = (user bulletin 2, ABI Prism 7700 Sequence Detection System; Applied Biosystems).
Plasma and tissue RNA samples were analyzed for vRNA by a quantitative bDNA assay (Dailey et al., Abstr. 13th Annu. Symp. Nonhum. Primate Models AIDS; Wingfield et al., Abstr. 20th Annu. Symp. Nonhum. Primate Models AIDS). Virus load in plasma samples is reported as log10 vRNA copy numbers per ml of plasma. The detection limit of this assay is 125 vRNA copies. Virus load in tissue samples is reported as log10 vRNA copy numbers per μg of total tissue RNA (2, 3, 27).
The percentages of CD3+ CD4+ and CD3+ CD8+ T cells, CD3− CD8+ NK cells, and CD3− CD20+ B cells within the lymphocyte population were determined by fluorescence-activated cell sorter analysis on a FACSCalibur instrument using rhesus macaque-specific antibodies from Pharmingen (Pharmingen/Becton-Dickinson, San Jose, CA; CD3 clone no. SP34, CD4 clone no. M-T477, and CD8 clone no. SK1) and Becton-Dickinson (Becton-Dickinson, San Jose, CA; CD20 clone no. L27). Expression of CCR5 on CD3+ CD4+ T cells was determined with CCR antibody clone no. 3A9 from Becton-Dickinson.
Intracellular cytokine staining for gamma interferon (IFN-γ) and tumor necrosis factor alpha (TNF-α) was performed on frozen cell suspensions from mesenteric LN and spleen as previously described (30).
Results were analyzed by one-way analysis of variance or the Student t test using the GraphPad Prism and InStat software programs, version 4 (GraphPad Software, Inc., San Diego, CA). Correlations between virus replication and cytokine mRNA levels were determined by the linear regression and correlation analysis tools in the Prism software.
At 1 week after the first oral exposure to SIVmac251, vRNA was detectable in the plasma of all 12 animals; thus, all animals were systemically infected (Fig. (Fig.1A).1A). In 8 of the 12 monkeys, plasma vRNA levels exceeded 106 copies/ml; in 1 of the 12 monkeys, plasma vRNA levels reached about 105 copies/ml; and 3 of the 12 animals had between 103 and 104 vRNA copies/ml of plasma (Fig. (Fig.1A1A).
To determine the pattern of virus dissemination at 1 week after multiple oral SIV exposures, we measured the levels of virus replication in tissues at the site of virus administration (gingiva) and the draining lymphoid tissues (tonsil, submandibular LN); in tissues of the gastrointestinal tract (esophagus, jejunum, colon), including the mesenteric LN; and in distal lymphoid tissues (axillary LN, spleen). The results show that oral SIV inoculation in newborn macaques results in extremely rapid virus dissemination (Fig. (Fig.1).1). At 1 week after the first oral SIV exposure, virus replication was detectable in local and systemic tissues in 10 of 12 monkeys. In the four monkeys with lower plasma vRNA levels (<105copies/ml), virus replication was confined mainly to the site of virus entry (Fig. (Fig.1B).1B). In monkey 34642, replicating virus was detectable only in the tonsils and the submandibular LN. In the other three monkeys with plasma vRNA levels below 105 copies/ml, vRNA was detectable in tissues at the portal of entry but also in some distal lymphoid tissues, albeit at low levels (Fig. (Fig.1B).1B). In the eight animals with higher plasma viremia (>106 copies/ml), vRNA was detectable at multiple anatomic sites: at the site of virus exposure, in gut-associated tissues, and in distal lymphoid tissues. With the exception of the colon, virus replication was higher in lymphoid tissues than in mucosal tissues. In fact, despite oral exposure to SIV, at 1 week p.i., the gingiva was not a major site of virus replication. All tissues, except the axillary lymph node, had statistically significant higher vRNA levels than the gingiva (P < 0.05). Further, virus replication was significantly (P < 0.05) higher in the tonsil, mesenteric lymph node, and spleen than in the esophagus. Consistent with lower virus replication in the gingiva and the esophagus, there was a significant positive correlation between plasma vRNA and tissue vRNA levels (P < 0.01) for all tissues except the gingiva and the esophagus. This is likely the result of the various cell populations present in mucosal versus lymphoid tissues, especially the fewer CD4+ T cells present in mucosal tissues. Although average tissue vRNA levels of all animals were not statistically different between tonsil, submandibular LN, colon, mesenteric LN, axillary LN, and spleen, in 7 of 12 animals tissue vRNA levels for an individual animal were highest in the tonsil or the submandibular LN, and in 3 of 12 monkeys tissue vRNA levels were highest in gut-associated tissues (Fig. (Fig.1B1B).
A main goal of the present study was to determine the relationship between virus replication and host immune responses in tissues of infant macaques after oral SIV exposure. In particular, we analyzed gene expression levels of cytokines with antiviral activity (IFN-α/β) and inflammatory cytokines important in immune activation and effector cell recruitment. Further, we tested whether SIV-specific CD8+ T-cell responses were detectable in tissues at this early time point after SIV exposure.
In infant macaques, the CD4+ to CD8+ T-cell ratio in the blood (46) (Table (Table1)1) and in tissues (reference 47 and data not shown) is higher than typically observed in SIV-naïve adult macaques. Despite the increased frequencies of potential target cells for SIV in infant macaques, higher CD4+ T-cell frequencies in an individual animal at the time of challenge were not predictive of postchallenge levels of virus replication in the same animal (P > 0.05; r2 = 0.08 [see also Table Table1]).1]). However, absolute CD4+ T-cell numbers were decreased in some animals after SIV infection (Table (Table1),1), and the average decrease in peripheral blood CD4+ T-cell numbers from week 4 (time of SIV exposure) to week 5 (1 week after the first SIV exposure) was significantly greater in SIV-infected animals than in age-matched (week 5 of age) SIV-naïve animals in peripheral blood (Mann-Whitney U test: P < 0.03) (Table (Table11).
Consistent with the antigen-naïve phenotype of T cells in infant compared to adult macaques, the expression of activation markers such as CCR5, HLA-DR, and CD69 on T cells was low (<5% [data not shown]). Thus, the results of the present study suggest that in infant macaques, the absolute frequencies of CD4+ T cells and their activation status in peripheral blood at the time of SIV exposure are not predictive of acute-phase SIV replication levels.
IFN-α and IFN-β exert antiviral function through the induction of several IFN-inducible genes (31), through the activation of innate effector cells, and are essential for the induction of adaptive antiviral immune responses (4, 10, 17, 18, 29, 31).
Oral exposure of infant macaques to SIV resulted in a marked increase in IFN-β and IFN-α mRNA levels in the tonsil (Fig. (Fig.2A).2A). The magnitudes of IFN-β and IFN-α mRNA expression levels in the tonsil were directly correlated with tonsil vRNA levels (P < 0.003; r2 = 0.62). A strong increase in IFN-α mRNA levels was also observed in all the LN but was less pronounced in mucosal tissues (Fig. (Fig.2A).2A). Thus, IFN-α mRNA levels were significantly more elevated in tonsil, submandibular LN, and mesenteric and axillary LN than were IFN-α mRNA levels in the gingiva, esophagus, and jejunum. In the colon, despite similar vRNA levels compared to the tonsil, IFN-α mRNA levels were only minimally induced. Further, no increase in IFN-α mRNA was observed in the spleen in any of the animals.
Consistent with the induction of IFN-α in the tonsil and LN, these tissues also had increased expression levels of the IFN-inducible genes Mx and CXCL10 (Fig. 2B and C). The exception was the mesenteric LN, in which CXCL10 was only minimally induced, despite up to a 1,000-fold increase in IFN-α mRNA levels in some animals. Despite the lack of evidence of a strong IFN-α response in mucosal tissues, the IFN-inducible genes Mx and CXCL10 were markedly increased in the gingiva and the esophagus of animals with detectable virus replication in these tissues (Fig. 2B and C). It is possible that the IFN-α induction was not detected in these tissues if IFN-α induction had occurred earlier than 1 week p.i. and was transient. Alternatively, viral proteins could have directly induced these antiviral genes independently of IFN-α induction (6, 20). Consistent with low IFN-α mRNA levels in the jejunum and colon, IFN-α-inducible genes showed the lowest expression levels in the jejunum and colon.
In general, increases in tissue IFN-α/β, Mx, and CXCL10 mRNA levels were less pronounced in animals with lower plasma and tissue vRNA levels than in animals with higher vRNA levels (Fig. (Fig.2).2). Thus, strong type I IFN responses did not seem to be associated with control of virus replication and prevention of virus dissemination but instead represent a marker of relative virus replication.
The inflammatory response is essential for the recruitment of various effector cells to the site of virus replication and for the activation of antigen-presenting cells that then prime antigen-specific T cells. Thus, we sought to determine whether cytokines and chemokines important in inflammation are rapidly induced in infants after oral SIV exposure and whether their expression levels correlate with virus replication.
Oral exposure of infant macaques with pathogenic SIV resulted in the rapid induction of several proinflammatory cytokines (Fig. (Fig.3).3). Despite the clear establishment of a systemic infection at 1 week after the first SIV exposure, the highest cytokine mRNA levels were found in local tissues at the nearest sites of virus exposure. Macrophage inflammatory protein 1α (MIP-1α) mRNA levels were significantly higher (P < 0.05) in the gingiva and tonsil than in all other tissues (Fig. (Fig.3A).3A). Consistent with the highest increase in MIP-1 mRNA levels in gingiva and tonsil, average mRNA levels for CCR5, the receptor for MIP-1α, were highest in the tonsil, but this difference did not reach statistical significance except for the jejunum (data not shown). In the tonsil, MIP-1α mRNA levels were positively correlated with levels of virus replication (P < 0.002; r2 = 0.69). IL-12 mRNA levels were also markedly increased in the gingiva and tonsil compared to IL-12 mRNA levels in more distal tissues (Fig. (Fig.3B).3B). In fact, IL-12 mRNA levels were significantly higher in the gingiva and tonsil than in the jejunum, the mesenteric LN, and the spleen (P < 0.05). Consistent with the conclusion (see above) that oral exposure can result in virus entry through the esophagus/gut-associated tissues, a marked increase in IL-12 mRNA levels was also observed in the esophagus. As in the gingiva, IL-12 mRNA levels in the esophagus were significantly higher than in the jejunum, the submandibular LN, the mesenteric LN, and the spleen (P < 0.01). However, although tissue vRNA levels in the colon and mesenteric LN were comparable to tissue vRNA levels in the tonsil, average mRNA levels of proinflammatory cytokines were much lower in these tissues. In the jejunum, proinflammatory cytokine mRNA levels were indistinguishable from cytokine mRNA levels in the jejunum of normal, SIV-negative animals. The only exception was monkey 34654, and this animal had the highest vRNA levels in the jejunum of all the animals. While MIP-1α and IL-12 mRNA levels were increased in both the gingiva and the tonsil, markedly increased IL-6 mRNA levels were observed only in the tonsil (Fig. (Fig.3C).3C). In fact, IL-6 mRNA levels in the tonsil were significantly higher than gingiva IL-6 mRNA levels (P = 0.007).
Thus, four distinct tissue vRNA-cytokine patterns emerge: (i) tissues close to the site of virus exposure that have relatively low vRNA levels but a strong inflammatory response at 1 week p.i. (gingiva, esophagus); (ii) tissues with high virus replication and strong innate immune responses, with immune responses being stronger in tissues closer to the portal of virus entry (tonsil) than in tissues more distal to the site of virus exposure (axillary LN, colon); (iii) tissues with relatively high vRNA levels (mesenteric LN, spleen) at 1 week p.i. but low immune responses; and (iv) tissues with low vRNA levels and weak or absent immune responses (jejunum).
IFN-γ is a key cytokine in the activation of innate (cytotoxic NK cell) and adaptive (cytotoxic T cell) antiviral effector cells (8, 13, 23). At the same time, IFN-γ-driven inflammation has the potential to promote SIV/HIV replication (2, 33).
Similar to the induction of type I IFN and proinflammatory cytokines, the strongest increase in IFN-γ mRNA levels was observed in the tonsil (Fig. (Fig.4A).4A). There was a positive correlation between tonsil IFN-γ mRNA and vRNA levels (P < 0.008; r2 = 0.57). The tissues with the highest IFN-γ mRNA levels—tonsil, esophagus, and axillary LN—also had the highest mRNA levels for the IFN-γ-inducible chemokines CXCL9 and CXCL10 (Fig. (Fig.4B).4B). Consistent with the fact that CXCL10 can be induced by both IFN-γ and by IFN-α/β (see above), CXCL9 and CXCL10 mRNA levels in the submandibular LN were comparable to CXCL9 and CXCL10 mRNA levels in those three tissues (Fig. (Fig.2C2C and and4B).4B). In the jejunum, colon, and mesenteric LN, the increase in CXCL9 and CXCL10 mRNA levels was significantly lower than in the tonsil, the submandibular LN, the esophagus, and the axillary LN (P < 0.01).
To determine whether the early IFN-γ response could be associated with cellular anti-SIV immune responses, the mRNA levels of the cytotoxic effector molecules granzyme and perforin and the response to in vitro SIVgag p27 stimulation were measured. It was a striking finding that the mRNA levels for granzyme and perforin were most strongly increased in the same tissues that had the highest mRNA levels for IFN-γ (Fig. 4C and D). Thus, the tonsil and the esophagus had significantly higher mRNA levels of perforin and granzyme than the gingiva, the submandibular LN, the jejunum, the colon, and the spleen (P < 0.05). Similarly, the axillary LN also had significantly higher perforin and granzyme mRNA levels than the jejunum and the colon. However, the observed increase in IFN-γ, perforin, and granzyme mRNA levels could be the result of increased mRNA levels in CD8+ T cells, NK cells, or both. Thus, to test for SIV-specific CD8+ T-cell responses, the in vitro response to SIVgag p27 peptide stimulation was measured by intracellular cytokine detection of IFN-γ and TNF-α by flow cytometry. Due to the relatively low numbers of cells available from infant tissues, this assay was limited to spleen and mesenteric LN cell suspensions. Consistent with relatively low IFN-γ, perforin, and granzyme mRNA levels in these tissues, SIVgag p27-specific CD8+ T-cell responses were not detectable. Similarly, no IFN-γ- or TNF-α-producing cells could be detected in the CD4+ T-cell population in response to SIVgag p27 peptide stimulation. Further, there was no bias in cytokine induction that was indicative of either a Th1 or Th2 polarization. In fact, the tissues with the highest increase in IFN-γ and IL-12 mRNA levels also had significantly higher IL-4 mRNA levels than all other tissues (P < 0.05) (data not shown).
With the caveat that SIV-specific CD8+ T-cell responses were not analyzed in tissues at or near the virus entry site, there was no direct evidence of SIV-specific T-cell responses. Instead the changes in IFN-γ mRNA levels paralleled the increases in mRNA levels observed for other proinflammatory cytokines; thus, it is likely that the observed early IFN-γ response is part of this inflammatory response.
As virus replication was detectable in the plasma of all animals at 1 week after the first oral SIV exposure, we tested whether immune responses were elicited and detectable in PBMC at this early time point. Importantly, the question of how well the tissue responses were reflected by blood responses was addressed. The majority of the animals showed evidence of an early antiviral response as measured by the induction of type I IFN or the IFN-stimulated genes Mx and CXCL10. Although PBMC mRNA levels for IFN-α were only slightly increased in 2 of 8 animals (data not shown), increased PBMC mRNA levels for Mx and CXCL10 were detectable in 10 of 12 and 12 of 12 animals, respectively (Fig. (Fig.5A).5A). Further, in 9 of 12 animals, PBMC mRNA levels for MIP-1α were elevated, suggesting that a systemic inflammatory response was induced in response to viral challenge (Fig. (Fig.5A).5A). Indeed, the induction of both IFN-stimulated genes and MIP-1α in PBMC was positively correlated with plasma vRNA levels (Fig. (Fig.5A).5A). The fact that several innate and inflammatory immune responses were detected in the blood at 1 week after the first SIV exposure at a mucosal site is indicative of rapid virus dissemination and consistent with the observed plasma viremia in all animals. Importantly, it shows the ability of the immune system of infant macaques to respond quickly to oral SIV challenge.
In contrast, PBMC mRNA levels for IFN-γ, granzyme, and perforin, effector molecules of cytotoxic T cells and NK cells, were indistinguishable from PBMC mRNA levels of uninfected, age-matched control monkeys (data not shown). Although an innate antiviral response was readily detectable in PBMC of infant macaques at 1 week after oral SIV exposure, SIV-specific immune responses were rarely detected in PBMC. In only 1 of 12 animals could TNF-α-secreting CD8+ T cells be detected in response to in vitro SIVgag p27 peptide stimulation (data not shown). It should be noted that in infant macaques, the number of CD8+ T cells is often relatively low while the ratio of CD4+ to CD8+ T cells in peripheral blood is higher than in adult animals (Table (Table1)1) (44). Thus, assuming that only about 1% of all CD8+ T cells are antigen specific, the resulting low frequencies of SIV-specific CD8+ T cells might be close to the detection limit of the assay. Further, and in sharp contrast to tissue responses, no increase in PBMC IFN-γ mRNA levels was observed. Instead IL-4 PBMC mRNA levels were elevated in 8 of 12 animals, suggesting a bias toward Th2-like responses (Fig. (Fig.5B).5B). Thus, at this early time point after oral SIV exposure, PBMC responses do not consistently reflect the full spectrum of responses detectable in tissues.
The development of an HIV vaccine remains a major goal in the control of the HIV pandemic. Pediatric HIV vaccine design faces additional obstacles, including developmental differences between the neonatal/infant and the adult immune systems and the relatively high frequency of HIV exposure via breast-feeding starting immediately after birth and continuing for several weeks to years. Thus, an effective HIV vaccine likely has to induce antiviral immune responses immediately after birth and directly in the local tissues at the site of virus exposure. Therefore, the present study assessed the relationship between virus replication and early host immune responses in a model designed to reflect the natural route of HIV-1 transmission through breast-feeding in children using infant macaques orally exposed to multiple doses of pathogenic SIV.
Systemic infection was established in all animals at 1 week after the first virus exposure, a result consistent with our earlier studies in which >85% of the animals became infected by the same multiple dosing regimen (43). The design of the present study did not allow us to determine exactly which SIV exposure resulted in infection. In fact, the various levels of virus replication observed in different animals by 1 week p.i. suggest that individual animals become infected after different numbers of exposures. Importantly, though, the relatively high percentage of viremic animals at this early time point after repeated exposure to SIV is remarkably similar to the observed infection rate in infant macaques after two high-dose oral inoculations with pathogenic SIV (28, 44). The results are also similar to findings observed in the repeated SIV exposure model of vaginal transmission in adult macaques (22) in which, once infection occurs, the virus dissemination and challenge outcome is similar after exposure to two high or multiple lower doses of SIV. However, as there are differences in pathogenesis between HIV-1-infected infants and adults, differences in age and route need to be considered in the evaluation of how early virus-host interactions can affect the long-term challenge outcome. While our colleagues and we have recently shown that virus rapidly disseminates in infant and juvenile macaques upon high-dose oral SIV exposure (28), only three infant macaques were included in the previous study. Further, in most tissues, the study was limited to the analysis of viral DNA. This is the first study providing direct evidence that SIV actively replicates (vRNA) within multiple tissues after repeated oral SIV exposure in infant macaques. Further, the results of the present study extend the observation of the previous high-dose SIV inoculation study to the repeated oral SIV exposure model as a model of breast milk transmission in human infants. Importantly, we can show that infant macaques have the ability to rapidly induce innate immune responses to oral SIV exposure. However, SIV can rapidly disseminate despite these locally elicited immune responses.
The primary target cell after oral transmission of HIV/SIV has not been identified. In general, CCR5-expressing CD4+ T cells are the preferred target cells for HIV and SIV in acute infection (19, 25, 48). The highest concentration of CCR5+ CD4+ T cells is found in the mucosa-associated lymphoid tissue of the gut and not in peripheral blood or lymphoid tissues (49). Milush et al. (28) reported that CD4+ T cells and CCR5+ cells are present in the oral mucosa of infant and juvenile monkeys and that SIV-infected T cells and macrophages were detectable at 4 to 7 days after oral inoculation in the epithelium of the esophagus (28). Consistent with our finding that the tonsil is a primary site of virus replication, there is also an abundance of CCR5-expressing CD4+ T cells in the tonsil (7, 26). Further, the results of the present study show that as early as 1 week after the first SIV exposure, the absolute number of CD4+ T cells in peripheral blood was significantly lower in SIV-infected than in age-matched (5 weeks of age) SIV-naïve animals, suggesting that CD4+ T cells are an early target of SIV infection after oral exposure. This rapid systemic effect of apparent CD4+ T-cell depletion in infant macaques might be a result of the relatively higher frequencies of CD4+ T cells in infants compared to adults, frequencies that gradually decrease in the first few weeks to months of life (44, 46). However, only SIV-infected animals had a decline in CD4+ T cell levels; thus, this decline is not simply an age-dependent effect. Thus, while there was no statistically significant correlation between the frequency and the activation status of CD4+ T cells in the blood prior to SIV exposure and vRNA levels of individual animals in the present study, it remains to be determined whether there is a relationship between the maturation and activation status of CD4+ T cells and levels of virus replication at mucosal entry sites for SIV in infant macaques.
An important question that remains to be answered is what viral and/or host factors contribute to the persistently high levels of viremia in infants and their more rapid disease progression compared to adults. Thus, a major goal of the present study was to determine whether infant macaques have the ability to respond rapidly to oral SIV challenge. The immune responses elicited after oral SIV exposure differed in kind and magnitude at distinct anatomic sites. In fact, four unique patterns emerged: (i) tissues close to the site of virus exposure that have relatively low vRNA levels and a strong inflammatory response at 1 week p.i. but that lack antiviral type I IFN responses (gingiva, esophagus); (ii) tissues with high virus replication and strong innate immune responses that are more pronounced in tissues closer to the portal of virus entry (tonsil) than in more distal tissues (axillary LN, colon); (iii) tissues with relatively high vRNA levels (mesenteric LN, spleen) at 1 week p.i. but lesser induction of immune responses; and (iv) tissues with low vRNA levels and weak or absent immune responses (jejunum). These unique response patterns observed in distinct tissues raise questions about what cell populations are present in different anatomic locations and/or whether cytokine responses are differentially regulated in different tissues. Further, we need to understand how these different immune responses at distinct anatomic sites can influence SIV pathogenesis and the establishment of virus reservoirs in the first few days of infection. The tissue-specific immune response patterns were similar to findings obtained with the vaginal transmission model of SIV in adults (3).
Importantly, although a marked increase in type I IFN was detectable at 1 week after the first SIV exposure, this activity was insufficient to prevent virus spread. Further, type I IFN mRNA levels were generally higher in lymphoid than in mucosal tissues. Thus, the data suggest that effector cell populations that respond to in vivo virus challenge with IFN-α production may not be present in sufficient numbers at these mucosal entry sites to effectively limit or prevent virus dissemination. In contrast to type I IFN, several proinflammatory cytokine and chemokine levels were increased in mucosal and lymphoid tissues. As inflammatory cytokines not only recruit and activate antiviral effector cells but also activate T cells, a target cell population of HIV/SIV, they have the potential to directly enhance virus replication. In fact, the induction of proinflammatory cytokines and the relative lack of antiviral type I IFN responses in the gingiva, esophagus, and colon result in a cytokine environment that is likely to promote immune activation and virus replication rather than controlling viral replication and preventing virus dissemination. Thus, prevention of the detrimental inflammatory response presents itself as a common goal in pediatric and adult HIV vaccine design.
Overall, despite the establishment of systemic infection by 1 week p.i., the most pronounced increase in cytokine mRNA levels was observed in tissues closest to the site of virus exposure, suggesting that even the innate host immune response was lagging behind virus replication. Thus, the highest cytokine mRNA levels were observed in the oral mucosa and the draining lymph nodes, suggesting that induction of cytokines is directly associated with the level and duration of virus replication and with the kinetics and pattern of virus dissemination. This result is similar to the findings in the high-dose vaginal transmission model (3); thus, the overall pattern of the infant immune response is not solely the result of immunological immaturity.
In contrast, anti-SIV/HIV CD8+ T-cell responses may be age dependent. The analysis of CD8+ T-cell responses was necessarily limited to PBMC, spleen, and mesenteric LN. It may be more difficult to detect SIV-specific CD8+ T-cell responses in the first few weeks of life, because antigen-specific T cells occur in low frequencies and infants have relatively few CD8+ T cells compared to adults. Indicative of the general ability of infant macaques to mount SIV-specific immune responses, low frequencies of SIV-specific CD8+ T cells could be detected in PBMC of 1 of the 12 infant macaques, though not in any of the tissues examined. A recent longitudinal study performed in HIV-1-infected African children showed that HIV-specific CD8+ T-cell responses in PBMC are rarely detectable in the first few weeks of life (21). Thus, the results for SIV-infected infant macaques are consistent with HIV-1-infected children. In addition, and consistent with our own unpublished observation, in vitro studies with human neonatal PBMC suggest that the activation threshold for infant T cells is higher than in adults (11, 32). Importantly, though, as the results of the present study clearly show that even innate immune responses to oral SIV challenge are generally strongest in the tonsil and the submandibular LN, future studies need to examine CD8+ T-cell responses at anatomic sites at or near the portal of virus entry and at later time points. In fact, the strongest increase in mRNA levels for the cytotoxic effector molecules perforin and granzyme was observed in the tonsil. Alternatively, the induction of virus-specific CD8+ T-cell responses could be delayed after mucosal, compared to systemic, HIV/SIV exposure. Although a direct comparison of virus-host interactions after vaginal heterosexual transmission in adults and oral HIV/SIV exposure in infants should be interpreted with caution, it is interesting to note that in vaginally exposed adult macaques, CD8+ T-cell responses are delayed until day 14 to 21 and even then are strongest in tissues closest to the site of virus entry (34). Thus, an important unanswered question is whether developmental differences in the infant compared to the adult immune system, the anatomic site of virus exposure, or both are the determining factor(s) for the observed absent or lower levels of HIV/SIV-specific CD8+ T-cell responses in infants early after infection.
In conclusion, several factors are likely to contribute to the observed rapid virus dissemination in infant macaques at 1 week after oral SIV exposure. These include high frequencies of CD4+ T cells, a major target cell population for SIV/HIV, and unique cytokine gene expression patterns in distinct anatomic compartments, with responses generally being stronger in lymphoid tissues than at mucosal entry sites. Further, the early inflammatory cytokine response in the absence of SIV-specific adaptive immune responses is likely to promote virus replication. Importantly, though, the results of the present study show that infant macaques can rapidly respond to in vivo virus challenge and mount strong innate immune responses in local and systemic tissues. While the induced responses are insufficient to control virus replication and stop virus dissemination, a better understanding of the kinetics of their induction in relation to virus replication in distinct anatomic compartments may lead the way to therapeutic intervention aimed at prevention of HIV transmission.
This work was made possible through the support of NIH/NIDCR grant R21 DE016541 to K.A. and NIH/NIAID grants AI046320 and AI062518 to M.M. The animal work was in part supported by grant RR00169 to the CNPRC.