Increased IL-15 Production Is Associated with Higher Susceptibility of Memory CD4 T Cells to Simian Immunodeficiency Virus during Acute Infection¹²

Acute SIV infection is characterized by a massive loss of memory CD4 T cells from both mucosal and peripheral tissues. These cells are primarily lost due to viral infection that peaks at day 10 postinfection (p.i.).⁵ Interestingly, the kinetics of infection appears to be independent of the route of infection; both i.v. and vaginal challenge display similar kinetics of infection and loss (1, 2).

A detailed analysis of the kinetics of viral infection shows that very few memory CD4 T cells in the mucosa or periphery are infected at day 7 p.i., whereas most of the memory CD4 T cells were infected at day 10 p.i. Li et al. (1) using SIV-specific riboprobes demonstrated that the level of infection in the vaginal mucosa significantly increased between days 7 and 10 p.i. Additionally, Mattapallil et al. (2) using a quantitative PCR assay for SIV-gag DNA showed that SIV-gag copies in either the jejunum or periphery increased from ∼1 to 10 × 10³ copies of SIV-gag/10⁵ memory CD4 T cells at day 7 p.i. to ∼2 to 2.5 × 10⁵ copies of SIV-gag/10⁵ memory CD4 T cells at day 10 p.i. Single-cell analysis revealed that there were approximately two copies of SIV DNA/cell (2). This suggests that <5% of memory CD4 T cells were infected at day 7 p.i., whereas by day 10 p.i. most of the memory CD4 T cells carried viral DNA.

What drives this massive amplification of viral infection within a short period of 2–3 days? It is generally believed that immune activation associated with acute infection likely plays a major role. Little is known, however, about the exact mechanisms behind this process. We have previously shown that most memory CD4 T cells express similar levels of CCR5 at days 7 and 10 p.i. (2), suggesting that factors other than SIV coreceptor expression play a role in making the memory CD4 T cells highly permissive to infection during this period.

Previous studies have shown that acute immune activation was associated with increased proliferation and activation of peripheral CD8 T cells (4, 5). The preferential activation and proliferation of memory CD8 T cells during acute SIV infection suggests that specific cytokines likely activate memory CD8 T cells. Cytokines such as IL-7 have been shown to induce homeostatic proliferation of both naive and memory CD8 T cells, whereas IL-15 primarily supports homeostatic and infection-mediated proliferation and survival of memory CD8 T cells under physiological conditions (6, 7, 8, 9, 10, 11). Recent studies (11) have shown that IL-15 has an effect on the proliferation of effector memory CD4 T cells in rhesus macaques. We hypothesized that immune activation during acute SIV infection is likely associated with production of IL-15 that makes memory CD4 T cells more susceptible to infection. We tested this hypothesis using experimental SIV infection of rhesus macaques by evaluating the kinetics of Ki-67 and HLA-DR expression on T cells and measuring plasma IL-15 levels and SIV infection in memory CD4 T cells. Additionally, we evaluated the expression of IL-15Rα and IL-2Rβ on T cells to better delineate the role of IL-15 in acute immune activation and amplification of viral infection. Finally, we performed in vitro culture experiments to directly assess the ability of IL-15 to increase SIV infection in CD4 T cells.

Eight colony-bred healthy Mamu*A01neg rhesus macaques (Macaca mulatta) of Indian origin housed at Advanced Bioscience Laboratories (Kensington, MD) were used in the longitudinal study. Animals were housed in accordance with the American Association for Accreditation of Laboratory Animal Care guidelines and were seronegative for SIV, simian retrovirus, and simian T cell leukemia virus type 1. All animal care and procedures were reviewed and approved by the Institutional Animal Care and Use Committee. Animals were infected with 100 animal infectious doses of uncloned pathogenic SIVmac251 i.v.; peripheral blood, rectal biopsies, and plasma samples were collected longitudinally at various time points before and after challenge. Additionally, archival blood and mucosal samples were obtained from SIVmac251-infected animals for analysis.

PBMC were isolated by density gradient centrifugation, whereas rectal biopsies were processed using Percoll gradients as previously described (2). Plasma viral loads were determined by real-time RT-PCR assay as previously described (12).

All Abs except for CD4-Qdot605 (courtesy of M. Roederer, National Institutes of Health, Bethesda, MD) used in this study were obtained from BD Biosciences. All reagents were validated and titrated using rhesus macaque PBMC. For phenotypic analysis and sorting of CD4 T cell subsets, freshly isolated cells were labeled simultaneously with the following combinations of Abs: CD3-Cy7-allophycocyanin, CD8-Alexa Fluor 700, CD4-PE or CD4-Qdot605, HLA-DR-Texas Red-PE, CD95-allophycocyanin, and CD28-Cy5-PE. T cell activation was determined by labeling freshly isolated cells using the above panel. After fixing and permeabilizing, the cells were stained with Ki-67-FITC. To determine IL-15Rα and IL-2Rβ expression, cells were labeled with anti-IL-15Rα (R&D Systems) and IL-2Rβ (R&D Systems) along with anti-CD3, CD4, CD8, CCR5, and CD95. Labeled cells were fixed with 0.5% paraformaldehyde and analyzed using a modified BD Biosciences Aria Sorter. One to two million total events were collected for analysis.

To determine which subsets supported viral infection, naive and memory CD4 T cells (discriminated based on CD28 and CD95 (13) expression) were sorted into tubes and subjected to qPCR assay for measuring SIV-gag DNA as described previously (2, 14).

For in vitro culture experiments, cells were labeled with anti-CD14-FITC, CD20-FITC, and CD8-FITC and resuspended in warm culture medium. Cells negative for these markers representing CD4 T cells were sorted into warm medium. After harvesting, cells were labeled with anti-CD3-Cy7-allophycocyanin, CD4-PE, CD8-Cy5-PE, CD95-allophycocyanin, and CD28-FITC.

CD4 T cell-associated viral DNA was measured by a quantitative PCR assay for SIV gag as previously described (2, 14) using SIV gag primers and probe as described by Lifson et al. (15). The assay was calibrated using a cell line that carried a single copy of proviral SIV-gag DNA as described previously (2).

Commercial ELISA kits were used to measure IL-15, IL-7 (human IL-15 and IL-7 kits from R&D Systems) and IL-2 (monkey IL-2 OptEIA kit from BD Biosciences) in the plasma as per the manufacturer’s instructions. Each sample was set up in duplicates. The detection limit for IL-15 was 2 pg/ml, IL-7 was 0.1 pg/ml, whereas the limit of detection for IL-2 was 7.8- 15.6 U/ml. ELISA for IL-2 was performed two times using undiluted plasma and once using diluted plasma.

To determine whether IL-15 up-regulates CD4 expression on CD4 T cells and increases the permissibility to SIV infection, we sorted CD4 T cells by negative selection (CD14⁻CD20⁻CD8⁻) from peripheral blood of eight healthy rhesus macaques and cultured them in the presence or absence of IL-15 and/or SIVmac251. The cells were not labeled with either CD3 or CD4 to avoid activating them through these molecules. Sorted cells were >99% pure (supplemental Fig. 1a⁶), and ∼8 × 10⁵–1 × 10⁶ of CD4 T cells were obtained from each animal. Since IL-15 can stimulate cells in trans, we used highly purified populations of total CD4 T cells (mix of naive and memory) to avoid any bystander effect of cytokines released from other cells. Cells were cultured for 7 days in the presence or absence of 5 ng/ml recombinant simian IL-15 (National Center for Research Resource for Non-human Primate Immune Reagents, Atlanta, GA) and/or 1000 TCID₅₀ of SIV_mac251. Previous studies (10, 16) had used 5 ng/ml in long-term culture studies.

Sorted cell cultures were set up after dividing cells from each animal (n = 4) equally as follows: 1) cells only, 2) cells plus IL-15, and 3) cells plus IL-15 plus IL-15Rα-IgG. An IL-15Rα-Ig-G (IL-15Rα-IgG) fusion protein (National Center for Research Resource for Non-human Primate Immune Reagents) that specifically blocked the effect of simian IL-15 was used in these experiments at a concentration of 10 μg/ml. Cells were harvested after 7 days and labeled with CD3/CD4/CD8/CD95/CD28 to determine the density of CD4 expression on memory CD4 T cells.

Additionally, purified CD4 T cells obtained from four animals were used to determine whether IL-15 increased infection in purified CD4 T cells and to determine whether this effect could be blocked with IL-15Rα-IgG. Sorted cell cultures were set up after dividing cells from each animal (n = 4) equally as follows: 1) cells plus SIVmac251, 2) cells plus SIVmac251 plus IL-15, and 3) cells plus SIVmac251 plus IL-15 plus IL-15Rα-IgG. Cells were harvested, washed, and directly used in a qPCR assay for SIV-gag DNA.

rIL-15Ra-IgG fusion protein was produced at the National Center for Research Resource for Non-human Primate Immune Reagents as per the procedures described below. The coding sequence of MamuIL-15Rα was cloned and sequenced from rhesus macaque of Indian origin as well as other species of macaques (cynomolgus and pig-tailed) (GenBank accession nos. FJ222743–FJ222748) (http://pathology.emory.edu/Villinger/ index.htm). Initially, cDNA was reverse transcribed from total mRNA isolated from PBMC with an IL-15Rα-specific primer IL-15Rα17 (GTA AAA TGG CAC TGA GTT GAG). Subsequently, standard PCR was performed using high-fidelity Taq polymerase (Roche) with primers IL-15Rα19cor2 (GACCATGGAATCACGTGCCCTCCCCCAGTGTCCG TGGAACACGCA) and IL-15Rα3 (GCAGAGAGGCTCCTTCACTCC). The fusion of the IL-15 antagonist, rMamuIL-15Rα-IgG, was performed using a strategy similar to the one described previously for the construction of a PD-1-IgG (17). The Fc portion of the IgG was mutated at positions corresponding to aa 235 and 331 (L235A and P331S) to inactivate potential binding of the fusion protein to complement and to Fc receptors, respectively, effectively inactivating complement and cell-mediated cytotoxicity. The schematic representation of the recombinant protein and nucleotide sequence encoding the mature protein with corresponding amino acid sequence is shown in supplemental Fig. 3.

Briefly, the extracellular domain of IL-15Rα (181 aa) amplified with primers pMT-IL15Rα (CCGGATCCATCACGTGCCCTCCCCCAGTGT) and IL15Rα-IgG2r (ACGTAGATCTACCGACGGTGGATGTAGAGATAGCC) and the IgG2 amplified with primers IL15Rα-IgGf (GGCTATCTCTACATCCACCGTCGGTAGATCTACGT) and IgG6ae (TATGACGTCGAATTCTCATTTACCCGGAGACACGGAGA) were concatamerized by another round of PCR using the primers pMT-IL15Rα and IgG6ae and the overlap created in the previous amplifications. The gene coding for this fusion construct was subcloned into the pMT-BIP vector designed to produce soluble excreted protein from Schneider-2 insect cells in spinner cultures (Invitrogen) using BamHI and EcoRI. All constructs wee verified by sequencing. The protein released in the supernatant of theses culture at a concentration of 7–10 mg/L was then purified by passage over a Protein G-Sepharose capture column and eluted with diluted acetic acid (pH 2.8). The ∼50-kDa purified protein was then dialyzed extensively against PBS and tested for purity, presence of endotoxin, protein content, and biological activity. The protein appeared >90% pure and devoid of any significant endotoxin activity. Blocking activity was tested against recombinant MamuIL-15 using the HT-2 indicator cell line as a readout (18). IL-15Rα-IgG (556 μg/ml) was serially diluted and used in the blocking assay. One inhibitory unit was the equivalent amount of IL-15Rα-IgG needed to inactivate the activity of 1 U of IL-15 bioactivity as measured in the HT-2 proliferation assay.

Flow cytometric data were analyzed using FlowJo version 8.6 (Tree Star). Statistical analysis was performed with GraphPad Prism version 4.0 software. The p values shown in all the figures are not corrected for multiple comparisons.

We first evaluated the kinetics of infection by measuring levels of SIV RNA in plasma, and SIV-gag DNA in total memory CD4 T cells (CD95⁺CD28⁺ and CD95⁺CD28⁻) sorted from peripheral blood and rectal or jejunal biopsies during the first 2–4 wk of infection. Plasma viral loads (Fig. 1,a) peaked around day 14 p.i. and declined thereafter. Fewer than 5% of memory CD4 T cells in either the mucosa or periphery carried viral DNA at day 7 p.i. (Fig. 1 b), confirming previous reports (1, 2). On average, at day 7 p.i., there were 6–9 × 10³ copies of SIV gag DNA/10⁵ memory CD4 T cells. However, by day 10 p.i., the frequency of SIV-gag DNA copies significantly increased to ∼1.9–2.5 × 10⁵ copies/10⁵ memory CD4 T cells, suggesting that most memory CD4 T cells were infected in both the mucosa and periphery by day 10 p.i.. We have previously shown that there are approximately two copies of SIV gag DNA/infected memory CD4 T cell at day 10 p.i. (2).

Following peak infection by day 10 p.i., we observed a massive loss of SIV-infected cells by day 14 p.i. that coincided with the loss of memory CD4 T cells (Fig. 1 c). The similar kinetics of viral infection in the mucosa and periphery argues against the hypothesis that mucosal CD4 T cells are preferentially infected as compared with memory CD4 T cells in other tissues. Rather, the explosion of infection between days 7 and 10 p.i. in both of these tissues suggests that other factors likely cause memory CD4 T cells to become more susceptible to infection.

To determine whether the increased susceptibility of memory CD4 T cells to infection was due to acute immune activation, we evaluated the expression of Ki-67 and HLA-DR on CD4 and CD8 T cells in parallel with SIV infection in memory CD4 T cells. Previous studies have shown that acute SIV infection is characterized by immune activation (4, 5, 19). Our results indicate that acute viral infection leads to a significant increase in the proliferation (Fig. 2,a) and activation (Fig. 2 b) of memory CD8 T cells as early as day 7 p.i., whereas there was no change in either the proliferation or activation of CD4 memory T cells during the first 10 days of infection. This suggests that activation of memory CD4 T cells was not the cause for the apparent increased permissiveness of these cells to SIV between days 7 and 10 p.i. Naive CD4 and CD8 T cells (CD95⁻) did not express either Ki-67 or HLA-DR (supplemental Fig 2, a and b). Consistent with our results, Kaur et al. (5) demonstrated that acute SIV infection was associated with a 3- to 5-fold increase in the Ki-67⁺ CD8 T cells that were activated and had a memory phenotype, whereas there was minimal change in the frequency of Ki-67⁺CD4 T cells.

Interestingly, increased Ki-67 expression on memory CD8 T cells significantly correlated (r = 0.83, p < 0.0001) with SIV infection in memory CD4 T cells (Fig. 2,d), suggesting that factors that drive acute immune activation likely play a role in determining the kinetics of viral infection in memory CD4 T cells. In contrast to the early activation of memory CD8 T cells, a significant increase in the proliferation of memory CD4 T cells did not appear until as late as day 14 p.i. (Fig. 2,a) and was coincident with the massive loss of memory CD4 T cells at day 14 p.i. (Fig. 1 c). This would suggest that the proliferation of memory CD4 T cells at day 14 p.i. likely represents a homeostatic response to the severe loss of memory CD4 T cells.

Next, we sought to determine why memory CD8 T cells were preferentially activated as compared with CD4 T cells. We hypothesized that this might be due to the secretion of factors that preferentially targeted memory CD8 T cells. IL-15 is a proinflammatory cytokine secreted by dendritic cells, monocytes/macrophages, and fibroblasts (20) and likely constitutes a part of the innate immune response to invading pathogens. Previous studies had shown that IL-15 plays a critical role in preferentially regulating homeostatic and infection-driven proliferation of memory CD8 T cells (6, 7, 8, 21). Others (9) have shown that treatment of SIV-infected rhesus macaques with rIL-15 during the early phase of infection led to a significant expansion of CD8 T cells. We hypothesized that the preferential expansion of memory CD8 T cells might be due to the production of IL-15 in response to early viral replication.

We observed a significant increase in IL-15 levels in the plasma of SIV-infected animals at day 7 p.i. (Fig. 3,a) that peaked at day 10 p.i. and declined by day 14 p.i.. Plasma IL-15 levels before day 10 p.i. significantly correlated with plasma (r = 0.77, p < 0.0001) and cell-associated viral loads (r = 0.78, p < 0.0001) and Ki-67 expression on memory CD8 T cells (r = 0.69, p < 0.0001) during the same period of infection (Fig. 3, c–e), suggesting that increased secretion of IL-15 is a response to early viral infection. There was no significant correlation (r = 0.33, p = 0.1156) between plasma IL-15 levels and Ki-67 expression on memory CD4 T cells between days 0 and 10 p.i. (supplemental Fig. 1c). The high levels of plasma IL-15 was surprising since free IL-15 is difficult to detect in the plasma and likely reflects an early response to viral infection since by day 35 the plasma IL-15 levels had reached baseline values. Previous studies (22) have shown that IL-15 was superior to IL-2 in the generation of long-term Ag-specific memory CD8 T cells. The effect of IL-15 on memory CD4 T cells is less clear. Tan et al. (23) showed that unlike CD8 memory T cells, homeostatic proliferation of memory CD4 T cells was independent of both IL-7 and IL-15, and others (24, 25, 26) have demonstrated that treatment of healthy or SIV-infected rhesus macaques with rIL-7 increased the proliferation of both naive and memory CD4 and CD8 T cell subsets.

Plasma IL-2 (Fig. 3 b) was below the level of detection and no significant changes were observed in plasma IL-7 levels (supplemental Fig. 1d), suggesting a minimal role of IL-2 and IL-7 in acute immune activation. Although IL-2 mediates its effect locally, the relative absence of IL-2 in plasma and the demonstration of little or no CD25 expression on CD4 T cells during acute SIV infection (5) argues against a role for IL-2 in the early stages of viral infection. Although surprising, the lack of IL-2 production likely reflects the delay in the generation of SIV-specific immune responses and the relatively low level of activation of CD4 T cells. Reynolds et al. (27) demonstrated that SIV-specific immune responses emerged much later after peak viral infection and loss of CD4 T cells. Likewise, the absence of proliferation and activation of naive CD4 and CD8 T cells along with memory CD4 T cells during the first 10 days of infection argues against the involvement of IL-7.

The absence of proliferation by memory CD4 T cells during the first 10 days of infection even in the presence of significant levels of IL-15 in plasma suggested that IL-15 likely had little effect on these cells. We hypothesized that this may be secondary to the inability of CD4 T cells to bind IL-15 due to the lack of IL-15R on their surface.

The IL-15R consists of the IL-15Rα subunit along with the IL-2Rβ (CD122) and IL-2Rγ (CD132) subunits of the IL-2R. Interaction of the cytokine with the IL-2Rβ and IL-2Rγ on T cells is required for IL-15-induced proliferation of T cells. Studies have, however, shown that IL-15 can bind IL-2Rβ independently of the IL-15Rα and IL-2Rγ chains (28, 29, 30). Similarly, IL-15 has been shown to bind to the IL-15Rα chain directly without the requirement of either IL-2Rβ and IL-2Rγ receptor subunits (31), but this binding does not induce proliferation or activation of T cells. IL-15Rα is known to have a short cytoplasmic tail and its role in cell signaling or activation is still under investigation. IL-15 is mostly presented in trans along with IL-15Rα in vivo.

To determine whether IL-15-mediated preferential proliferation of CD8 T cells was due to the differential expression of IL-15R subunits, we evaluated the expression of IL-15Rα and IL-2Rβ on CD8 T cells and CD4 T cells. CD8 T cells expressed both receptor chains with most cells expressing high levels of IL-2Rβ (Fig. 4, a and b). In contrast, CD4 T cells expressed primarily IL-15Rα but little or very low levels of IL-2Rβ, suggesting that the low level of IL-2Rβ expression on CD4 T cells likely contributes to the low level of proliferation in these cells.

The relative lack of activation in memory CD4 T cells between days 7 and 10 p.i. suggested that other factors contribute to making the memory CD4 T cells more permissive to SIV. Since SIV like HIV uses CD4 as its primary receptor and CCR5 as its coreceptor, we hypothesized that the increased permissiveness of memory CD4 T cells to SIV between days 7 and 10 p.i. might be due to the changes in the level of expression of either CD4 and/or CCR5 on these cells. We addressed this question by evaluating the density of CD4 and CCR5 expression on memory CD4 T cells during early SIV infection.

We observed a significant increase in the level of CD4 expression on memory CD4 T cells by day 10 p.i. as compared with day 7 p.i. (Fig. 5, a and b). The increased density of CD4 expression significantly correlated (r = 0.78, p = 0.0002) with SIV infection in memory CD4 T cells (Fig. 5,e). In contrast to the level of CD4 expression, CCR5 expression decreased between days 7 and 10 p.i. (Fig. 5, c and d) and negatively correlated (r = −0.58, p = 0.0084) with SIV loads in memory CD4 T cells. These findings suggest that the high level of CD4 expression likely contributed to more efficient infection in the memory CD4 T cells. In contrast, the negative correlation of CCR5 expression with cell-associated viral loads indicates that viral infection was associated with the down-regulation of CCR5 expression on memory CD4 T cells. The lower frequency and density of CCR5 expression on memory CD4 T cells at day 10 relative to days 0 and 7 p.i. suggests that CCR5 was not the limiting factor in determining the level of infection in memory CD4 T cells. We have previously shown that memory CD4 T cells expressed enough CCR5 at days 3–10 p.i. to be infected with SIV (2).

To determine whether IL-15 plays a role in up-regulating the density of CD4 expression on memory CD4 T cells and whether this increase in the expression of CD4 coincided with a higher level of infection in memory CD4 T cells, we sorted CD4 T cells by negative selection and incubated them with or without IL-15 in the presence/absence of SIVmac251. Additionally, we cultured negatively sorted CD4 T cells in the presence of IL-15 plus IL-15Rα-IgG fusion protein to block the effects of IL-15 to confirm whether the increase in CD4 expression and infection in sorted cells was primarily mediated by IL-15. IL-15Rα-IgG fusion protein specifically binds and inactivates IL-15 (Fig. 6 a).

As shown in Fig. 6,b, the density of expression of CD4 on the memory CD4 T cells treated with IL-15 was significantly up-regulated as compared with untreated cells. IL-15Rα-IgG fusion protein blocked IL-15-mediated up-regulation of CD4 expression on these cells. Increased expression of CD4 was associated with a significant increase in SIV infection (Fig. 6 c) in sorted CD4 T cells that was significantly blocked with 15Rα-IgG fusion protein. Taken together, the data from these in vitro culture experiments confirm our findings that IL-15-mediated up-regulation of CD4 expression plays an important role in making memory CD4 T cells more susceptible to SIV infection.

Acute HIV infection is associated with massive viral replication and loss of memory CD4 T cells in both mucosal and peripheral tissues, key aspects of viral pathogenesis that are well modeled in experimental infection of macaques with SIV. Beginning from a known time of inoculation, there is a lag, with limited viral replication for the first few days, after which infection amplifies explosively within a period of 2–3 days to reach peak by day 10 p.i.; this process appears to follow the same kinetics in the mucosa and periphery (1, 2). Given the highly activated mucosal tissue microenvironment, it has been generally believed that mucosal CD4 T cells were likely to be preferentially infected as compared with CD4 T cells elsewhere. However, the similarity with which infection proceeds in the mucosa and periphery suggests other factors likely play a role in this process.

Early viral infection is associated with an acute phase-type response that likely induces a cascade of events leading to acute immune activation. Our results show that in experimental SIV infection, acute immune activation occurs very early and before the amplification of viral infection within the memory CD4 T cell compartment. Interestingly, however, the activation seen during early stages of infection was restricted to memory CD8 T cells without any apparent activation of CD4 T cells, observations in line with previous studies (5, 19). The selective activation of memory CD8 T cells but not memory CD4 T cells suggests that specific cytokines such as IL-15 play a role in this process. We found a significant correlation between the activation of CD8 T cells and plasma IL-15 levels.

It was surprising that increased levels of plasma IL-15 was not associated with increased proliferation or activation of memory CD4 T cells, but significantly correlated with the extent of viral infection in these cells. This apparent paradox was resolved by our findings that few memory CD4 T cells express IL-2Rβ (Fig. 4, a and b), and those that express them do so at significantly lower levels as compared with CD8 T cells (Fig. 4 c). Because IL-2Rβ signaling is critical for the activation and proliferation of T cells, these data indicate that the relative absence of increased activation and proliferation of memory CD4 T cells during the first 10 days of infection was likely due to the low levels of IL-2Rβ expression.

Burkett et al. (32) found that expression of IL-15Rα on T cells was dispensable for the generation of memory CD8 T cells. In murine studies, Kamimura et al. (33) showed that CD8 memory cells proliferated even after the depletion of IL-2 in vivo. These memory CD8 T cells expressed IL-2Rβ and blockade of IL-2Rβ signaling completely abolished the division of memory CD8 T cells. In contrast, our results suggest that the low levels of IL-2Rβ expression on CD4 T cells likely allows IL-15 to interact with these cells. However, the relative lack of any change in Ki-67 or HLA-DR expression on CD4 T cells during the first 10 days after infection indicates that IL-15 interaction with IL-2Rβ on CD4 T cells was likely not strong enough to induce activation and proliferation, but was sufficient to make the CD4 T cells more susceptible to SIV infection. Previous studies (34, 35) have shown that IL-15 can enhance HIV replication in the absence of proliferation.

Kaur et al. (5) reported little change in Ki-67⁺CD4⁺ T cells during the first 4 wk of acute SIV infection. In contrast, Picker et al. (11) reported an increased proliferation of effector memory CD4 T cells after administration of IL-15 protein to rhesus macaques. We did not specifically evaluate effector memory CD4 T cells. However, it is highly likely that the high dose of exogenous IL-15 (10 μg/kg body weight/twice a week) used by Picker et al. (11) may have contributed to the proliferation of effector memory CD4 T cells by inducing the secretion of other factors, whereas the highest level of plasma IL-15 we observed was ∼25–30 pg/ml at day 10 p.i. Importantly, our findings indicate that increased activation or proliferation of memory CD4 T cells was not the primary cause for the increase in the level of infection in these cells at day 10 p.i. as compared with day 7 p.i.

Although the low levels of IL-2Rβ explained the absence of any significant level of activation of memory CD4 T cells, it did not explain the highly significant correlation between plasma IL-15 levels and viral infection in memory CD4 T cells. We hypothesized that IL-15 interaction with the low levels of IL-2Rβ was sufficient to up-regulate the expression of the receptors for SIV, namely, CD4 and CCR5. In fact, we found a significant increase in the density of CD4 expression on memory CD4 T cells between days 7 and 10 p.i. that coincided with the increased level of SIV infection in these cells. SIV-gp120 binds to CD4 on CD4 T cells along with CCR5, and higher densities of CD4 expression on memory CD4 T cells significantly correlated (r = 0.78, p = 0.0002) with viral infection in these cells, suggesting that SIV may have bound to memory CD4 T cells more efficiently between days 7 and 10 p.i., leading to a higher a higher frequency of these cells being infected by day 10 p.i.

Previous studies have shown that HIV binds weakly to cells expressing low levels of CD4 (36) and that the low level of CD4 expression on macrophages played an important role in restricting the entry of T-tropic SIV strains such as SIVmac239 (37); this restriction could not be overcome by the overexpression of CCR5 on these cells. Kozak et al. (38) found that CD4 rather than CCR5 or CXCR4 expression determines the kinetics and pathways for gp120 binding, endocytosis, and proteolysis on cells that contain sufficient coreceptor for efficient infection. Platt et al. (39) showed that cells with large amounts of CD4 and low traces of CCR5 are sufficient for maximum susceptibility to HIV-1 strains. Finally, Pesenti et al. (40) demonstrated that susceptibility of macrophages to HIV-1 is significantly increased when they express high levels of CD4 as compared with macrophages that expressed low levels of CD4, whereas there was no significant difference in susceptibility based on CCR5 expression.

These studies support our observations that the low level of infection in memory CD4 T cells coincided with the lower expression of CD4 on these cells, whereas an increased level of expression was associated with near total infection of most memory CD4 T cells. It is highly unlikely that the increased viral replication is the cause for the increased density of CD4 expression. If this were true, then at day 14 p.i. when the plasma viral loads are 2 logs higher than at day 7 p.i., all of the memory CD4 T cells present should have been infected and carry viral DNA. This is however not the case because the frequency of CD4 memory T cells in both peripheral and mucosal tissues (Fig. 1 c) that carry viral DNA at day 14 p.i. are <20% whereas most of the memory CD4 T cells at day 10 p.i. are infected and carry viral DNA. Likewise, the plasma viral loads are quite high at day 7 p.i., yet the level of infection in memory CD4 T cells is much lower than what is seen at day 10 p.i.

In vitro culture experiments using highly purified CD4 T cells in the presence or absence of rIL-15 confirmed our hypothesis that IL-15 plays a major role in early infection. IL-15 treatment up-regulated the level of CD4 expression on highly purified CD4 T cells, an effect that was accompanied by a higher frequency of sorted CD4 T cells being infected with SIV (Fig. 6). The use of a defined system comprised of purified populations of CD4 T cells, and purified rIL-15 minimized potential bystander effects from other cells.

We found no significant change in the level of plasma IL-7 levels during the early phase of infection (supplemental Fig. 1d). IL-7 is a cytokine produced by nonlymphoid cells and plays a role in the homeostatic expansion of both naive and memory CD4 and CD8 T cells (21, 25, 26, 41). Although we cannot completely rule out the potential role of IL-7 in acute immune activation we observed due to the small sample size, the relative lack of proliferation and activation of naive CD4 and naive CD8 T cells (supplemental Fig. 2, a and b) and the minimal changes in proliferation and activation of memory CD4 T cells (Fig. 2, a and b) all argue against a prominent role for IL-7 in immune activation seen during the first 10 days of acute SIV infection. The exact role of IL-7 may need to be evaluated using a larger group of animals to better clarify its contribution to early infection.

In conclusion, our data show that IL-15 plays an important role in acute immune activation and increasing the susceptibility of memory CD4 T cells to SIV infection by up-regulating the expression of CD4, the primary receptor for SIV, on memory CD4 T cells. Although our studies focused on peripheral memory CD4 T cells, the similarities in the kinetics of infection between peripheral and mucosal CD4 T cells suggest that IL-15 likely plays a role in increasing the susceptibility of CD4 T cells in the mucosa to SIV infection. However, additional studies need to be performed using a larger cohort of animals to better understand the role IL-15 plays in immune activation and immunopathogenesis of SIV infection in mucosal tissues.

We thank Nancy Miller at the National Institute of Allergy and Infectious Diseases for help with the animals; Enrico Lugli, Kaimei Song, and Diane Bolton at the Vaccine Research Center; Karen Wolcott and Kateryna Lund at the Biomedical Instrumentation Center; and Dr. Deborah Weiss and Jim Treece at ABL, Inc. (Rockville, MD) for expert assistance with the animals.

The authors have no financial conflict of interest.