Simian Immunodeficiency Virus Infection of Rhesus Macaques Results in Delayed Zika Virus Clearance

Immunocompromised individuals often become symptomatic with infections which are normally fairly asymptomatic in healthy individuals. The particular mechanisms that underlie susceptibility to coinfections in human immunodeficiency virus (HIV)-infected individuals are multifaceted. ZIKV and other flaviviruses are sensitive to neutralizing antibodies, whose production can be limited in HIV-infected individuals but are also sensitive to type I interferons, which are expressed at high levels in HIV-infected individuals. Data in this study highlight how individual components of the innate and adaptive immune responses which become perturbed in HIV-infected individuals influence ZIKV infection.

Adaptive immune responses are also clearly important for the clearance of ZIKV and other flaviviruses (19)(20)(21). Human immunodeficiency virus (HIV) infection of humans and simian immunodeficiency virus (SIV) infection of Asian macaques result in a loss of memory CD4 ϩ T cells, which provide critical immunological defense functions. Most HIV/SIV-infected individuals who are not treated with combination antiretroviral therapy will lose sufficient numbers of these CD4 ϩ T cells to render them susceptible to opportunistic infections (22,23). While this sensitivity to opportunistic infections generally occurs very late after the acquisition of infection, several studies have shown that HIV-infected individuals and SIV-infected rhesus macaques (RMs) mount suboptimal adaptive immune responses to vaccination and viral infections (24)(25)(26)(27). The decreased humoral responses to vaccination, in particular, have been attributed to preferential infection of lymph node follicle-resident T follicular cells, which provide critical help to B cells during the germinal center reaction (27)(28)(29)(30)(31)(32)(33).
While HIV/SIV-infected individuals become immunocompromised and mount suboptimal adaptive immune responses, it is also important to note that another hallmark of these progressive immunodeficiency lentiviral infections is chronic immune activation (34)(35)(36). This immune activation is very commonly associated with increased expression of ISGs (37)(38)(39)(40). Indeed, transcriptional analysis of cell types of multiple lineages from HIV/SIV-infected individuals indicates recent type I IFN signaling (41). Thus, it is also reasonable to predict that the type I IFN production which occurs in HIV infection might limit ZIKV replication.
Both ZIKV and SIV have individually been isolated from wild nonhuman primates (NHPs) and, when experimentally introduced into captive NHPs, have been shown to mirror aspects of human disease progression (42)(43)(44)(45). To explore how HIV infection might impact ZIKV antiviral responses and disease progression, we followed ZIKV viremia, ZIKV RNA levels within cells targeted by ZIKV in vivo, and innate and adaptive immune responses in peripheral blood and lymph nodes (LNs) in non-SIV-infected and SIV-infected RMs that were infected with ZIKV. We found that ZIKV viremia peaked later and was prolonged in SIV-coinfected RMs. These data provide insight into how preexisting lentiviral infection modulates ZIKV viremia in a relevant NHP model for both infections.

RESULTS
Experimental sampling and infection establishment. To determine how SIV infection might influence viremia and the host response to ZIKV infection, we infected 7 Asian RMs with 3,000 50% tissue culture infective doses (TCID 50 ) of SIV mac239 intravenously (Table 1). Three weeks thereafter, we subcutaneously infected these animals with 10 4 PFU of ZIKV isolate Nicaragua/2016 (SIV/ZIKV animals) ( Table 1). For a control, we also infected 12 rhesus macaques (RMs) with ZIKV only ( Table 1). The time point 3 weeks post-SIV infection was chosen to correspond to the peak inflammation and expression of ISGs during acute SIV infection in rhesus macaques but prior to the animals becoming dramatically immunocompromised (38,46). We sampled peripheral blood and lymph nodes at numerous time points both pre-and post-SIV and/or ZIKV infection (Fig. 1A) to study viremia and innate and adaptive immune responses. To measure SIV progression, we measured SIV RNA levels in plasma (Table 1) as well as counted peripheral blood CD4 ϩ T cells (the number of cells per microliter of blood) (Fig. 1B). We found that the SIV/ZIKV cohort had a significant loss of memory CD4 ϩ T cells prior to ZIKV challenge (Fig. 1B) (P ϭ 0.0156), consistent with these animals being progressively SIV infected, but ZIKV infection did not exacerbate the loss of CD4 ϩ T cells in the peripheral blood of either the SIV-infected or the uninfected group of animals (Fig. 1B).
Peak plasma ZIKV viremia is delayed in SIV-infected RMs. We next measured plasma ZIKV viremia between animals that were and were not SIV infected (Fig. 1C). Consistently with the results of previous reports, in non-SIV-infected RMs, ZIKV RNA was detectable in plasma by 2 days postinfection (dpi), peaked to approximately 500,000 copies/ml at 2 to 3 dpi, and then declined to undetectable levels by 7 dpi (47)(48)(49)(50)(51)(52)(53). In contrast, the dynamics of ZIKV viremia were delayed in SIV-infected animals ( Fig. 1C; see also the ZIKV viremia in individual animals discussed in Fig. S1 in the supplemental material). Peak ZIKV RNA in SIV-infected animals occurred at approximately 4 days postinfection and was 30% of that in non-SIV-infected animals. Furthermore, the SIV-infected RMs had significantly lower ZIKV plasma RNA at both day 2 ( Fig. 1C) (P ϭ 0.0020) and day 3 ( Fig. 1C) (P ϭ 0.0023) post-ZIKV infection. Likely as a consequence of this delayed peak, ZIKV RNA levels were higher in SIV ϩ animals than in SIV-negative animals at day 7 after ZIKV infection, a time when ZIKV had been cleared from most non-SIV-infected RMs (Fig. 1C) (P ϭ 0.0029), suggesting that lentiviral coinfection might prolong the flavivirus transmission period.
Peripheral lymph node macrophages are a source of ZIKV replication. To understand ZIKV tropism in our coinfection model, we used flow cytometry to sort T cells, B cells, and macrophages from LNs of animals from our cohort at day 5 and day 27 post-ZIKV infection (our flow cytometric sorting strategy is shown in Fig. S2) and measured ZIKV RNA by quantitative reverse transcription-PCR (qRT-PCR). We found ZIKV RNA in lymph node-resident macrophages and, to a lesser extent, in B cells, with higher levels of ZIKV RNA in macrophages isolated from non-SIV-infected animals at day 5 post-ZIKV infection (Fig. 1D) than in SIV-infected animals (P ϭ 0.021). Interestingly, we also found ZIKV RNA in macrophages isolated from LNs of animals at day 27 post-ZIKV infection, a time point when plasma ZIKV RNA was undetectable in plasma (Fig. 1E). This is consistent with previous reports of the persistence of ZIKV in the lymph nodes of RMs, with virus being detected in peripheral LNs for up to 72 days (47).
Fluorescence in situ hybridization for ZIKV RNA in lymph node sections verified that ZIKV RNA could be detected both within B cell follicles and in the paracortex (Fig. 1F, where ZIKV RNA is indicated with red arrows), with the majority of the ZIKV RNA at days 5 and day 27 post-ZIKV infection found within the B cell follicle (Fig. 1G). To understand how lymph node levels of ZIKV RNA compared between SIV ϩ and non-SIV-infected animals, we performed quantitative image analysis ( Fig. 1G and H). Consistently with our plasma ZIKV RNA analysis, we observed modestly higher levels of ZIKV RNA within lymph nodes of non-SIV-infected animals at day 5 post-ZIKV infection than in SIV ϩ animals (P ϭ 0.056) (Fig. 1H) but significantly higher levels of ZIKV RNA in lymph nodes of SIV ϩ animals at day 27 post-ZIKV infection (P ϭ 0.029) (Fig. 1H). These data are  consistent with delayed clearance of ZIKV in SIV ϩ animals and further demonstrate that the LN serves as a prolonged ZIKV reservoir in RMs. SIV coinfection reduces immune responses to ZIKV. Using flow cytometric analysis, we next measured innate and adaptive immunological responses during ZIKV infection of the animals in our cohort (the flow cytometric gating strategy is shown in Fig. S3). We found that RMs mounted robust proliferative responses to SIV and ZIKV infections in several cellular populations as measured by Ki67 expression (Fig. 2). The SIV-infected animals mounted NK cell proliferative responses to both SIV infection and ZIKV infection ( Fig. 2A). However, the NK cell proliferative response during ZIKV replication among the SIV ϩ animals was restricted compared to that of the non-SIVinfected animals, showing a delayed, dampened, and shorter duration of proliferation ( Fig. 2A) (area under the curve [AUC], P ϭ 0.05). We also measured the proliferation of memory CD4 ϩ T cells (Fig. 2B), memory CD8 ϩ T cells (Fig. 2C), and memory B cells (Fig. 2D) in our longitudinal peripheral blood samples. These proliferative responses tended to peak around day 11 post-ZIKV infection, with a significant response for both cohorts in the Ki67 ϩ memory CD8 ϩ T cell population (P ϭ 0.0368 in the ZIKV cohort and P ϭ 0.0009 in the SIV/ZIKV cohort). We did not observe any proliferative response to ZIKV among B cells (Fig. 2D). SIV infection was also associated with robust proliferative responses from NK cells, T cells, and B cells (Fig. 2). Proliferative responses of T and B cells to ZIKV infection among SIV ϩ animals were, thus, less clear given their alreadyheightened proliferation ( Fig. 2B to D).
The humoral immune response plays a significant role in controlling flavivirus infection (20,21). We next measured titers of ZIKV-specific neutralizing antibody (NAb) responses in longitudinal plasma samples (Fig. 2E). In non-SIV-infected animals, NAbs to ZIKV became detectable by day 7 post-ZIKV infection and then peaked at days 14 to 21, after which, titers plateaued until at least day 56 (Fig. 2E). In SIV/ZIKV-coinfected RMs, the NAb response to ZIKV was slightly delayed and peaked at a lower titer than that in non-SIV-infected animals ( Fig. 2E) (AUC, P ϭ 0.0051). The SIV/ZIKV-coinfected RMs had a 5.3-fold-lower average EC 50 NAb titer at day 11 and a 2.9-fold-lower titer at day 14. Taken together, these data suggest that SIV-infected RMs have reduced innate and adaptive immune responses to ZIKV infection.
ZIKV has no significant effect on CD4 T cell functionality in SIV/ZIKV-coinfected RM. HIV and SIV infection are associated with decreased T cell functionality and loss of CD4 ϩ T cells, particularly Th17 cells from mucosal tissues, supporting the delayed and decreased immune responses observed in our ZIKV/SIV-coinfected animals (54)(55)(56). This immunological perturbation is thought to be due, in part, to alterations to antigenpresenting cells, which result in an environment conducive to the maintenance and development of Th1 cells (57). Given that ZIKV is also an RNA virus which induces Th1 cells (58), we sought to explore whether ZIKV coinfection further exacerbates immune dysfunction. Thus, we assessed whether SIV/ZIKV coinfection changed the functional profile of memory CD4 ϩ T cells. We, therefore, measured the expression of CD40L, interleukin 22 (IL-22), IL-17, IL-2, tumor necrosis factor alpha (TNF-␣), and gamma interferon (IFN-␥) from mitogenically stimulated memory CD4 ϩ T cells by flow cytometry. We then determined the overall functional capacity of Th1 (IFN-␥ ϩ IL-17 -) and Th17 (IL-17 ϩ IFN-␥ -) CD4 ϩ T cells using a simplified presentation of incredibly complex evaluations (SPICE) analysis (Fig. 3). SIV infection decreased the polyfunctional capacity (defined here as the ability to simultaneously express CD40L, TNF-␣, and IL-2) of Th1 cells (Fig. 3A). Consistently with previous reports (54,56), SIV infection resulted in decreased polyfunctionality (simultaneous expression of CD40L, IL-22, IL-2, and TNF-␣) of Th17 cells (Fig. 3B). ZIKV infection did not influence the functional capacity of memory CD4 ϩ T cells (either Th1 or Th17 cells) (Fig. 3).
Interferon signaling predicts peak ZIKV plasma viremia. The observed delays in both ZIKV viremia and immunological responses in coinfected RMs suggested that SIV viremia might, itself, either directly or indirectly limit ZIKV replication. However, when we compared SIV plasma viral loads at day 0 of ZIKV infection with ZIKV plasma viral loads at day 3 post-ZIKV infection (peak ZIKV viremia in non-SIV-infected animals) we found no clear evidence that SIV viremia correlates with the extent of subsequent ZIKV replication ( Fig. 4A) (r ϭ -0.1081, P ϭ 0.8190). Considering that both HIV and SIV infections induce significant type I IFN production and signaling (37-40) and that  (59)(60)(61)(62), we next sought to examine whether there was an association of ISGs with levels of ZIKV viremia in our SIV/ZIKV animals. To measure systemic levels of IFN signaling, we assayed plasma levels of IP-10/CXCL10 in our animals longitudinally after SIV and ZIKV infections (Fig. 4B). This IFN-induced protein has been shown to be high in acutely ZIKV-infected individuals and strongly associated with exanthema (63). In non-SIV-infected animals, ZIKV infection led to increased plasma levels of IP-10 that were maintained until day 14 post-ZIKV infection (Fig. 4B). In SIV/ZIKV-coinfected animals, SIV itself (pre-ZIKV) induced and maintained high plasma levels of IP-10 (higher than those observed in the animals infected only with ZIKV) that did not further increase following coinfection (Fig. 4B). We, SIV Infection of Macaques Delays Zika Virus Clearance ® thus, questioned whether the IFN response elicited by SIV infection might predict lower levels of peak ZIKV viremia and compared the plasma levels of IP-10 prior to ZIKV infection with the levels of ZIKV viremia at day 3 (Fig. 4C). Higher levels of IP-10 prior to ZIKV infection, produced in response to SIV infection, were predictive of lower levels of ZIKV viremia. Moreover, the prolonged ZIKV viremia observed in SIV/ZIKV-coinfected RMs at day 7 post-ZIKV infection was associated with plasma levels of IP-10 prior to ZIKV infection (Fig. 4D). Thus, the IFN-stimulated response mounted against SIV correlated with a delayed peak and prolonged clearance of ZIKV viremia.
Transcriptional analysis of ZIKV and SIV immune responses. Since we observed associations between increased IP-10 levels and lower peak ZIKV viremia at day 3 post-Zika infection in our coinfected RMs, we next sought to explore global transcriptional responses that coincide with SIV and ZIKV infection. Gene expression was measured from mRNAs extracted from peripheral blood mononuclear cells (PBMCs) by NanoString nCounter analysis. We studied samples corresponding to pre-ZIKV infection (day 0) and day 3 post-ZIKV infection, with and without SIV coinfection. Principalcomponent analysis revealed that transcript profiles of PBMCs from the ZIKV RMs were distinct from those of the SIV/ZIKV RMs (Fig. 5A). Most of the variation within the gene expression data was related to SIV or ZIKV infection, most of which was attributable to  (Fig. S4A). Many of the pathways which were induced by SIV infection were also induced by ZIKV infection, albeit to lower levels. Indeed, when we compared non-SIV-infected and SIV-infected RMs prior to ZIKV infection (ZIKV day 0), 163 genes were significantly differentially expressed between the two cohorts ( Fig. 5B and Table S1).
Ingenuity pathway analysis (IPA) of the gene expression profiles revealed a robust type I IFN response to SIV (Fig. 5C and Fig. S4B, ZIKV monoinfection). We compared levels of mRNA for IP-10/CXCL10 in peripheral blood mononuclear cells (from normalized counts from nanostring analysis) to protein levels in plasma (from ELISA) and found a strong, positive correlation (Fig. 5D) (r ϭ 0.9167, P ϭ 0.0013). We further observed an association between levels of IP-10 mRNA at day 0 and ZIKV viremia at day 3 (Fig. 5E), as we did with IP-10 protein levels (Fig. 4C). This association was not limited to IP-10/CXCL10, as we also identified message levels of interferon-stimulated IFIT3 as a predictor of subsequent ZIKV peak viremia (Fig. 5F). To further understand the robustness of these findings, we also examined and found a relationship between ZIKV viremia and levels of IRF7, a central ISG (Fig. 5G). Indeed, IRF7 correlated positively with both CXCL10 and IFIT3 (Fig. 5H) (P ϭ 0.0061 (Fig. 5I) (P ϭ 0.0108). Taken together, our findings demonstrate that the SIV-induced type I IFN response may lead to diminished ZIKV viremia during coinfection.

DISCUSSION
NHPs are often used as models to study viral vaccine efficacy and antiviral therapeutic safety and efficacy and to understand aspects of viral pathogenesis. Owing to Ͼ90% genetic coding conservation between humans and nonhuman primates, viral infections in NHP models result from infections of the same cell types and induce nearly identical pathologies (44). Although NHPs are well-known models for lentiviral and flaviviral infections, to date, their use in coinfection models has been limited. Here, we studied how preexisting SIV infection influences subsequent ZIKV infection. We find that the type I IFN responses elicited in SIV-infected RMs correlates with reduced ZIKV replication, that these coinfected animals mount inferior humoral responses to ZIKV, and that these animals have delayed ZIKV clearance. A recent study, using a small cohort of simian-human immunodeficiency virus (SHIV)-and SIV-infected RMs also found that ZIKV viremia was uncharacteristically low during coinfection; however, the authors concluded that SHIV/SIV infections did not dramatically influence the course of ZIKV infection (64). This small study of only 4 Asian macaques used historical controls and did not study immune responses to ZIKV; thus, the study is somewhat difficult to interpret.
Decreased adaptive immune responses are a hallmark of progressive lentiviral infection (27)(28)(29)(30)(31)(32)(33). Although the underlying mechanisms are multifactorial, preferential HIV/SIV infection of follicle-resident CD4 ϩ Tfh cells (reviewed in reference 65), loss of pathogen-specific memory CD4 ϩ T cells (66,67), exhaustion of adaptive lymphocytes (68,69), and aberrant inflammation (70) are all described contributors. In HIV and SIV infection, antibody/virus complex trapping by follicular dendritic cell (FDC) networks promotes the increased HIV/SIV exposure and subsequent infection of Tfh cells, and this is exacerbated by an inability of virus-specific CD8 ϩ T cells to enter the follicle (65). Indeed, replication-competent HIV can be recovered from the FDC network for years after HIV-infected individuals are treated with antiretroviral therapy (71). That we also find higher levels of ZIKV RNA in the lymph node follicles than in the paracortex (irrespective of coinfection) (Fig. 1G) coupled with delayed ZIKV clearance (Fig. 1C) in SIV-coinfected animals suggests that replication-competent ZIKV may persist in lymphoid follicles and suggests that similar mechanisms may be operative for other flaviviruses. However, it is unclear which particular types of cells become ZIKV RNA ϩ in lymphoid follicles or whether FDCs can become infected themselves. Additional work is required.
Consistently with previous reports, our data clearly demonstrate that SIV infection leads to a robust production of ISGs (37)(38)(39)(40)(41). While infection with ZIKV was sufficient to induce expression of ISGs, the response generated in acute SIV infection far exceeded that induced by ZIKV (Fig. 4B). The differences in ISG induction may be due to distinct differences between the duration of viremia and magnitude of ZIKV infection, which is typically cleared from plasma within 7 days, and those of HIV/SIV, which is persistent in plasma at a typical set point viral load of 1 million copies per milliliter of plasma. Although we further demonstrate that the level of IP-10/CXCL10 at the time of ZIKV infection predicts subsequent ZIKV replication ( Fig. 4C and D), the SIV-generated IFN response was insufficient to completely prevent ZIKV replication. Moreover, recent work suggests that particular ISGs may competitively bind to flavivirus receptors (72)(73)(74), which may explain the delayed ZIKV viremia that we observed in SIV ϩ animals. Paradoxically, this may result from the interferon-responsive genes themselves, in that reduced antigenemia may lead to decreased B cell stimulation and in turn limit the anti-ZIKV antibody responses (75)(76)(77), as observed in Fig. 2E. The delayed antibody responses, in turn, may lead to moderately prolonged ZIKV viremia in SIV ϩ animals. Interestingly, levels of IP-10 have also been shown to differentiate between patients with and without neurological complications after ZIKV infection in Brazil and between infants with and without congenital deformities born to ZIKV-infected mothers, consistent with the idea that IFN signaling is associated with ZIKV pathogenesis (78).
Tfh cells are an important reservoir for HIV/SIV (27)(28)(29)(30)(31)(32)(33), and their predicted infection by SIV in our coinfection model may compromise the generation of ZIKV antibodies. A combination of Tfh loss and ISG-mediated hypoantigenemia may be cumulatively operative in decreasing ZIKV-specific NAbs. Although the degree to which lower levels of antibodies influence susceptibility to flaviviral coinfection is unclear, these data coupled with the prolonged ZIKV viremia demonstrate that SIV infection shapes the resultant anti-ZIKV response.
Importantly, we coinfected our RMs during the early phase of SIV infection, before they were clinically symptomatic and prior to the onset of immunodeficiency. Although vaccine immunogenicity is known to be compromised in chronically HIV-infected individuals, our data further suggest that vaccine efficacy may be similarly compromised or biased during hyperacute infection.
Delayed clearance of ZIKV in SIV-infected RMs coupled with the higher levels of virus within the lymphoid follicle might result in prolonged ZIKV transmission risks. This may be related to antibody complex interactions with follicular dendritic cells, as in previous reports of FDC trapping of HIV/SIV virions (71).
Taken together, this study provides mechanistic insights into how simultaneous viral infections of primates can influence one another and highlights the importance of the innate arm of the immune system in limiting RNA virus replication in vivo. These data also highlight that perturbations to immune responses to secondary infections can be influenced by immunodeficiency virus infection, even before progression to AIDS.

MATERIALS AND METHODS
Sample processing. Whole blood was centrifuged for plasma collection, and subsequently, peripheral blood mononuclear cells (PBMCs) were isolated by standard density centrifugation and cryopreserved. LN biopsy samples were collected from axillary or inguinal sites and then either processed into single-cell suspensions for flow cytometry analysis or fixed in 4% paraformaldehyde and paraffin embedded for immunohistochemistry (79). For flow cytometry experiments, all time points from a specific animal were stained and run together after their collection on day 56. For all other experiments, assayed time points were run within the same experiment.

SIV Infection of Macaques Delays Zika Virus Clearance
® Plasma viral-load assessment. SIV viral RNA (vRNA) levels in plasma were determined by real-time RT-PCR using the ABI Prism 7900 sequence detection system (Applied Biosystems) as previously described (79). Primer pairs for the assay corresponded to the forward nucleotides 1181 to 1208 and reverse nucleotides1338 to 1317 of the SIV mac239 gag gene (80).
To assess ZIKV RNA levels in the plasma of our study RMs, viral RNA was extracted using EZ1 virus minikit v2.0 (Qiagen), and cDNA was synthesized with the SuperScript III first-strand synthesis system (Life Technologies). Absolute quantification of virus genome equivalents (GE; log 10 per milliliter) was determined by real-time PCR using a TaqMan probe (Integrated DNA Technologies) targeting the ZIKV E protein. ZIKV RNA levels were measured in all animals to at least day 11 post-ZIKV infection. All animals, except RHDFiV (SIV/ZIKV cohort) measured below the limit of detection (LOD) of 3.2 log 10 GE/ml at 11 days postinfection (dpi).
RVP production (for neutralization assays). Reporter virus particles (RVPs) incorporating the structural proteins of ZIKV (strain H/PF/2013) (81), dengue virus 2 (DENV2; strain 16681) (82), or West Nile virus (WNV) lineage I (strain NY99) (83) were produced by complementation of a subgenomic green fluorescent protein (GFP)-expressing replicon derived from a lineage II strain of WNV as previously described (81)(82)(83)(84). To determine virus titer, 2-fold dilutions of RVPs were used to infect Raji cells that express the flavivirus attachment factor DC-SIGNR (Raji-DCSIGNR cells) (85) in duplicate technical replicates at 37°C. GFP-positive infected cells were detected by flow cytometry 2 days later. In subsequent neutralization assays, RVPs were sufficiently diluted to within the linear range of the virus infectivity dose-response curve to ensure antibody excess at informative points.
Neutralization assays. For neutralization studies, ZIKV RVPs were mixed with serial dilutions of heat-inactivated macaque plasma for 1 h at 37°C, followed by infection of Raji-DCSIGNR cells in duplicate technical replicates. Infections were carried out at 37°C, and GFP-positive infected cells were quantified by flow cytometry 2 days later. Results were analyzed by nonlinear regression analysis to estimate the concentration of plasma required to inhibit 50% of infection (IC 50 ). For neutralization assays, all plasma samples were initially tested at a starting dilution of 1:60 (based on the final volume of cells, virus, and plasma per well), which was designated the limit of detection.
ZIKV infection assessment in LN cells. Lymph node cells from day 5 and day 27 post-ZIKV infection were sorted into T cell, B cell, and macrophage populations (Fig. S2) for assessment of ZIKV viral RNA via qRT-PCR. RNA was extracted from sorted cells using the Qiagen RNeasy kit (Qiagen) with on-column DNase I treatment per the manufacturer's instructions. ZIKV positive-strand RNA was measured by qRT-PCR using the primers and probe described by Lanciotti et al., with a slightly modified probe to more closely correspond to the Nicaragua ZIKV sequence (86). The following primer and probe sequences were used: forward, 5=-CCGCTGCCCAACACAAG-3=; reverse, 5=-CCACTAACGTTCTTTTGCAGACAT-3=; and probe, 5=-6-carboxyfluorescein (6FAM)-AGCCTACCTTGACAAGCAATCAGACACTCAA-MGBNFQ-3=. The qRT-PCR was performed in triplicate using Applied Biosystems TaqMan fast virus 1-step master mix and a QuantStudio 6 instrument (ThermoFisher). A standard curve was generated using in vitro-transcribed ZIKV RNA using an Ambion MEGAscript kit (ThermoFisher) and a pCRII-TOPO plasmid containing French Polynesia ZIKV bases 700 to 1500 (kindly provided by Thomas C. Friedrich, University of Wisconsin, Madison, WI). The standard curve was used to extrapolate the number of copies of ZIKV RNA per 1,000 cells. Samples were considered positive if at least one of the triplicate wells had a threshold cycle (C T ) value of less than 38. If not all of the wells of the triplicate sample were positive, then any positive C T s from that sample were averaged and divided by 3 to calculate the copy number.
IFN response assessment. IP-10 levels were determined in plasma using the CXCL10/IP-10 human Quantikine enzyme-linked immunosorbent assay (ELISA; R&D Systems). Plasma from day Ϫ14 to day 40 (SIV/ZIKV cohort) and day 0 to day 40 (ZIKV cohort) were run at a 1:2 dilution in duplicate.
In situ hybridization. Rhesus macaque LNs were biopsied 5 or 27 days post-ZIKV infection in both the ZIKV and SIV/ZIKV cohorts. LNs were then fixed overnight in 4% paraformaldehyde at 4°C. LNs were then dehydrated by washing them in sequentially increasing concentrations of ethanol, embedded in paraffin, and allowed to sit overnight at 4°C. LNs were sectioned at 5 m (Histoserv). Sectioned slides were deparaffinized in a dry oven for 1 h at 60°C, followed by sequential washes in 100% xylene, ethanol of decreasing concentrations, and distilled water (dH 2 O). Antigen retrieval was performed by steaming slides immersed in a 0.5% citraconic acid solution (pH 7.4) at 95°C for 20 min, followed by two 10-min rinses in dH 2 O. Sections were treated with Protease Plus pretreatment solution (Advanced Cell Diagnostics) diluted 1:5 in cold phosphate-buffered saline (PBS) for 20 min at 40°C, rinsed twice in dH 2 O, and then treated with hydrogen peroxide for 10 min at room temperature and rinsed twice in dH 2 O. Fluorescent in situ hybridization was conducted using an RNAscope 2.5 HD Brown detection reagent kit (Advanced Cell Diagnostics) according to the manufacturer's suggestions. ZIKV RNA was detected using the V-ZIKA-pp RNAscope probe, and SIV RNA was detected using the SIV mac239 RNAscope probe (Advanced Cell Diagnostics). Following the final amplification step, tissue was stained with Trypticase soy agar (TSA) plus cyanine 3.5 amplification reagent (Perkin Elmer).
Confocal imaging and analysis. Stained LN sections were imaged on an inverted SP5 confocal microscope (Leica Microsystems), and images were analyzed using Imaris software (Bitplane). Threedimensional (3D) surfaces were constructed according to an absolute intensity threshold for fluorescent channels detecting either CD20 (B cell follicle), DAPI (all tissue), or the RNA-FISH probe (ZIKV/SIV RNA). Summed volumes of constructed 3D surfaces were then calculated. 3D surface volume overlap was determined manually. ZIKV/SIV RNA quantifications were normalized as a function of their total tissue volume (e.g., cubic micrometers of the RNA-FISH probe per cubic micrometer of DAPI).
RNA expression profiling. RNA was isolated from 3 million PBMCs for each animal at day 0 and day 3 post-ZIKV infection using the MagMAX-96 total RNA isolation kit (Thermo Fisher Scientific). Each RNA sample was normalized to 50 ng/l. Preparation, hybridization, and detection of RNA samples were carried out by following the NanoString manufacturer's instructions (NanoString Technologies). Subsequent analyses were performed using the nCounter analysis system (NanoString Technologies) and TM4 MeV microarray software suite (http://mev.tm4.org). Principal-component analysis was performed with the R package FactoMineR at default settings. Reads from NanoString were normalized to internal positive and negative controls and housekeeping genes. For ingenuity pathway analysis (IPA), log 2 fold change values (SIV ϩ day 3/SIVday 0) were uploaded to the IPA software (Qiagen) to determine potential upstream regulators.
Quantification and statistical analysis. All analyses were performed using Prism v7.0 software (GraphPad). Mann-Whitney U tests were used when Zika mean titers and infection cell frequencies were compared between cohorts. The Wilcoxon matched-pair signed-rank test was used for cell subset comparisons between time points within the same cohort (i.e., 1B, P ϭ 0.0156; memory CD4 ϩ T cell count preinfection to postinfection). AUC analysis for phenotypic proliferation analysis was carried out from day 2 to day 21 post-ZIKV infection for each animal, and then group cohort comparisons were made using the Mann-Whitney U test (Fig. 2). AUC analysis for Zika mean titers was carried out similarly, except that it was limited to the curve defined by day 2 to day 7 of ZIKV infection (Fig. 1C). Comparisons of ZIKV 50% effective concentrations (EC 50 s) were analyzed via 2-way analysis of variance (ANOVA) of log-transformed data (Fig. 2E). Responses to SIV and ZIKV were determined by repeated-measures one-way ANOVA analysis ( Fig. 2A to D and 4B). Pie charts and polyfunctional capability comparisons were assessed using SPICE v5.3 (National Institute of Allergy and Infectious Diseases), with the permutations set at 10,000 (Fig. 3). The P values included within the SPICE analysis reflect the differences between T cell polyfunctionality based on the expression of different combinations of cytokines and not differences in the total number of functions (e.g., 1ϩ function, 2ϩ functions). Spearman's rank-transformed correlation analysis was utilized to evaluate any associations between immune parameters (Fig. 4A, C, and D and Fig. 5D to I).
Contact for reagent and resource sharing. Further information and requests for resources and reagents should be directed to Jason Brenchley.
Experimental model and subject details. Nineteen healthy RMs (Macaca mulatta) were randomly divided into 2 cohorts, consisting of 7 RMs that were first infected with SIV mac239 (3,000 TCID 50 intravenously) (87) and then subsequently challenged subcutaneously on day 21 post-SIV infection with Zika virus isolate Nicaragua/2016 (UCB 7420) at 10 4 PFU (SIV/ZIKV cohort); 12 RMs were infected with the Zika virus isolate Nicaragua/2016 alone (ZIKV cohort) ( Table 1). This ZIKV challenge virus was chosen due to its relatively high viral replication in RMs. The original isolate (UCB 7420) was from patient serum collected at the University of California, Berkley (kindly provided by Eva Harris, GenBank accession number MN577550). The cohort sizes were based on previous studies of experimental manipulations of SIV progression in Asian macaques (45,79,88). All animals for this coinfection study were between 4.7 and 17 kg and 4 and 18 years of age (Table 1).
All study macaques were housed in certified facilities and cared for in accordance with standards recommended by the American Association for the Accreditation of Laboratory Animal Care (AAALAC). All procedures were performed in accordance with protocols approved by the Institutional Animal Care and Use Committee (IACUC) of the National Institute of Allergy and Infectious Diseases (animal protocol number LVD26). All procedures were carried out under ketamine anesthesia by trained personnel with veterinary staff supervision. This study followed recommended guidelines outlined in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (89) as well as the 2006 Weatherall report (90). RMs were fed twice daily with standard commercial monkey chow and produce and provided with continuous access to water. Each of the animals was singly housed in adjoining cages to allow for noncontact social interaction. Environmental enrichment was provided in the form of primate puzzle feeders, mirrors, and other appropriate toys.
Peripheral blood and lymph node biopsy specimens were collected pre-and post-ZIKV infection as shown in Fig. 1A. These sampling time points were chosen to focus extensively on early ZIKV infection prior to clearance from plasma circulation, as well as to include later time points in order to determine viral immune response dynamics and persistence. Study animals were prospectively screened and, with the exception of RHA0P003, free of the presence of antibodies for West Nile virus, dengue virus, and ZIKV prior to study infection. Neutralizing antibody titers from RHA0P003 were left out of the analysis in Fig. 2E due to clear evidence of prior WNV infection detected in a prospective antibody-dependent enhancement WNV reporter virus particle assay. The NAb titers from this animal were high enough to drive significant differences at day 7 (a 17.8-fold higher average NAb titer in non-SIV-infected RMs), which was no longer significant when the RHA0P003 titers were left out.

ACKNOWLEDGMENTS
We thank Heather Kendall, JoAnne Swerczek, Richard Herbert, and all the exceptional veterinary staff at the National Institutes of Health Animal Center. We thank Shurjo Sen, Giorgio Trinchieri, and the Genomics Laboratory/Leidos Biomedical Research, Inc., for technical assistance with the NanoString NHP assay.
The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does the mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government. We declare no competing interests.