Hippocampal Neuronal Loss in Infant Macaques Orally Infected with Virulent Simian Immunodeficiency Virus (SIV)

The neurological impact of Human Immunodeficiency Virus (HIV) on children includes loss of brain growth, motor abnormalities and cognitive dysfunction. Despite early antiretroviral treatment (ART) intervention to suppress viral load, neurological consequences of perinatal HIV-1 infection persist. Utilizing the pediatric simian immunodeficiency virus (SIV) infection model, we tested the hypothesis that early-life SIV infection depletes neuronal population in the hippocampus. A total of 22 ART-naïve infant rhesus macaques (Macaca mulatta) from previous studies were retrospectively analyzed. Infant macaques were either intravenously (IV) inoculated with highly virulent SIVmac251 at ~1 week of age and monitored for 6–10 weeks, or orally challenged with SIVmac251 from week 9 of age onwards with a monitoring period of 10–23 weeks post-infection (19–34 weeks of age), and SIV-uninfected controls were euthanized at 16–17 weeks of age. We have previously reported that the IV SIVmac251-infected neonatal macaques (Group 1) displayed a 42% neuronal reduction throughout the hippocampal cornu ammonis (CA) fields. The orally-infected infant macaques displayed a 75% neuronal reduction in the CA1 region compared to controls and 54% fewer neurons than IV SIV infants. The CA2 region showed a similar pattern, with a 67% reduction between orally-infected SIV subjects and controls and a 40% difference between IV-and orally-infected SIV groups. In the CA3 region, there were no significant differences between these groups, however both SIV-infected groups had significantly fewer pyramidal neurons than control subjects. There was no correlation between plasma viral load and neuronal populations in any of the CA fields. The loss of hippocampal neurons may contribute to the rapid neurocognitive decline associated with pediatric HIV infection. While each subfield showed vulnerability to SIV infection, the CA1 and CA2 subregions demonstrated a potentially enhanced vulnerability to pediatric SIV infection. These data underscore the need for early diagnosis and treatment, including therapeutics targeting the central nervous system (CNS).


Introduction
Globally, there are over 3.2 million children under the age of 15 currently living with human immunodeficiency virus (HIV), with an estimated one new diagnosis every 2 min [1,2]. transitorily alters sensory-motor function [43,44]. Design based stereology demonstrated that neonatal intrahippocampal Tat 1-72 and gp120 administration resulted in detrimental and regionally selective cell loss within the hippocampus [40]. The protein gp120 was found to reduce the neuronal population within CA2 and CA3 regions of the neonatal hippocampus while intrahippocampal Tat administration reduced neuronal populations in the CA2/3 subfields and the hilus of the dentate gyrus [40].
Additional research findings supported the hypothesis that Tat plays a significant role in pediatric HIV-1 neuropathogenesis and the overall psychological deterioration that is observed in HIV-1 infected children [42]. Results from in vitro cell culture studies indicate that the release of Tat occurs when the rate of cell death is low (as seen during infection), or when cell death is absent (as seen after transfection), and Tat expression is high [45]. In the former study, the highest expression of Tat was observed 48 hours post-transfection with 5%-10% of transfected cells expressing Tat. Results implied that cell growth requires a lower concentration of extracellular Tat than the transactivating effect [45]. A Tat concentration of 100 ng/mL or higher did not produce any growth stimulation, however at concentrations of 0.05 to 50 ng/mL, with peak activity at 0.1 to 1 ng/mL, growth stimulation was observed [45]. Although Tat has been shown to be a part of the neuropathology in perinatal HIV-1 infection, the question whether neuropathology is related to viral load remains unanswered. In adults, neuronal loss can be related to vial load, however, a similar link has not been established in children who experience HIV pathogenesis at a time of neuronal development.
The simian immunodeficiency virus (SIV) infection model in macaques offers a valid alternative because SIV and HIV-1 have similar pathogenesis, including infection of CD4+T cells and macrophages, immune suppression, disease progression, neurological complications in juvenile and adult primates, and routes of transmission [46]. MTCT of SIV can occur by the same routes in monkeys and humans [47].
In a vertical infection model in pigtailed macaques (Macaca nemestrina), offspring of dams that were intravenously (IV)-inoculated with HIV-2287 during the third trimester displayed significant motor and cognitive delays that were correlated with cluster of differentiation 4 (CD4+) T lymphocyte cell counts at birth [48]. We have previously shown that neonatal rhesus macaques infected with SIVmac251 (IV) showed significant neuronal reductions in the hippocampus, loss of immature neurons and gross demylination [38,49]. Here, we expand on these data and investigate the effects of oral SIVmac251 infection at nine weeks of age, simulating oral HIV-1 acquisition by breastfeeding human infants, on hippocampal neuronal populations.

Subjects
The brain samples used in the current study were all collected as part of previously conducted studies [50][51][52][53]. The infant rhesus macaques (Macacca mulatta) in those former studies were nursery-reared in pairs at the California National Primate Research Center (CNPRC) in accordance to American Association for Accreditation of Laboratory Animal Care Standards, with all protocols being approved by the UC Davis Institutional Animal Care and Use Committee. All procedures were performed under ketamine-HCl anesthesia (10 mg/kg i.m.; Parke-Davis, Morris Plains, NC).
The 22 infant rhesus macaques consisted of SIVmac251 IV-inoculated neonates (n = 3) with a monitoring time of 6-10 weeks (hippocampal neuronal counts for this group were previously reported; see [49]). Infant macaques in Groups 2A-C were part of a vaccine study [50,51] and received a weekly oral (PO) low-dose SIVmac251 inoculation regimen (5000 tissue culture infective dose (TCID) 50 per dose) starting at~week 9 of age until infected; then infants were monitored for 10-23 weeks of infection (Table 1). It should be noted that the animals in Groups 2B and 2C had received different SIV vaccine regimens prior to oral SIV challenge (Table 1) [50,51]. Blood samples were drawn on a weekly basis for the PO groups and plasma SIV levels were analyzed [50,51]. Infants in Group 3 represented SIV-naïve control infant macaques (n = 4) that were euthanized between 16-17 weeks of age. * Plasma and cerebrospinal fluid (CSF) SIV RNA levels at time of euthanasia. Group 2 was exposed, starting at 9 weeks of age, once weekly to SIVmac251 by the oral route until infection was verified [50,51]. 1

Brain Preservation and Histology
Upon euthanasia, brains were extracted, post-fixed in 10% buffered formalin phosphate, blocked into 1-cm slabs in the coronal plane, cryoprotected in 30% buffered sucrose, and frozen at −80 • C until further processing. Ten parallel series of coronal sections (50 µm) were obtained from each animal, the first series being Nissl stained with cresyl-violet for design-based stereology, with the remaining sections placed in antigen preserve for future studies [49].
Systematic sections through the hippocampus were Nissl stained and hippocampal pyramidal neurons were quantified using design-based stereology. Although there is an age variance between the groups, hippocampal neuronal populations through the CA fields has been shown to be stable from birth through at least two years of age [54,55]. SIVmac251 was obtained from the Analytical and Resource Core at CNPRC. Plasma viral loads were quantified by real-time reverse transcriptionpolymerase chain reaction (RT-PCR) [50,56].

Design-Based Stereology
Total estimation of the pyramidal neuronal population of the hippocampus cornu ammonis fields (CA1-3) was achieved using the optical fractionator method. Sampling parameters are similar to previously described methods [49] where the topography (4x objective) and superimposed counting frames (dissectors 16,000 µm 3 ; plan fluor oil-immersion 100x numerical aperture (N.A.) = 1.3) were generated through the MicroBrightField StereoInvestigator program (Williston, VT, USA). The stereological parameters are the same as previously used for Group 1 [49] where every 20th section was selected, a standard scan x-y grid was employed (500 µm × 500 µm, 250 µm × 250 µm, and 350 µm × 350 µm for CA1, CA2 and CA3, respectively), and the total estimation of pyramidal neuronal numbers (N) was calculated by the following equation: where ssf is the section-sampling fraction, asf is the area-sampling fraction, tsf is the thickness-sampling fraction, and ΣQ − is total number of objects of interest counted. The average coefficient of error (CE) for total number of neurons was determined to assess reliability of measurement was less than 0.12 for all subjects.

Statistical Analysis
Due to the small group sizes, statistical differences were determined only between two groups at a time applying both one-tailed [49] and two-tailed non-parametric Mann-Whitney U test of significance using the GraphPad Prism V7, and InStat3 programs (La Jolla, California, CA, USA). Correlation analyses between viremia and neuronal data were performed using the Spearman Rank test. Significance was determined as p < 0.05. Viral loads were log transformed and then calculated as area under the curve (AUC) as a way to express overall exposure to avoid acute peaks or multiple peaks in viremia from carrying a relatively larger weight (i.e., skewing the data) than for example if a subject has no peaks but a persistent intermediate viremia.

Results
We have previously reported that IV SIVmac251-infected infants (Group 1) displayed a 42% neuronal reduction throughout the hippocampal CA fields [49]. Here, we show that orally SIVmac251-infected infants (Group 2A) subjects showed a significant neuronal reduction compared to SIV-naïve controls in all three CA regions (p = 0.0286; Figures 1 and 2). It should be noted though that the duration of SIV infection in Group 2 animals was slightly longer (10-12 weeks) compared to Group 1 SIV-infected macaques (7-10 weeks). Vaccinated and orally SIV-challenged infant macaques in Groups 2B and 2C also displayed a significant neuronal reduction in all three CA regions compared to controls ( Figure 2). In regions CA1 and CA3, these neuronal reductions of Group 2B and Group 2C animals were comparable to those observed in non-vaccinated orally SIV-infected macaques of Group 2A ( Figure 2). However, in the CA2 region, the neuronal reduction of infant macaques that had received two doses of the Mycobacterium tuberculosis (Mtb)-SIV vaccine prior to SIV infection had an even more severe neuronal reduction than non-vaccinated SIV-infected Group 2 macaques (Figure 2).
We have previously reported that IV SIVmac251-infected infants (Group 1) displayed a 42% neuronal reduction throughout the hippocampal CA fields [49]. Here, we show that orally SIVmac251-infected infants (Group 2A) subjects showed a significant neuronal reduction compared to SIV-naïve controls in all three CA regions (p = 0.0286; Figures 1 and 2). It should be noted though that the duration of SIV infection in Group 2 animals was slightly longer (10-12 weeks) compared to Group 1 SIV-infected macaques (7-10 weeks). Vaccinated and orally SIV-challenged infant macaques in Groups 2B and 2C also displayed a significant neuronal reduction in all three CA regions compared to controls (Figure 2). In regions CA1 and CA3, these neuronal reductions of Group 2B and Group 2C animals were comparable to those observed in non-vaccinated orally SIV-infected macaques of Group 2A (Figure 2). However, in the CA2 region, the neuronal reduction of infant macaques that had received two doses of the Mycobacterium tuberculosis (Mtb)-SIV vaccine prior to SIV infection had an even more severe neuronal reduction than non-vaccinated SIV-infected Group 2 macaques ( Figure  2).   Volume differences were found only in the CA1 and CA2, but not the CA3 region between the various groups ( Figure 2). Both CA1 and CA2 volumes were significantly lower in Groups 1 and 2A-C compared to controls (Group 2) (Figure 2). The volume in infant macaques orally infected with SIVmac251 (Group 2A) was comparable to that observed in Group 2B and Group 2C infants that were vaccinated prior to SIV infection ( Figure 2). There was a significant correlation between volume and neuronal population (Figure 3).
The changes in neuronal frequencies and volumes between SIV-naïve and SIV-infected animals were not directly correlated to plasma viremia at the time of euthanasia (Figures 4 and 5). Similar results were obtained when we tested for a correlation between CSF viremia and neuronal frequencies and volumes. As previous studies have shown that pathogenesis outcome is dependent on peak and set-point viremia [57,58], we also tested for a potential correlation between acute and Groups 2A-C display significantly lower neuronal populations compared to Group 3. Volume differences were restricted to the cornus ammonis (CA)1 and CA2 subfields. p values reflect differences between each of the treated groups (G2A-C and G3) and the simian immunodeficiency virus-naïve (SIV-naïve) control group (G1). * p = 0.029, ** p = 0.006 except # p = 0.03 which is a comparison between Groups 2A and 2C; bar in the middle of the data points represents mean. One-tailed test between G3 and G1 [49] all other comparisons were two-tailed tests.
Volume differences were found only in the CA1 and CA2, but not the CA3 region between the various groups ( Figure 2). Both CA1 and CA2 volumes were significantly lower in Groups 1 and  (Figure 2). The volume in infant macaques orally infected with SIVmac251 (Group 2A) was comparable to that observed in Group 2B and Group 2C infants that were vaccinated prior to SIV infection ( Figure 2). There was a significant correlation between volume and neuronal population (Figure 3). SIVmac251 (Group 2A) was comparable to that observed in Group 2B and Group 2C infants that were vaccinated prior to SIV infection (Figure 2). There was a significant correlation between volume and neuronal population (Figure 3).
The changes in neuronal frequencies and volumes between SIV-naïve and SIV-infected animals were not directly correlated to plasma viremia at the time of euthanasia (Figures 4 and 5). Similar results were obtained when we tested for a correlation between CSF viremia and neuronal frequencies and volumes. As previous studies have shown that pathogenesis outcome is dependent on peak and set-point viremia [57,58], we also tested for a potential correlation between acute and early chronic viremia, as assessed by area-under-the-curve analysis (AUC), with neuronal changes. The degree of neuronal frequency reduction in SIV-infected macaques was not associated with acute, chronic or terminal plasma viremia (Figure 4). Likewise, the volume of the subregions was not correlated with acute, chronic or terminal plasma viremia ( Figure 5).  The changes in neuronal frequencies and volumes between SIV-naïve and SIV-infected animals were not directly correlated to plasma viremia at the time of euthanasia (Figures 4 and 5). Similar results were obtained when we tested for a correlation between CSF viremia and neuronal frequencies and volumes. As previous studies have shown that pathogenesis outcome is dependent on peak and set-point viremia [57,58], we also tested for a potential correlation between acute and early chronic viremia, as assessed by area-under-the-curve analysis (AUC), with neuronal changes. The degree of neuronal frequency reduction in SIV-infected macaques was not associated with acute, chronic or terminal plasma viremia ( Figure 4). Likewise, the volume of the subregions was not correlated with acute, chronic or terminal plasma viremia ( Figure 5).    There was a lack of correlation between regional volume and plasma viral loads. Figure 5. Volume correlation. There was a lack of correlation between regional volume and plasma viral loads.

Discussion
We previously reported significant pyramidal neuronal loss in the CA1-3 fields of the hippocampus following IV inoculation with SIVmac251 within 72 h after birth with an average infection time of 8.3 weeks [49]. Here, we expanded on these data and tested the effect of oral SIV infection on neuronal damage. The oral route of SIV infection better represents HIV-1 acquisition by breastfeeding, the most common route of new mother-to-child transmissions of HIV-1. Furthermore, the age period of 9-23 weeks is more representative of the time of breastfeeding or mixed feeding practices in the clinical setting [13,14].
Our data suggests a particular vulnerability of the CA1 and CA2 region of the hippocampus. While both IV and PO SIV-infected infant macaques displayed significant reductions in pyramidal neuronal populations in the CA1-3 fields compared to controls, this effect appeared to be even more pronounced in the orally-infected infants compared to the perinatally IV-infected subjects. However, this difference was not significant and should not be over-interpreted as infants in these two groups differed not only in age at the time and route of SIV infection, but also in the duration of SIV infection. Thus, future studies should determine whether increased viral exposure time is associated with a progressive neuronal loss.
Studies of neonatal rodent models of intracranial HIV-1 tat antigen administration support clinical evidence that the neurons of the hippocampus, and hippocampal neurogenesis, are specifically susceptible to the neurotoxic cascade of HIV-1 proteins [40]. Upon intracranial injections of tat and gp-120 in neonatal rats, there is a reduction of neuronal populations in the CA1, CA2 and CA3 fields, delays in eye opening and reflex development, and altered prepulse inhibition, which are correlated to a decline in spatial memory [40,41]. In the current investigation, we examined a narrow parameter of hippocampal integrity, neuronal population and volume. With the decreased volumes of the CA1-2 regions and the reduced neuronal population in CA1-3 in all SIV-infected infants (Groups 1-2A-C), it would be expected that changes in neural circuitry extend beyond the hippocampus. Clinical data indicates significant deficits in working memory and executive functioning [28,59], which would indicate a disrupted hippocampal-prefrontal circuitry [60][61][62].
Much of pre-existing clinical research focuses on global cortical changes and the basal ganglia [23,31,63], but there is increasing evidence of HIV-1 related toxicity in the hippocampus [49]. Previously, neural atrophy of the hippocampus has been reported in young adult monkeys following SIV infection [64]. CA1 pyramidal neurons are probably the most studied class of neuron in the brain, and better understood from a structural and functional standpoint compared to other types of neurons in the hippocampus [65]. This is due to relative ease of obtaining intracellular recordings and field potential recordings from these regions [65]. Despite our findings of a decrease in hippocampal neurons of SIV-infected subjects, a direct connection between neuronal populations and viral load in plasma or CSF at the time of brain collection could not be established.
Clinical studies have shown that HIV RNA levels are extremely high during infancy and early childhood, in comparison to adults [66]. A study found a trend in the mean HIV-1 RNA load rising from low values (<10,000 copies/mL) at birth, to extremely high (>100,000 copies/mL) within the first 24 months of life, and then decreasing very slowly until the end of those 24 months [67]. This same study compared the viral load of infants with early (in utero transmission) vs. perinatal infection and discovered significantly varying median HIV-1 RNA values during the early months of life. At birth these values were 10,800 copies/mL, and less than 400 copies/mL; and after one month of life these values were 716,000 copies/mL and 100,000 copies/mL; with peak values of 780,000 copies/mL and 243,000 copies/mL, respectively [67]. Data from a perinatal cohort demonstrated a strong connection between high HIV RNA viral load during early infancy and poor clinical outcome, and also found that children with a higher viral load experienced rapid progression of the disease compared to children with lower loads [67]. Similar correlations between plasma viral RNA levels and disease-free survival have also been described in SIV-infected infant macaques [68,69].
Although HIV-1 has not been shown to infect neurons, there is evidence that the virus exerts its actions indirectly on neurons [70]. There is also evidence that HIV proteins are responsible for disrupting ion homeostasis and depolarizing neurons; while other proteins are restricted to neurons to disrupt neuronal membranes or induce apoptosis of the proteins [71]. Indirect neurotoxicity of HIV may involve the activation of glial cells, which have shown to result in the production and release of inflammatory cytokines like tumor necrosis factor (TNF)-alpha [70,72]. Our finding that infant macaques in Group 2C had the most severe neuronal reduction in CA2 is consistent with this conclusion. We have previously shown that vaccination with Mtb-SIV caused persistent systemic immune activation prior to and post-SIV infection [51]. Although the prior study did not analyze brain inflammation, mycobacteria-induced immune activation was persistent and systemic, and thus, could have affected the brain as well [51]. Astrocytes, microglia and macrophages are significant HIV-1 targets in the brain and research suggests that indirect mechanisms could, in part, account for severe neuronal damage observed in HIV-positive patients [73][74][75].

Conclusions
The loss of hippocampal neurons may contribute to the rapid neurocognitive decline associated with pediatric HIV infection. While each subfield showed vulnerability to SIV infection, the CA1 and CA2 subregions demonstrated a potentially increased vulnerability to SIV infection. These data underscore the need for early and continued ART as well as the development of therapeutics targeting the CNS [38].