Altered gene expression in asymptomatic SHIV-infected rhesus macaques (Macacca mulatta)

Simian-Human immunodeficiency virus is a chimeric virus which, in rhesus macaques (Macacca mulatta) closely imitates immunodeficiency virus infection in human (HIV). A relatively new way to study pathogenesis of viral infection is to study alterations in host gene expression induced by the virus. SHIV infection with certain strains does not result in clinical signs. We hypothesized that alterations in gene expression relating to the immune system would be present in SHIV-infected animals despite the lack of clinical signs. Splenic tissue from four adult male Indian-origin Rhesus monkeys serologically positive for non-pathogenic SHIV 89.6 was processed by cDNA microarray analysis. Results were compared with the corresponding outcome using splenic tissues from four unexposed adult male Rhesus monkeys. Subsequent gene analysis confirmed statistically significant variations between control and infected samples. Interestingly, SHIV-infected monkeys exhibited altered expression in genes related to apoptosis, signal transduction, T and B lymphocyte activation and importantly, to immune regulation. Although infected animals appeared asymptomatic, our study demonstrated that SHIV-infected monkeys cannot reliably be used in studies of other infectious agents as their baseline gene expression differs from that of normal Rhesus monkeys. The gene expression differences in SHIV-infected animals relative to uninfected animals offer additional clues to the pathogenesis of altered immune function in response to secondary infection.


Background
Simian immunodeficiency virus (SIV) infection of rhesus macaques exhibits many similarities to human immunodeficiency viral (HIV) infection of humans. Most pathogenesis and vaccine studies for HIV-1 have been undertaken in either SIV-macaque or a chimeric simianhuman immunodeficiency (SHIV)-macaque model [1]. SHIV strains have the viral envelope of HIV but the gag/ pol genes of SIV. Pathogenesis is similar with respect to macrophage and T lymphocyte cell tropism, histopathologic changes, CD4-cell depletion and clinical signs of auto-immune deficiency syndrome (AIDS) in virulent strains. HIV and SIV additionally cause cognitive and motor impairments in infected patients and monkeys, respectively [2]. Host factors may play a role in degree of pathogenesis between varying SHIV constructs, as one study reported observing similar viral loads in rhesus monkeys infected with pathogenic and non-pathogenic SHIV constructs [1].
Gene expression studies have grown increasingly popular as a tool to mine large amounts of data from treated and control populations. Such data can be used to examine host factors involved in SHIV, and thereby HIV, pathogenesis. To our knowledge, microarray data from SHIVinfected Rhesus macaques have not yet been examined for genes affecting immune response and inflammation. Gene expression data have the potential to greatly expand the understanding of SHIV-host interaction beyond the limited number of cell types or cytokines generally examined.
In animals free of clinical signs of SHIV, altered baseline gene expression data may give clues to the pathogenesis of altered immune response to secondary infections. Studies involving HIV-infected humans demonstrated suppression of IL-2 in response to select antigens and increase in TNF-α even prior to the onset of CD4+ T-cell depletion [3,4]. Gene expression data collected in this study from SHIV 89.6-infected monkeys demonstrate that these animals are not genetically 'normal' and cannot ethically be used for studies involving other infectious agents, if at all, without an explicit caveat listing their SHIV status. Comparison of gene expression patterns collected from SHIVinfected and uninfected animals to that of the matched animals exposed to select bacterial and viral agents would provide a more complete understanding of SHIV effect on immune response to particular infectious agents. Extrapolation to the HIV-patient response to secondary agents may then be attempted. Gene expression data may also provide clues to pathogenesis of cognitive and related ailments arising with HIV infection.

Clinical history
A brief description of treated and control animals is given in Table 1. All monkeys were male; while two of them (one SHIV-positive, one SHIV-negative) were Herpes Bpositive. Table 2 summarizes abnormalities in clinical chemistries including complete blood counts of the SHIV-infected animals. Abnormalities were minimal. Attending veterinary clinicians considered these animals asymptomatic with respect to SHIV infection.

Micro-array analysis of SHIV-infected versus uninfected
Using the 38 most varying genes between SHIV-infected and SHIV-uninfected animals, we performed Principle Component Analysis, a non-hierarchal clustering tool, to revalidate the t-test result. Figure 2 demonstrates that the SHIV positive and negative groups were clustered together, keeping a significant distance between them along the first principal component (X-axis), which shared the highest fraction of group variation. The pattern of clustering also suggested that the gene expression variability was independent of the animals' Herpes B status.
Gene ontology study, using GeneCite [6], associated the members of the differentially expressed genes to a range of important biological and pathological functions including immune defense, cell death or apoptosis, cell growth, signal transduction and others. Table 4 represents the functional classification of some of the genes of interest.

Discussion
Simian immunodeficiency virus (SIV), previously referred to as simian T-cell lymphotropic virus type III (STLV-III), induces an AIDS-like disease in its natural host, rhesus macaques. HIV and SIV, members of the lentivirus subfamily of retroviruses, not only resemble each other by their antigenicity, but also bear remarkable similarity in their biological properties, such as cytopathic effect and tropism for CD4-bearing cells. These criteria render the chimeric SHIV the best animal model currently available for HIV study.
In this study, we examined gene expression in SHIVinfected male rhesus macaques of Indian origin using a genomic perspective and compared the results to uninfected age, gender and Herpes B-status-matched controls. Although infected animals were without clinical signs related to SHIV infection, a significant number of genes exhibited significantly altered expression concurrent with SIV infection.
Ontological research revealed that several genes, namely Reports suggest that overexpressed ALDH5A1 changes the concentration of gamma-aminobutyric acid (GABA) and glutamate, commencing henceforth excitotoxic damage, a well-established clinical marker of HIV activity [9]. Underexpressed SDHD and CTNS are associated with immunodeficiency through curbed monocyte and CD4+ T cell -induced immunoregulation [10], respectively.
In summary, in this small sample of SHIV-infected Rhesus macaques, expression was consistently altered in specific groups of genes which regulate a broad range of biochemical functions. A few important members of the genome of interest are discussed here. The present study, along with correlating some genes with SHIV and HIV model, identifies several novel genes as potential therapeutic markers for immune deficiency studies. Furthermore, results of this study suggest that SHIV infection of rhesus macaques may influence immune response to a second agent, even if baseline levels of clinical measurements appear normal. This study substantiates and validates the concern that an infected (i.e., antibody-producing) but asymptomatic animal should not be used in any other study involving infectious agents unless the pattern of gene expression to that agent is compared to normal animals' pattern, one agent at a time.
Note: microarray data have been submitted to the Gene Expression Omnibus (GEO) and can be searched using the Platform ID: GPL3395.

Animals and virus
Four adult (7-8 years old) male Rhesus macaques (one Herpes B-positive and three Herpes B-negative) that were previously exposed to SHIV 89.6 strain (Animal identifications: FFG, JGH, PHB and TTH) were euthanized due to being declared 'excess' and no longer usable due to their serologically positive SHIV status. Splenic tissue was collected from each animal upon euthanasia and immersed in RNA Later ® for 30-60 minutes before freezing at -80C.
SHIV 89.6, like all SHIV strains, has the env gene from the HIV-1 strain. All four animals had been challenged with 1.0 ml intravenous SHIV 89.6, a non-pathogenic strain, and became seropositive. Previous studies by the same researchers showed that seropositive animals were PCR positive as well (WRAIR Protocol TO03-98). All animals remained free of clinical signs. Complete blood counts and serum chemistry profiles were performed on the SHIV-positive animals and were within or very close to normal limits. The negative control animals were Indianorigin adult male Herpes B-negative Rhesus macaques. Splenic tissues were kindly provided by Scripps Institute, the National Institute of Health, and the Oregon National Primate Research Center. Tissue from a SHIV-negative animal (DB-87, provided by the Tulane National Regional Hieratically clustered Tree-view of genes differentially expressed between the SHIV positive and negative animals Figure 1 Hieratically clustered Tree-view of genes differentially expressed between the SHIV positive and negative animals.

Control SHIV
Primate Research Center) was Herpes B-positive to control for the Herpes B-positive status of one SHIV-infected animal Table 1 represents an overview of the Rhesus macaques used in this study. Table 2 shows the clinical results of the SHIV-positive rhesus macaques

RNA isolation
Splenic tissue samples stored in RNALater ® (Ambion, TX) at -80C were thawed in 1.5 mL tubes on ice. Tissue was submerged in Trizol ™ (Invitrogen, CA) solution and RNA isolation was carried out paccording to the Trizol ™ Reagent manufacturer's recommended instructions. RNA was ethanol-precipitated, air-dried and re-suspended in 20 ul/ sample of nuclease-free water. RNA quantity was measured via spectrophotometry followed by analysis with a Bioanalyzer 2100 (Agilent Technologies, CA)

Custom made cDNA microarray SlidePreparation and hybridization
The gene library for the present project was commercially obtained from Research Genetics (Invitrogen, CA), containing 7489 genes, including 7019 known genes, 249 unknown genes and 110 expressed sequence tagged genes (ESTs). Superamine coated Telechem slides (Telechem Inc., OR) were used for printing the cDNA clones using 12 × 4 pin format, on a Virtek chip writer professional microarrayer in KemTek, Inc, MD. The printed slides underwent UV cross-linking, followed by post-processed by succinic anhydride treatment. The Micromax™ Tyramide Signal Amplification (TSA)™ Labeling and Detection Kit (Perk-inElmer, Inc., MA) was used as directed by the manufacturer to determine relative gene expression of the collected samples. Custom-made reference RNA was prepared by

Figure 2
Principal component analysis was performed over the SHIV infected and non-infected population. Though the animals were clinically reported asymptomatic, the SHIV treated and control samples cluster far from each other along PCA1 axis. The result also suggests that the Herpes B status does not affect the outcome. Here PCA1 has 61.7% population, while PCA2 and PCA3 shares 12.6% and 8.56% of the population respectively.

PCA 82.86%
combining aliquots of RNA from 33 normal Rhesus tissues and was used on every slide as the array controller, to check overall sensitivity of array printing, and to monitor reverse transcription, labeling and hybridization efficiency. Sample hybridization was carried out at 55°C for sixteen hours. A laser detection system was used (GenePix 4000b, Axon Instruments, CA) to scan the finished slides. Intensity of the scanned images was digitalized through Genepix 4.0 software (Axon Inc., CA).

Microarray analysis
Data cleansing and statistical analysis was carried out using Genespring ® 7.0 (Agilent Tech., CA). Local background was subtracted from individual spot intensity. Genes that failed this 'background check' in any of the eight given experiments were eliminated from further analysis. Each chip was next subjected to intra-chip normalization (LOWESS). The genes that varied most between control and treated sample sets were selected via t-test analysis. The p-value cutoff was set at 0.05. Four hundreds and thirty two genes were differentially expressed between SHIV -infected and control uninfected animals with p < 0.05.
The pattern of gene expression variability of the experimental set having reduced dimension was evaluated using principal component analysis (PCA) classifying SHIV positive and negative samples as the two variable classes [16].

Real Time PCR
The t-test result was corroborated through real time polymerized chain reaction (Real-time PCR). A webbased primer designing tool was used to design the primers for the selected genes [17]. The specificity of each primer sequence was further confirmed by running a blast search. Reverse transcription and Real-time PCR reactions were carried out using reverse transcription kit (Invitrogen, CA) and Real-time PCR kit (Roche, IN), respectively. Each reaction with five technical duplicates was run in I-Cycler machine (Bio-Rad, CA). Each sample was also amplified against the house-keeping probe of the experiment: glyceraldehyde 3 phosphate dehydrogenase (GAPDH). The resultant cycle threshold data from each real-time-PCR 'run' was converted to fold-change using an established algorithm [5].
Quantitative and qualitative verification of the PCR product was accomplished by performing 1% agarose gel electrophoresis using SYBR Green I (Kemtek, Rockville, MD). Gel images were captured using PharosFX Molecular Imager system (Bio-Rad, CA) scanner and analyzed using Quantity One software (Bio-Rad, CA).

Authors' contributions
EEC participated in the design of the study, carried out the microarray and real time PCR studies and participated in drafting the manuscript. RH participated in the design of  -TCC AAG TCC TGT GTC TCC TC-3'  5'-TCT TGA GTT CAG CCC TCT CC-3'  109 bp  Bit1  AI339248  CGI-147 protein  5'-TGG CTG TTG GAG TTG CTT G-3'  5'-TGT GTG TCT TGC TCG TCT TG-3'  93 bp  CLCA2  AI675394  chloride   The first, second and third columns list the GeneBank ID, Symbol and Gene Name respectively. The Fourth column stands for the corresponding fold change of SHIV positive animal with respect to that of the control animal, averaged over the entire population, i.e. (Average fold change for all SHIV positive animals)/(Avg FC for all control animals) the study, carried out the microarray data analysis, data mining and participated in drafting the manuscript. NC participated in the microarray data analysis and participated in drafting the manuscript. AP participated in the microarray and real time PCR studies.
SAM participated in the microarray and real time PCR studies. MJ conceived of the study, and participated in its design and coordination. All authors read and approved the final manuscript.