|Home | About | Journals | Submit | Contact Us | Français|
Three of six morphine-dependent monkeys progressed rapidly to AIDS and died by 20 weeks in our SIV/SHIV non-human primate model of drug addiction and AIDS. We studied the evolution of the SIV vpr gene in both cerebrospinal fluid (CSF) and plasma in these rapid progressors, in their normal progressor counterparts and in infected, drug-free controls at 12 and 20 weeks post infection. Viral RNA was amplified, cloned, and sequenced to permit phylogenetic analyses of diversity and divergence of the vpr locus. As we found for SIV tat and env, the vpr gene evolves inversely to the rate of disease progression. Further, we found evidence that compartmentalization of the virus in plasma and CSF is significantly greater in the normal progressors than in the morphine-dependent, rapid progressors. Interestingly, although our previous work with the accessory gene nef indicated no association between disease progression and evolution, the accessory factor vpr, behaves similarly to the essential lentiviral genes tat and env.
All retroviruses share common elements in genomic structure including the gag, pol, and env genes. In addition to these universally essential genes the primate lentiviruses, including both HIV-1 and SIV have a well characterized set of accessory gene products(Vogt, 1997). Only two of these accessory genes, tat and rev, are absolutely required both in vivo and in vitro for viral replication(Dull et al, 1998;Luciw, 1996). Another accessory gene, vpr (encoding the viral protein R) is common to both HIV-1 and SIV, yet unlike the tat and rev gene products, vpr has been shown to be dispensable for viral replication in a number of cell culture and animal model settings(Le Rouzic & Benichou, 2005;Aldrovandi & Zack, 1996). In in vivo studies, however, a vpr− SIV is only half as virulent as the wild type SIVmac239, while a deletion of two accessory gene products, vpr and nef, is 99.5% less virulent and does not lead to simian AIDS(Desrosiers et al, 1998), indicating the importance of this ~100aa protein.
Vpr is packaged in the virus via an association with the p6 component of the viral gag gene(Kondo et al, 1995;Paxton et al, 1993). During viral replication, Vpr has been attributed a variety of roles including increasing the accuracy of reverse transcription, assisting in the nuclear import of the preintegration complex, transactivation of the viral promoter, regulation of cell cycle and in some cells, regulation of apoptosis(Le Rouzic & Benichou, 2005;Mueller & Lang, 2002;Patel et al, 2002;Chen et al, 2004;Mahalingam et al, 1997;Muthumani et al, 2002b;Piller et al, 1999;Di Marzio et al, 1995). In the specific setting of SIV, some of the functions of Vpr (in HIV) are shared or overlapping with SIV Vpx. However, at least the transactivation, apoptosis and cell cycle regulation appear to be conserved in Vpr proteins of primate lentiviruses(Planelles et al, 1996;Zhu et al, 2001;Philippon et al, 1999). Thus Vpr protein has demonstrable roles in a variety of aspects of viral infection and pathogenesis. Still, there is little information of the possible pathogenic role of Vpr during the setting of drug abuse regardless of the large impact of drugs of abuse on the HIV/AIDS epidemic both in the US and worldwide.
In the US alone, drug abuse is a co-factor in the acquisition and or pathogenesis in nearly 1/3 of the HIV-1 infection cases(Purcell et al, 2004). Although this clearly presents a considerable burden on the health care system and the economy, studies of the impact of drug abuse (morphine) have provided mixed message of harm(Chuang et al, 2005) versus reduced pathogenesis(Donahoe, 2004;Kapadia et al, 2005). In part to address this ongoing controversy, we have developed a non-human primate model of AIDS/drug abuse using rhesus macaques addicted to morphine and then infected with a combination of viruses (SIV17E-Fr, SHIV89.6P, SHIVKU-1B) to more rapidly induce simian AIDS (sAIDS), including a precipitous loss in CD4+ T cells characteristic of AIDS in humans(Kumar et al, 2004a). In this model, we have observed that half of the morphine addicted animals progress rapidly to AIDS and death by 20 weeks post-infection (Table 1)(Kumar et al, 2004a;Noel, Jr. & Kumar, 2006). While all animals in the study showed a rapid peak in virus and severe drop in CD4+ T-cells during the acute phase of infection, the rapid progressors never regained CD4+ T-cells; nor did they establish control of the viral replication in either the plasma or the cerebrospinal fluid (CSF)(Kumar et al, 2004a). Macaques that did not show rapid progression, including half of the morphine group and all members of the morphine-free cohort, established some recovery of CD4+ T-cells and control over plasma and CSF viral load. One goal of our work with these cohorts has been to evaluate the impact of viral evolution in rapid progression in the setting of drug abuse. We have already established a relationship between viral evolution and the rate of disease progression for a number of critical viral genes in this system. Both the essential accessory gene tat (in plasma and in CSF) and the structural gene env (plasma) have shown an inverse correlation between evolution and disease progression in the setting of morphine-dependence(Noel, Jr. et al, 2006;Noel, Jr. & Kumar, 2006;Tirado & Kumar, 2006). However, our experience to date with an in vitro dispensable accessory gene (nef) showed no correlation between evolution rate and disease progression(Noel et al, 2006). This suggested that some non-essential accessory genes could prove less influential to rapid pathogenesis. We have now extended this analysis to the vpr gene where we find that unlike the dispensable accessory gene nef, and like the essential accessory gene tat, evolution is inversely correlated to disease progression both in plasma and CSF, and furthermore, that compartmentalization of virus does not occur in rapid progression.
As with our previous studies, viral nucleic acids were extracted for two times points roughly 12 and 20 weeks post infection from cell-free fluids and subjected to RT-PCR amplification, cloning and sequencing(Noel, Jr. et al, 2006;Noel, Jr. & Kumar, 2006;Tirado & Kumar, 2006;Noel et al, 2006). Phylogenetic comparisons were made of all clones in both plasma and CSF within each individual monkey. The resulting phylogenetic trees (Figure 1) show that vpr evolution appears to be inversely correlated with disease progression. Each tree includes the sequence of the SIV virus used in the initial infection (inoculum) as well as the original SIV17E-Fr sequence (Genebank #AY033146). Although all three inoculum virus have identical Vpr amino acid sequences there are silent nucleotide changes present in the SHIVs. We found no evidence in our clones for amplification of vpr from either SHIV, indicating that our primers were specific for SIV as was the case for tat(Noel, Jr. et al, 2006;Noel, Jr. & Kumar, 2006). We found no evidence of recombination among the three viral forms as we did for nef(Noel et al, 2006), thus we did not include the SHIVs in our trees. Interestingly, as we observed for tat, the clustering patterns of the trees were more evident in normal progressors (Figure 1, Groups B and C), and in particular with CSF samples as opposed to the plasma(Noel, Jr. et al, 2006). Thus, not only does evolution appear to be inversely correlated with rate of disease progression, but the complexity of the evolution and the ability of a variant to start a persistent evolutionary path appear to be related to slower progression rather than the presence of morphine itself (compare Group A vs. B and C, Figure 1). Recent studies that show no detectable binding or neutralizing antibodies nor virus specific CTL among the rapid progressors, but modest responses by twenty weeks in normal progressors(Kumar et al, 2006), are suggestive of a role for immune pressure in the enhanced compartmentalization we detected for both tat(earlier) and vpr in this work.
Similar to what has been reported for tat and the 5′ part of env(Noel, Jr. et al, 2006;Noel, Jr. & Kumar, 2006;Tirado & Kumar, 2006), the SIV vpr gene had the lowest overall diversity in Group A (Figure 2). The average diversity of all clones was significantly lower in the rapid progressors than in all normal progressors (Figure 2A, 0.31% Group A vs. 0.59% Group B + C, p < 0.05) as well as compared to the normal progressor, morphine dependent group alone (0.31% Group A vs. 0.64% Group B, p < 0.05). Diversity was the lowest in Group A animals in both plasma and CSF when compared only in those compartments (Figure 2B and 2C); however, the trend did not achieve statistical significance.
The vpr divergence in rapid progressors was also significantly less than in all normal progressors (Group A 0.16% vs. Groups B + C 0.34%, p < 0.05, Figure 3A). Further, the divergence was significantly less than either Group B (0.16% vs. 0.35%, p < 0.05, Figure 3A) or Group C (0.16% vs. 0.32%, p < 0.05, Figure 3A) when all samples from both plasma and CSF were considered; as well as versus Group C specifically in the CSF (0.19% vs. 0.39%, p < 0.05, Figure 3C). Divergence in plasma alone was lowest in Group A, but was not statistically less than in normal progressors (Figure 3B).
The relationship between the full array of vpr sequence changes within an individual animal (the total diversity) and the evolution of the viral population from the origin sequence (divergence from the inoculum) was characterized by two measures. First, we compared the ratio between diversity to divergence within each monkey. This ratio was always the greatest (near 2.0) in the rapid progressors, including all samples (Figure 4A), plasma only (4B) and CSF only (4C). In comparisons with rapid progressors, the ratio of diversity to divergence is statistically less for Groups B or C, with the single exception of Group C plasma for which data from one animal, 2/AC42, was unavailable. As a second measure, we assessed the divergence between the plasma and CSF populations in all three groups (Figure 3D). The divergence was significantly greater in both Groups B and C as compared to Group A for all samples (Figure 3D, p < 0.05). This indicates a greater degree of specific selection of virus in these compartments in normal disease progression. Although we did not find evidence of different immune pressure by comparing the rates of synonymous/non-synonymous mutations (data not shown) as was the case for tat previously (Noel, Jr. et al, 2006;Noel, Jr. & Kumar, 2006), the greater divergence between plasma and CSF in Groups B and C suggests a different level of selective pressure, including the possibility of the emergence of a functional immune response. In fact, none of the Group A animals developed detectable virus-specific immunity, while both Group B and C macaques showed evidence of a specific immune response to SIV, including binding and neutralizing antibodies as well as virus-specific CTL(Kumar et al, 2006).
Work reported previously by a number of groups have indicated the functional regions of Vpr responsible for a variety of the biological roles of this conserved lentiviral protein. For example, both the transactivation and cell cycle arrest functions are compromised by a change from arginine to serine at position 73(Sawaya et al, 2000). Cell cycle arrest can also be compromised by alteration of the c-terminus (amino acids 83–89) and at least once such mutations have been linked to long-term non-progression(Wang et al, 1996). We examined the deduced amino acid sequences for all vpr clones with unique, non-synonymous mutations (Figure 5) to determine if we could find evidence of early changes in known functional domains leading to slowed progression. No qualitative differences are readily evident among the three groups in terms of absolute number nor distribution of amino acid substitutions. Furthermore, R73 described previously by Sawaya, et al., is universally conserved in our clones. Interestingly, only monkeys in Groups B and C show any changes between 83–89 amino acids, that may be involved in G2 arrest(Wang et al, 1996), however, these mutants were not a majority for any macaque.
This work examines, for the first time, the evolution of the SIV vpr gene during rapid progression to simian AIDS in the setting of morphine abuse. Much like the SIV tat gene in both plasma and CSF or SIV env in plasma, evolution of vpr shows a significant inverse correlation with disease progression. As with tat, the virus appears to show compartmentalization only when disease progression occurs at a normal rate, regardless of the morphine status. This is supported both in the trees and in a direct comparison of the divergence between plasma and CSF viral populations which was significantly less in the rapid than in the normal progressors. In addition, in this report we provide the first direct comparison of the ratio of diversity to divergence for rapid and normal progressors. In all cases the total population diversity was greater than the average evolution away from the inoculum (divergence); however, the ratio was statistically greater in the rapid progressors, perhaps suggesting that biological selection had begun to promote evolution of particular branches from among the viral quasispecies only in the normal progressors. In fact, only in normal progressors was there any evidence of a modest immune response by 20 weeks post-innoculation (Kumar et al, 2006), indicating that immune selection may have contributed to the reduced diversity to divergence ratio. In contrast to the accessory gene product Nef(Noel et al, 2006), this study provides the first indication that the evolution is related to rapid disease progression during drug abuse for a gene product that is dispensable in vitro.
Previous work with HIV-infected humans has indicated that the evolution of Vpr is related to long-term nonprogression (Cali et al, 2005). Some long-term survivors have shown predominantly intact and functional vpr sequences, while lack of progression in an injecting drug user was associated with subtle changes in nef and the G2 arrest domain in the c-terminus of vpr(Saksena et al, 1996;Zhang et al, 1997). Thus, the role of vpr evolution in slow disease progression in humans may be distinct from that of rapid disease in our model, although the study of an individual drug user may lend a parallel to our system as we carry out analysis beyond 20 weeks. Perhaps a better comparison can be made with studies of the role of vpr sequence changes on HIV transmission. In this case, viral vpr sequences showing greater hetereogeneity proved less infectious during vertical transmission(Yedavalli & Ahmad, 2001), perhaps due to greater likelihood of loss of important functional components of the Vpr protein(Yedavalli et al, 1998). In these studies as well as our own work, greater virulence is associated with greater conservation of the infecting form.
The lack of humoral responses to Vpr have also been implicated in rapid progression(Richardson et al, 2003). Thus the well known roles in immune interference (Muthumani et al, 2002a)and T cell apoptosis (Muthumani et al, 2002b)played by Vpr in vivo may be involved in our system, even though we are looking predominantly at rapid pathogenesis. We do see greatly depleted levels of CD4+ Tcells and, predominantly in the rapid progressors, we have reported a failure to mount any detectable immune response(Kumar et al, 2006). Although, these data do not clearly define the spectrum of amino acid changes that differentiate our rapid from our normal progressors, based on the pathogenicity of the inoculum virus and the more rapid rate of change in the normal progressors, it is possible that the mutations that occur within the first 20 weeks of infection in normal progressors contribute to the slower progression in their respective hosts beyond this time period.
Finally, although we still can not describe the full role of morphine in our system, we do find that only morphine-dependent monkeys progress rapidly, and at a rate of 50%. A picture is emerging that the evolution of some (tat, env, vpr) but not all (nef) viral genes is directly related to the rate of pathogenesis in the setting of drug abuse, but the cause effect relationship is unclear. Morphine has well characterized effects on cells of the immune system, but in spite of years of study, there remains the question of the role of morphine in HIV progression – whether it advances or prolongs the time to (and severity of) AIDS. Our current work does suggest, as with tat and env, that vpr evolution is a potentially important component of early determination of disease progression and rapid pathogenesis in morphine-dependent AIDS.
This study used a previously described rhesus macaque model of drug abuse and AIDS(Kumar et al, 2004a;Noel, Jr. & Kumar, 2006). Nine male monkeys (Macaca mulatta) were divided into morphine-dependent (N=6) or non-morphine controls (N=3). Prior to viral infection, morphine was administered for 18 weeks (5mg/kg, T.I.D.) following a period of morphine introduction with gradual dose increase from 1 to 5mg/kg over two weeks(Kumar et al, 2004a). Morphine administration was maintained throughout the study to avoid withdrawal effects. Infection was by intravenous route with a 2mL inoculum containing 104 50% tissue culture infective doses each of simian-human immunodeficiency virus KU-1B (SHIVKU-1B), (Singh et al, 2002) SHIV89.6P,(Reimann et al, 1996) and SIV 17E-Fr (SIV/17E-Fr)(Flaherty et al, 1997). This three-virus combination produces uniform disease leading to clinical AIDS in a relatively short time(Kumar et al, 2002). In this group of animals, the SIV 17E-Fr virus shows the greatest tissue distribution and generates the predominant immune response(Kumar et al, 2006). All animal protocols were approved by the local animal care committee (IACUC) in accordance with the Guide for the Care and Use of Laboratory Animals.
Viral RNA was extracted from cell-free fluids (plasma and CSF) at approximately 12 and 20 weeks post infection. In cases where samples were unavailable, samples at 14 or 18 weeks served as substitutes. Primers (specific for SIV 17E-Fr) and PCR conditions were as reported previously(Noel, Jr. et al, 2006;Noel, Jr. & Kumar, 2006). Briefly, an RT-PCR reaction (primers F1 5′-GGCAGGGGGATGGAGACCAGG and R1 5′-GCACAAAAAAGGGGAATTGTCGC) was followed by a secondary PCR reaction (primers F2 5′-AAATGAAGGACCACAAAGGGAACC and R2 5′-CCCATAGACACTTAAAAGCAAGATGGC) to generate a 491 base pair DNA fragment. The 5′ end of this product, that encodes the C-terminal 94 amino acids of the SIV Vpr protein, was sequenced after cloning into pPCR2.1 (Invitrogen, Carlsbad, CA). A target of 6–10 positive clones per time point were sent for sequencing to the DNA Sequencing Facility of Florida State University, Department of Biological Sciences. In general, two or more independent PCR/cloning reactions were used to generate sufficient clones. All sequences in this study were deposited into Genebank (DQ839744-DQ840024).
Sequence files (*.ab1 format) were decoded and edited using BioEdit version 7.0.1(Hall, 1999). Alignments were performed using BioEdit after removal of primer sequences and included the 282 nucleotides encoding the final 94 amino acids of SIV Vpr. Amino acid sequences were inferred by translation of the nucleotide alignments using the standard genetic code. Intraindividual diversity for all clones as well as divergence from the inoculum clone were calculated using MEGA3.1(Kumar et al, 2004b). Phylogenetic trees were formed using a neighbor-joining method(Saitou & Nei, 1987). Group diversity and divergence were compared statistically using an unpaired t-test. The statistical cut-off for significance in these analyses was p=0.05.
This work was supported by National Institute on Drug Abuse (DA015013), National Institute for General Medical Sciences (GM008239) and National Institute on Alcohol and Alcoholism (AA015045).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.