Genetic Composition of Human Immunodeficiency Viru
http://www.100md.com
病菌学杂志 2005年第3期
University of California, San Diego, La Jolla
HIV Neurobehavioral Research Center
Veterans Affairs San Diego Healthcare System, San Diego
Applied Biosystems, Foster City, California
Division of Infectious Diseases and Hospital Epidemiology, University Hospital of Zurich, Zurich, Switzerland
Northwestern University, Chicago, Illinois
ABSTRACT
Human immunodeficiency virus (HIV) infection of the central nervous system (CNS) is a significant cause of morbidity. The requirements for HIV adaptation to the CNS for neuropathogenesis and the value of CSF virus as a surrogate for virus activity in brain parenchyma are not well established. We studied 18 HIV-infected subjects, most with advanced immunodeficiency and some neurocognitive impairment but none with evidence of opportunistic infection or malignancy of the CNS. Clonal sequences of C2-V3 env and population sequences of pol from HIV RNA in cerebrospinal fluid (CSF) and plasma were correlated with clinical and virologic variables. Most (14 of 18) subjects had partitioning of C2-V3 sequences according to compartment, and 9 of 13 subjects with drug resistance exhibited discordant resistance patterns between the two compartments. Regression analyses identified three to seven positions in C2-V3 that discriminated CSF from plasma HIV. The presence of compartmental differences at one or more of the identified positions in C2-V3 was highly associated with the presence of discordant resistance (P = 0.007), reflecting the autonomous replication of HIV and the independent evolution of drug resistance in the CNS. Discordance of resistance was associated with severity of neurocognitive deficits (P = 0.07), while low nadir CD4 counts were linked both to the severity of neurocognitive deficits and to discordant resistance patterns (P = 0.05 and 0.09, respectively). These observations support the study of CSF HIV as an accessible surrogate for HIV virions in the brain, confirm the high frequency of discordant resistance in subjects with advanced disease in the absence of opportunistic infection or malignancy of the CNS, and begin to identify genetic patterns in HIV env associated with adaptation to the CNS.
INTRODUCTION
The primate lentiviruses human immunodeficiency virus type 1 (HIV-1) and the simian immunodeficiency virus SIVmac invade the central nervous system (CNS) within days to weeks of primary infection (11, 37, 102), and in the case of HIV result in neurocognitive dysfunction in up to 50% of untreated individuals in late disease (31, 36, 57, 71, 74). Such neurocognitive changes have been linked to postmortem neuropathologic findings of HIV encephalitis (10) and injury to the synaptodendritic apparatus (24, 56). This neuropathology occurs independently of known opportunistic infection (OI) or neurologic malignancy and requires both productive HIV infection and its immunologic sequelae, in particular, the activation and recruitment of macrophages into the CNS (27, 77, 78, 103).
Vigorous but ultimately ineffective host immune responses impel the rapid evolution of HIV-1 in blood and lymphatic tissues (6, 47, 80, 93, 98), but whether similar forces are in play in the CNS is not known. Indeed, the CNS constitutes an immunologically distinct and possibly privileged site that, in the case of other viral pathogens, serves as a chronic sanctuary (12). In addition, HIV predominantly infects microglial cells and perivascular macrophages in the CNS (44, 95), while the principal targets are activated CD4+ lymphocytes in blood and the lymphoid tissues (101). Selective adaptation of replicating HIV populations in the CNS to these cell types might therefore be expected. Thus, findings from earlier studies indicating that HIV populations in the CNS are genetically distinct from virus present in blood or lymphoid tissues (23, 29, 35, 46, 66, 69, 72, 73) are not surprising and may reflect either differences of the principal cellular targets or differences of immune selection. Whether particular sequence motifs define either neurotropism or neurovirulence remains unresolved (29, 46, 66, 73, 79). Observed sequence differences may arise due to genetic drift rather than to increased replicative efficiency or immune selection in the CNS (42).
The potential for differential effects of antiretroviral treatment on CNS and systemic HIV populations both has clinical implications and provides opportunities for the investigation of HIV genetics in the CNS (20, 21, 59, 75, 87, 90). Although many studies using CSF virus as a surrogate for virus in brain parenchyma suggest that highly active antiretroviral therapy (ART) is usually effective over short periods in the CNS (26, 33), the long-term durability of these responses remains unaddressed. Differences in baseline HIV genetics, the longevity of infected microglia, and limited drug penetration justify concern for long-term outcomes. Indeed, studies suggest that neurological manifestations of HIV infection constitute an increasing proportion of AIDS-defining illness in some settings (17, 18, 64), and when drug resistance develops, discordant resistance patterns between CNS and blood virus have been observed in many though not all studies of HIV in brain (59, 99) or cerebrospinal fluid (CSF) (14, 15, 84, 91, 94).
In this study, we examined the genetics of HIV-1 in subjects either failing therapy or not on treatment who underwent paired blood and CSF sampling in the absence of opportunistic neurological disease. Population sequences of pol and clonal sequences of C2-V3 env from these subjects revealed genetically distinct populations in the CSF of most subjects irrespective of antiviral therapy. Analysis by logistic regression identified three to seven codons in C2-V3 that are important as tissue-specific genetic patterns. Subjects harboring these signature patterns in CSF virus were very likely to exhibit discordant patterns of drug resistance in CSF and plasma. These results provide evidence that HIV sequence evolution in the CNS is shaped by compartment-specific selective pressures reflected, in a majority of cases, by genetically discrete CSF HIV populations.
(Presented in part at the 10th Conference on Retroviruses and Opportunistic Infections, Boston, Mass., February 2003, and the Workshop on HIV Genetics and NeuroAIDS, Washington, D.C., December 2002.)
MATERIALS AND METHODS
Subjects. Twenty-one individuals enrolled in longitudinal clinical studies at the HIV Neurobehavioral Research Center between years 1998 and 2002 were initially studied. All subjects had stable or no antiviral therapy for at least 2 months prior to study, had plasma and CSF HIV RNA of >500 copies/ml, and had no evidence of systemic or CNS OIs or malignancy based on clinical, laboratory, and neuro-imaging studies. Data were available on past and present therapy, current HIV RNA and CD4 counts, nadir CD4 counts, and CSF cell counts. All studies were conducted in compliance with local Institutional Review Board guidelines and with subjects' written informed consent.
Specimen processing. Paired blood from peripheral venipuncture in acid citrate dextrose tubes and CSF from lumbar punctures were collected (typically collected within 1 h of each other) and processed within 2 h of collection. Plasma and cell-free CSF were aliquoted, frozen, and stored at –70°C until processing. All subsequent plasma and CSF processing was performed separately to minimize the within-subject cross-contamination of samples.
HIV RNA quantitation. Plasma and CSF HIV RNA levels were measured by the Roche Amplicor Monitor assay (lower limit of detection of 400 copies/ml) or the Ultrasensitive assay (lower limit of detection of 50 copies/ml) with modifications as previously described (21).
NP testing. Neuropsychological (NP) testing was performed within 2 days of CSF sampling for 17 of 18 evaluable subjects except for subject N, whose testing was performed 17 months earlier. Overall NP performance was summarized by calculating global deficit scores (GDS). In brief, the GDS approach quantifies the number and degree of impaired NP test scores across the battery while attaching relatively less significance to performances that are within (or above) normal limits. For each of the NP test variables, raw scores were converted to T-scores (mean, 50; standard deviation, 10) using published demographically corrected normative data. The T-scores were then transformed into deficit scores using the following conversions: >40T = 0; 39T to 35T = 1; 34T to 30T = 2; 29T to 25T = 3; 24T to 20T = 4; and <19T = 5. The deficit scores from each NP test were then averaged to derive a GDS for each participant (range, 0 to 5). Prior research supports the construct validity of the GDS approach as an indicator of global NP functioning in persons with HIV infection and indicates that the GDS is a reliable and valid alternative to clinical ratings and, more importantly, can be generalized across different NP test batteries (9, 36). GDS above 0.5 were interpreted as signifying NP impairment.
Nucleotide sequencing. (i) pol sequencing. pol sequencing was performed using the Viroseq version 2.0 system (Applied Biosystems, Alameda, Calif.) on an ABI 3100 genetic analyzer according to the manufacturer's protocol. In brief, 0.5 to 2.0 ml of cell-free CSF or plasma was thawed and centrifuged at 23,900 x g and 4°C for 1 h to pellet virus. We chose to concentrate virus even though most samples contained sufficient HIV copy numbers for assays and successful amplification without further concentration in order to improved representative sampling. RNA was extracted and resuspended in 50 to 100 μl of diluent according to the protocol. A 10-μl aliquot of resuspended RNA was used in pol sequencing; the remaining 40 μl was used for clonal C2-V3 sequence studies. Raw data from pol sequencing were edited and interpreted using Viroseq HIV genotyping system software version 2.5 (Applied Biosystems, Foster City, Calif.) to generate a genotypic drug resistance profile. The default base-caller for this software identifies a minority HIV population when it is present in at least a 30% proportion.
(ii) env sequencing. Reverse transcription and PCR amplification of C2-V3 env for each sample was performed in triplicate or quadruplicate using the Finnzyme one-step reverse transcription-PCR (RT-PCR) kit (MJ Research, Waltham, Mass.) and primers V3Fout and V3Bout as previously described (32) in a 25-μl reaction volume. A 2.5-μl aliquot of first-step RT-PCR product was used in the second, nested PCR with primers V3Fin and V3Bin (32) and Tris (pH 8.0), 20 pmol of each primer, 2.5 mM MgCl2, 200 μM deoxynucleoside triphosphates, and bovine serum albumin (10 mg/ml) in a total volume of 50 μl. Cycling parameters were 94°C for 2 min; 94°C for 30s, 54°C for 30 s, and 72°C for 1 min for 35 cycles; and 72°C for 10 min. All assays were conducted under conditions to minimize the potential for PCR contamination by utilizing aerosol-resistant pipette tips, dedicated PCR reagents, and laminar flow hoods. All assays included negative controls. Nested PCR products were visualized by agarose gel electrophoresis and ethidium staining. Replicate PCR products were proportionately pooled and cloned using the TOPO-TA cloning system (Invitrogen, Carlsbad, Calif.). Clones were selected by blue-white screening and expanded in 4.0-ml broth cultures. Plasmid clones were purified using QIAGEN Mini-prep kits (QIAGEN). Purified plasmids were sequenced in both directions with –20M13 primer (5'-GTAAAACGACGGCCAG-3') and Topo forward primer (5'-TGGATATCTGCAGAATTCG-3') using Prism Dye terminator kits (ABI) on an ABI 3100 genetic analyzer. Sequences were compiled, aligned, and edited using Sequencher 4.0 (Genecodes, Ann Arbor, Mich.) and Clustal (version 1.81).
Genetic analysis. Phylogenetic analysis was performed using DNAML (Phylip 3.6a; J. Felsenstein) (T/T ratio of 2.0, gamma-distributed substitution rate) on clonal sequences for individual subjects and DNADIST, NEIGHBOR, and CONSENSE for 100 bootstrapped data sets from all sequences (Phylip 3.5; J. Felsenstein). Trees were displayed using Treeview (Roderic DM Page). Assessment of degree of intercompartment segregation was performed by testing for panmixis using gene phylogenies (38, 86) as implemented in MacClade (Sinauer, Sunderland, Mass.). In brief, from the maximum likelihood trees for each individual subject's C2-V3 sequences and their characterization according to compartment of origin, the minimum number of intercompartment migration events allowed by the tree was tallied. This result was compared to the distribution of migration events for 1,000 randomly generated trees. Evidence of restricted gene flow (compartmentalization) was documented when <1% (or between 1 and 5%) of the random trees required the same or a fewer number of migration events as for the sample data (65, 86). Assessment of the role of selection on HIV genetic composition was performed using a Z test for selection with a modified Nei-Gojobori method and p-distances (63) as implemented in MEGA (49). Tests for positive selection, neutral evolution, and purifying selection were performed at sites identified as important for distinguishing CSF from plasma virus.
Analysis of sequence differences between CSF and plasma virus. Prediction of coreceptor usage was based on composition at residues 11 and 25 of the V3 loop (codons 306 and 322 of HXB2 gp120) (25, 85). Residues at which amino acid (AA) frequencies varied significantly between compartments were identified by Monte Carlo approximation of the Mantel-Haenszel and likelihood ratio tests, based on identity of individual AA or on chemical class of AA (positive, negative, neutral, or hydrophobic), with a P value of <0.05 considered significant. Positions with intercompartmental variation by both of these univariate measures were included in a stepwise logistic regression (with entry cutoff P value of <0.10 and retention cutoff P value of <0.05). Forward logistic regression was performed with the entry criterion P value of <0.05. Analyses were performed on the entire data set as well as on "pruned" data, including only those subjects exhibiting phylogenetic segregation of CSF and plasma sequences verified to be significant by the Maddison-Slatkin test (P < 0.01) (86).
Statistical analysis of group characteristics. Fisher's exact test was used to compare proportions between groups. In some cases as noted, parameters were logistically fitted and a whole-model test of significance was applied. Statistical analyses were performed using SAS V8 and JMP IN (SAS Institute, Cary, N.C.).
Nucleotide sequence accession number. The pol and env sequences have been to GenBank and assigned accession numbers AY859056 to AY859091.
RESULTS
PCR amplification for both C2-V3 env and resistance genotyping was successful in 18 of 21 subjects but failed in 3 subjects. HIV RNA concentrations did not appear to predict failure for PCR amplification of either pol or env sequences. The median levels of HIV RNA (log10 copies per milliliter) for subjects whose amplifications failed versus succeeded were similar in plasma (5.29 versus 5.35) and CSF (4.08 versus 4.15). Subject characteristics, treatment history, and NP test scores for the remaining 18 subjects are shown in Table 1. Subjects had relatively advanced disease (median CD4, 128 cells/mm3) and the majority (13 of 18) were neurocognitively impaired as defined by a GDS of >0.5.
C2-V3 env sequences. Most subjects (13 of 18) exhibited partitioning of CSF from plasma C2-V3 sequences (Fig. 1). For 11 subjects, the Maddison-Slatkin test for compartment-based partitioning was highly significant (P < 0.01), and all exhibited clear segregation of the majority of the CSF from plasma sequences on visual inspection of phylogenetic trees. Two other subjects had evidence of partitioning (0.01 < P < 0.05) but appeared to have a greater degree of intermingling of CSF and plasma sequences with only subsets of sequences that were segregated. In 17 of 18 subjects, sequences from CSF and plasma from the same subject clustered more closely to one another than did either CSF-CSF or plasma-plasma sequences between subjects, reflecting the impact of common ancestry of within-subject virus populations on the phylogenetic reconstruction (Fig. 2). In subject Q, the sequences formed two divergent but related clusters. While these sequences were more closely related to one another than to sequences from other subjects, the bootstrap value was marginal (43). Sequences from subject B comprised two very divergent, seemingly unrelated clusters, one having both CSF and plasma sequences and one with only plasma sequences (Fig. 1 and 2). Genetic distance between sequences from these two clusters was approximately 10%. These sequences did not appear to cluster with sequences from any other subjects or with lab strains. Of note, pol sequences from plasma virus in this subject also demonstrated genotypic mixtures at ratios of approximately 50:50 based on peak heights of chromatograms (data not shown). These two cases appear consistent with cocirculation of markedly divergent HIV variants reflecting either replication under very different selective environments or, in the case of subject B, coinfection or superinfection with a second HIV variant (2, 40, 43).
Analysis of differences in env sequence between CSF and plasma virus. Predicted coreceptor usage did not distinguish CSF from plasma viruses in the clonal sequences examined for most subjects. With few exceptions the majority of clones from both CSF and plasma were predicted to be from virus using CCR5 rather than CXCR4, based on the composition of residues 11 and 25 of the V3 loop (codons 306 and 322, respectively, of env) (25, 85) (Fig. 3). Basic residues at position 11 or 25 were found in 3 of 14 plasma and 1 of 13 CSF clones from subject I, 5 of 15 plasma and 0 of 14 CSF clones from subject M, and 11 of 15 plasma and 0 of 14 CSF clones from subject N. Notably, all CSF clones from subject O exhibited an arginine at position 25, but this position was deleted from all 10 plasma clones sequenced. Thus, while we encountered clones predicted to utilize CXCR4 more frequently in plasma virus, this type occurred in only a small number of subjects. CXCR4 using HIV was found occasionally in CSF, consistent with recent reports of its identification from brain parenchyma (66).
Regression analyses were used to verify positions that distinguished CSF from plasma clones for each subject and for all the sequenced clones. Residues 293, 308, and 341 were associated with CSF compartmentalization in the entire sequence set (data not shown). The set of unique sequences in the subset of subjects exhibiting clear partitioning of CSF and plasma sequences (n = 11, pruned subset) were separately analyzed. The P value for evidence of partitioning was <0.01 by both 2 analysis and stepwise logistic regression (Table 2; Fig. 4a). Categorization of amino acids by charge and hydrophobicity (as positive, negative, neutral, or hydrophobic) improved the predictive value of positions 293, 308, and 341 (Table 3; Fig. 4b).
In order to assess whether the sequence differences between plasma and CSF virus were due to genetic drift of isolated HIV populations or resulted from selection, a Z-test using a modified Nei-Gojobori method (p distances) for comparing synonymous and nonsynonymous substitutions at codons 293, 308, and 341 was performed for all CSF sequences and for all plasma sequences. For both CSF and plasma virus, the P values strongly supported positive selection at these sites (P < 0.001 for both). When all seven sites were included, the P value for positive selection for CSF virus increased to 0.07.
Drug resistance. Fourteen of 18 subjects exhibited mutations conferring drug resistance (70), 8 of 9 receiving antiretrovirals and 6 of 9 not on treatment at the time of study (Table 4). Among the eight subjects with drug resistance on ART, five had discordance in resistance patterns between CSF and plasma HIV. Similarly, three of four subjects with past but not current ART had discordance in resistance patterns between CSF and plasma HIV. Two subjects without a known history of drug treatment exhibited drug resistance. Overall, 50% of subjects demonstrated discordant resistance mutations, a proportion comparable to that in previous reports (14, 91).
Twelve of 14 subjects with resistance had mutations associated with resistance to more than one class of antiretroviral drug. Phylogenetic reconstruction of pol sequences showed that CSF and plasma sequences from the same subject were most closely related to one another, which excluded contaminating sequences as an explanation for discordant resistance patterns (data not shown).
Of the nine subjects with discordant drug resistance, resistance mutations were noted in plasma but not in CSF for eight, while only one subject (A) exhibited more resistance in virus from CSF than from plasma. This pattern suggests lesser selective pressure by ART or a smaller population of preexisting HIV with mutations for resistance in the CNS compartment. In several cases of concordant multidrug resistance, the patterns of resistance were extremely complex but were nonetheless identical between CSF and plasma virus, arguing for the emergence of resistance as single rather than parallel events in these cases.
Poor penetration of individual antiviral drugs into the CNS could favor the development of discordant resistance for some drugs. We could not systematically address this issue, since the study subjects were treated with a variety of different antiretroviral regimens. Nevertheless, the frequency of discordance for resistance mutations to nucleoside reverse transcriptase inhibitors, a class with good CNS penetration (26, 82), was no less than for resistance mutations to protease inhibitors, which generally penetrate poorly (3, 28, 50, 52) (Table 4).
Correlates of drug resistance patterns and partitioning of C2-V3 sequence. In subjects with drug-resistant virus, concordant resistance patterns were associated with lack of partitioning between CSF and plasma env sequences, reflecting similar HIV populations in both compartments (P = 0.035) (Table 5). This observation is compatible with selection of drug resistance in the HIV population in one compartment, which then spreads to the other perhaps because of mixing of the two populations. Conversely, discordance of resistance patterns correlated with envelope partitioning (P = 0.035) and with the presence of differences at signature positions in env as summarized in Table 2 (P = 0.007), consistent with the independent evolution of resistance in segregated and autonomously replicating HIV populations. Discordant resistance tended to be more common in subjects with lower nadir CD4 counts, as suggested by Cunningham and colleagues (14).
Treatment history tended to correlate with segregation of CSF and plasma env sequence clusters (P = 0.09) (Table 5), suggesting an effect of treatment on the composition of HIV env genotypes; however, there were relatively few subjects who were drug na?ve (n = 5) (Table 1). The apparent association of treatment and env sequence partitioning may have reflected a prestudy bias towards treating subjects with more advanced disease. In this small study population, CSF pleiocytosis was not significantly associated with resistance patterns or with the observed partitioning of C2-V3 sequences (data not shown).
Correlates of neurocognitive dysfunction. Nadir but not current CD4 count was correlated with severity of neurocognitive impairment (P = 0.05) (Table 5). A whole-model test for the logistic fit of nadir CD4 and presence or absence of neurocognitive impairment also demonstrated a significant association (P = 0.04). These correlations are consistent with the increasing prevalence of neurocognitive dysfunction with advancing disease. Of note, partitioning of blood and CSF C2-V3 sequences was not significantly associated with neurocognitive abnormalities. However, subjects with concordant resistance or no resistance had a lower GDS than those with discordant resistance (median GDS = 0.8 for concordant or no resistance, median GDS = 1.6 for discordant resistance; P = 0.07). Thus, more advanced disease and the presence of discordant resistance were positively associated with the presence and severity of neurocognitive dysfunction.
DISCUSSION
Infection of the CNS is an important contributor to the morbidity of HIV disease, but study of the in vivo biology of CNS infection by HIV is complicated by the relative inaccessibility of tissues of the CNS. Because of the impracticality of sampling the brain for the antemortem study of HIV genetics, many investigators have examined virus in CSF as a surrogate marker for parenchymal disease (8, 22, 55, 58, 76, 87). Whether HIV RNA in CSF accurately reflects the predominant HIV species replicating in brain parenchyma is unknown, but HIV RNA concentrations in brain and CSF appear to be correlated (58, 96) and the predictive value of CSF HIV RNA but not HIV RNA in blood for neurocognitive impairment suggests a significant relationship (8, 21, 58). In addition to the relative ease of sampling CSF, the study of CSF virus provides additional theoretical advantages. Whereas parenchymal DNA can disproportionately reflect archival and in some cases replication-defective virus, RNA in CSF has been shown to turn over relatively rapidly (20, 87), reflecting the kinetics of recently formed virions, in a manner analogous to the differences observed between plasma virus and HIV DNA in peripheral blood mononuclear cells (34, 61, 88). Relatively few HIV RNA genetic studies of postmortem brain specimens have been reported, possibly due to the lability of RNA and their degradation during the postmortem period. However, in the SIV-infected macaque model, DNA and RNA comprised distinct sequence variants when recovered from brain and isolated microglia of freshly euthanized animals (4). Underscoring this particular advantage of CSF RNA sequences, only 3 of the 225 CSF RNA clones we sequenced contained frameshifts or nonsense mutations compared with high frequencies of defective forms reported from prior studies of HIV DNA sequence in brain parenchyma (53, 54, 68). Thus, genetic studies of CSF HIV in subjects with relatively advanced disease complement HIV studies using brain samples from autopsies or biopsies (29, 39, 62) by providing only recently produced virions. We studied cell-free CSF samples to minimize effects from sampling of HIV sequences from lymphocytes that may traffic into the CSF. This approach may account in part for the relatively high proportion of samples demonstrating segregation of HIV populations.
Genetic studies comparing HIV from the CNS with that in blood or lymphoid organs have identified sequence differences in env, pol, gag, nef, tat, and the long terminal repeat (1, 5, 7, 13, 42, 46, 62, 66, 72, 81, 83, 99). The greatest attention has been paid to env sequences, in particular those corresponding to the V3 loop, an important (though not exclusive) determinant of both cell tropism and neutralization susceptibility by antibody (45). The reported differences could result from genetic drift of segregated populations or from selection by immune or other factors within the CNS. Our findings that C2-V3 sequences in CSF differed from those in plasma therefore corroborate many earlier studies that have catalogued genetic differences between virus in CNS and virus in blood or lymphoid tissues. However, while some earlier studies have suggested the existence of env genotypes associated with neurotropism and neurovirulence (46, 48, 73), others have disputed this (16, 66, 79). An aim of the present study was to revisit this question.
Recently proposed models for the origin of CSF virus implicate an increasing contribution of parenchymal CNS sources over peripheral lymphoid sources with advancing disease (20, 21, 87, 90). Our results are consistent with such a view, since we found a trend towards more segregation between CSF and plasma sequences in individuals with advanced disease. Discordant resistance was encountered more frequently for individuals with a nadir CD4 of less than 200/mm3, although this association did not reach statistical significance, owing perhaps to the small number of individuals studied with CD4 counts of >200 mm3.
Anticipating that HIV from both systemic and CNS sources might be sampled from CSF in any given subject, we sought a CNS signature sequence in C2-V3 from both the entire data set as well as from a pruned data set by using phylogenetic criteria to exclude those individuals likely to have CSF virus produced by systemic sources. We did not find a single exclusive signature pattern that predicted the CNS origin of a virus. However, regression analyses identified a small number of positions that appeared to be important for discriminating HIV populations from the two compartments. In related work, construction of a decision tree based on the physico-chemical characteristics of AA at six positions permitted the proper classification of up to 90% of sequences when evaluating only the pruned data (M. C. Strain, J. K. Wong and S. Pillai, unpublished data). Of note, position 308 is identical to one of two positions implicated by Power and colleagues (referred to as position 305 of the macrophage-tropic HIV consensus) to be important in determining neurovirulence (73) and to one of several positions in the V3 loop identified by Korber and colleagues as possible CNS signatures (46). Together, these findings suggest that the HIV phenotype(s) that confers neurotropism is genetically complex, and although the phenotype(s) may not be described by simple signature AA substitutions, it is nevertheless associated with discernible and predictable covariation. Further, the finding that codons 293, 308, and 341 are under positive selection and can be used to classify sequences by anatomic source based either on specific AA or physico-chemical characteristics of the AA provides compelling evidence for effects of selective forces generic to the CNS acting on these sites in the HIV envelope.
Despite the identification of residue 308 in this study as an important CNS-defining position and the implication of the same position by Powers and colleagues as a position conferring neurovirulence, we did not observe any evidence that CNS signature mutations correlated with cognitive abnormalities. Thus, we interpret these patterns as indicative of HIV neurotropism but not necessarily neurovirulence. However, such determinants of neurovirulence might be outside of the C2-V3 domain (66, 100) or might be found only in minority HIV populations that infect small but functionally critical cellular targets, such as astrocytes (60, 67, 97) and microvascular endothelial cells (89, 97). In the macaque model of neuro-AIDS, adaptation of SIV by serial animal passage in microglial cells appears to confer a phenotype that reliably produces SIV encephalitis (51, 92), providing further evidence that either or both neurotropism and neurovirulence of primate lentiviruses are selectable traits.
The seven signature positions identified in this study do not include either of the canonical positions in the V3 loop that help to define coreceptor usage (25, 85). This observation is consistent with others indicating that coreceptor usage does not discriminate HIV populations in the CNS from systemic HIV (30, 66). Phenotypic studies were not performed as a part of the present study, and such studies will be necessary to define the biologic requirements for neurotropism. Potential phenotypic characteristics for CNS adaptation might include such features as a lower requirement for primary and coreceptor density, as supported by in vitro models (19) and in vivo studies on small numbers of subjects (30). While one or more of the residues that we identified here might confer such a phenotype, they may also interact with sites outside of V3 and play indirect roles.
Segregated, CNS-adapted HIV populations in the CSF exhibited distinct resistance patterns. The high frequency of discordant resistance in our subjects, all of whom had no concurrent OIs, confirms recent results from subjects who underwent CSF sampling because of suspected and documented CNS OIs (14, 91). Concordance of resistance patterns was associated with similar V3 sequences in CSF and plasma virus, suggesting either that CNS and systemic HIV populations were not segregated or a resistant HIV variant had replaced wild-type virus in both compartments in the absence of recombination between env and pol coding regions. In contrast, we observed both discordant resistance patterns and the simultaneous partitioning of V3 sequences in the majority of cases. These cases are consistent with either the independent emergence of resistance in isolated HIV populations or the emergence of resistance in one compartment that then spreads to the other with the subsequent divergence of HIV sequences (in such cases, possibly aided by recombination). However, without longitudinal data, these possibilities cannot be distinguished.
We encountered discordant resistance both in individuals receiving failing ART at the time of sampling as well as in individuals off therapy at the time of study. This suggests that HIV populations may evolve independently during both the selection of resistance on therapy and reversion to wild-type off therapy. One subject harbored a substantially more resistant virus in CSF off therapy, while the plasma virus had apparently reverted to a less resistant sequence type. In such a scenario, treatment selection based on resistance testing of plasma virus alone would have failed to account for this highly resistant, actively replicating HIV population.
This and other recent studies of HIV in CSF reveal a high frequency of discordant resistance compared with earlier studies (84, 94). Possible explanations for these discrepant findings are differences in the complexity and composition of antiretroviral regimens the study subjects received. Whereas many earlier studies had focused on resistance to zidovudine, a drug with reasonably good CSF penetration, many contemporary regimens include drugs with marginal CSF penetration, such as the protease inhibitors (3, 28, 50, 52). Heterogeneity of drug concentrations has been predicted to promote the emergence of resistance (41) and may also explain the maintenance of populations of virus with differing resistance patterns in this and other recent studies (14, 91). Finally, the presence of discordant drug resistance was associated with more severe neurocognitive dysfunction as reflected by higher GDS. This trend demonstrated only borderline statistical significance due to the relatively small sample size but is important to note, since subjects with more severe neurocognitive deficits for whom control of virus replication in CNS would be more clinically important appear more likely to harbor virus in the CNS that has a resistance profile discordant from that of blood virus.
In summary, our results demonstrate that CSF HIV appears genetically distinct from contemporaneous plasma HIV in a majority of individuals with advanced disease, confirming both compartmental segregation and autonomous replication within the CNS. Furthermore, a small number of signature positions in and near the V3 loop may reflect genetic adaptation to replication in the CNS. An important correlate of independent V3 genotypes was the frequent observation of discordant resistance patterns between virus in plasma and CSF. These findings may have important implications for drug treatment and subject monitoring. For example, drug failure based on plasma HIV resistance pattern might not mean that CSF HIV has also developed resistance. Continuation of drugs that were previously effective may be indicated for CNS protection, even as new agents are added to address drug-resistant plasma virus. Clearly, establishment of the clinical significance of such discordant resistance patterns will need to await results of detailed longitudinal studies. Nevertheless, these results help to substantiate the value of CSF sampling in studies aimed at elucidating HIV neuropathogenesis and guiding neuro-AIDS treatment.
ACKNOWLEDGMENTS
We are grateful to Sharon Wilcox, Roma Sysyn, Melanie Sherman, and Darica Smith for administrative and technical support, to Clayton Wiley for helpful comments on the manuscript, and to the participating subjects from the UCSD HNRC.
This study was supported by a VA Research Merit award and NIH grants AI 43752 (J.K.W.); MH 58076 (J.K.W., S.L., R.E., I.G., and J.A.M.); MH 62512 (I.G., R.E., T.M., S.L., J.A.M., J.M.-B., and J.D.); MH 59745 (I.G., R.E., T.M., and J.A.M.); DA12065 (I.G., R.E., T.M., and J.A.M.); AI 27670, 38858, 29164, 47745, and 36214 (D.D.R.); an NSF grant (M.S.); and Swiss National Science Foundation grant 3345-65168 (H.F.G.).
Supplemental material for this article may be found at http://jvi.asm.org/.
REFERENCES
Ait-Khaled, M., J. E. McLaughlin, M. A. Johnson, and V. C. Emery. 1995. Distinct HIV-1 long terminal repeat quasispecies present in nervous tissues compared to that in lung, blood and lymphoid tissues of an AIDS patient. AIDS 9:675-683.
Altfeld, M., T. M. Allen, X. G. Yu, M. N. Johnston, D. Agrawal, B. T. Korber, D. C. Montefiori, D. H. O'Connor, B. T. Davis, P. K. Lee, E. L. Maier, J. Harlow, P. J. Goulder, C. Brander, E. S. Rosenberg, and B. D. Walker. 2002. HIV-1 superinfection despite broad CD8+ T-cell responses containing replication of the primary virus. Nature 420:434-439.
Aweeka, F., A. Jayewardene, S. Staprans, S. E. Bellibas, B. Kearney, P. Lizak, T. Novakovic-Agopian, and R. W. Price. 1999. Failure to detect nelfinavir in the cerebrospinal fluid of HIV-1-infected patients with and without AIDS dementia complex. J. Acquir. Immune Defic. Syndr. Hum. Retrovir. 20:39-43.
Babas, T., D. Munoz, J. L. Mankowski, P. M. Tarwater, J. E. Clements, and M. C. Zink. 2003. Role of microglial cells in selective replication of simian immunodeficiency virus genotypes in the brain. J. Virol. 77:208-216.
Blumberg, B. M., L. G. Epstein, Y. Saito, D. Chen, L. R. Sharer, and R. Anand. 1992. Human immunodeficiency virus type 1 nef quasispecies in pathological tissue. J. Virol. 66:5256-5264.
Borrow, P., H. Lewicki, X. Wei, M. S. Horwitz, N. Peffer, H. Meyers, J. A. Nelson, J. E. Gairin, B. H. Hahn, M. B. Oldstone, and G. M. Shaw. 1997. Antiviral pressure exerted by HIV-1-specific cytotoxic T lymphocytes (CTLs) during primary infection demonstrated by rapid selection of CTL escape virus. Nat. Med. 3:205-211.
Bratanich, A. C., C. Liu, J. C. McArthur, T. Fudyk, J. D. Glass, S. Mittoo, G. A. Klassen, and C. Power. 1998. Brain-derived HIV-1 tat sequences from AIDS patients with dementia show increased molecular heterogeneity. J. Neurovirol. 4:387-393.
Brew, B. J., L. Pemberton, P. Cunningham, and M. G. Law. 1997. Levels of human immunodeficiency virus type 1 RNA in cerebrospinal fluid correlate with AIDS dementia stage. J. Infect. Dis. 175:963-966.
Carey, C. L., S. P. Woods, R. Gonzales, E. Conover, T. D. Marcotte, I. Grant, R. K. Heaton, et al. 2004. Predictive validity of global deficit scores in detecting neuropsychological impairment in HIV infection. J. Clin. Exp. Neuropsychol. 26:307-319.
Cherner, M., E. Masliah, R. J. Ellis, T. D. Marcotte, D. J. Moore, I. Grant, and R. K. Heaton. 2002. Neurocognitive dysfunction predicts postmortem findings of HIV encephalitis. Neurology 59:1563-1567.
Clements, J. E., T. Babas, J. L. Mankowski, K. Suryanarayana, M. Piatak, Jr., P. M. Tarwater, J. D. Lifson, and M. C. Zink. 2002. The central nervous system as a reservoir for simian immunodeficiency virus (SIV): steady-state levels of SIV DNA in brain from acute through asymptomatic infection. J. Infect. Dis. 186:905-913.
Cook, M. L., V. B. Bastone, and J. G. Stevens. 1974. Evidence that neurons harbor latent herpes simplex virus. Infect. Immun. 9:946-951.
Corboy, J. R., and P. J. Garl. 1997. HIV-1 LTR DNA sequence variation in brain-derived isolates. J. Neurovirol. 3:331-341.
Cunningham, P. H., D. G. Smith, C. Satchell, D. A. Cooper, and B. Brew. 2000. Evidence for independent development of resistance to HIV-1 reverse transcriptase inhibitors in the CSF. AIDS 14:1949-1954.
Di Stefano, M., F. Sabri, T. Leitner, B. Svennerholm, L. Hagberg, G. Norkrans, and F. Chiodi. 1995. Reverse transcriptase sequence of paired isolates of cerebrospinal fluid and blood from patients infected with human immunodeficiency virus type 1 during zidovudine treatment. J. Clin. Microbiol. 33:352-355.
Di Stefano, M., S. Wilt, F. Gray, M. Dubois-Dalcq, and F. Chiodi. 1996. HIV type 1 V3 sequences and the development of dementia during AIDS. AIDS Res. Hum. Retrovir. 12:471-476.
Dore, G. J., P. K. Correll, Y. Li, J. M. Kaldor, D. A. Cooper, and B. J. Brew. 1999. Changes to AIDS dementia complex in the era of highly active antiretroviral therapy. AIDS 13:1249-1253.
Dore, G. J., A. McDonald, Y. Li, J. M. Kaldor, and B. J. Brew. 2003. Marked improvement in survival following AIDS dementia complex in the era of highly active antiretroviral therapy. AIDS 17:1539-1545.
Edinger, A. L., J. L. Mankowski, B. J. Doranz, B. J. Margulies, B. Lee, J. Rucker, M. Sharron, T. L. Hoffman, J. F. Berson, M. C. Zink, V. M. Hirsch, J. E. Clements, and R. W. Doms. 1997. CD4-independent, CCR5-dependent infection of brain capillary endothelial cells by a neurovirulent simian immunodeficiency virus strain. Proc. Natl. Acad. Sci. USA 94:14742-14747.
Ellis, R. J., A. C. Gamst, E. Capparelli, S. A. Spector, K. Hsia, T. Wolfson, I. Abramson, I. Grant, and J. A. McCutchan. 2000. Cerebrospinal fluid HIV RNA originates from both local CNS and systemic sources. Neurology 54:927-936.
Ellis, R. J., K. Hsia, S. A. Spector, J. A. Nelson, R. K. Heaton, M. R. Wallace, I. Abramson, J. H. Atkinson, I. Grant, J. A. McCutchan, et al. 1997. Cerebrospinal fluid human immunodeficiency virus type 1 RNA levels are elevated in neurocognitively impaired individuals with acquired immunodeficiency syndrome. Ann. Neurol. 42:679-688.
Ellis, R. J., D. J. Moore, M. E. Childers, S. Letendre, J. A. McCutchan, T. Wolfson, S. A. Spector, K. Hsia, R. K. Heaton, and I. Grant. 2002. Progression to neuropsychological impairment in human immunodeficiency virus infection predicted by elevated cerebrospinal fluid levels of human immunodeficiency virus RNA. Arch. Neurol. 59:923-928.
Epstein, L. G., C. Kuiken, B. M. Blumberg, S. Hartman, L. R. Sharer, M. Clement, and J. Goudsmit. 1991. HIV-1 V3 domain variation in brain and spleen of children with AIDS: tissue-specific evolution within host-determined quasispecies. Virology 180:583-590.
Everall, I. P., R. K. Heaton, T. D. Marcotte, R. J. Ellis, J. A. McCutchan, J. H. Atkinson, I. Grant, M. Mallory, E. Masliah, et al. 1999. Cortical synaptic density is reduced in mild to moderate human immunodeficiency virus neurocognitive disorder. Brain Pathol. 9:209-217.
Fouchier, R. A. M., M. Groenink, N. A. Kootstra, M. Tersmette, H. G. Huisman, F. Miedema, and H. Schuitemaker. 1992. Phenotype-associated sequence variation in the third variable domain of the human immunodeficiency virus type 1 gp120 molecule. J. Virol. 66:3183-3187.
Foudraine, N. A., R. M. Hoetelmans, J. M. Lange, F. De Wolf, B. H. van Benthem, J. J. Maas, I. P. Keet, and P. Portegies. 1998. Cerebrospinal-fluid HIV-1 RNA and drug concentrations after treatment with lamivudine plus zidovudine or stavudine. Lancet 351:1547-1551.
Gendelman, H. E., S. A. Lipton, M. Tardieu, M. I. Bukrinsky, and H. S. L. M. Nottet. 1994. The neuropathogenesis of HIV-1 infection. J. Leukoc. Biol. 56:389-398.
Gisolf, E. H., R. H. Enting, S. Jurriaans, F. de Wolf, M. E. van der Ende, R. M. Hoetelmans, P. Portegies, and S. A. Danner. 2000. Cerebrospinal fluid HIV-1 RNA during treatment with ritonavir/saquinavir or ritonavir/saquinavir/stavudine. AIDS 14:1583-1589.
Gorry, P. R., G. Bristol, J. A. Zack, K. Ritola, R. Swanstrom, C. J. Birch, J. E. Bell, N. Bannert, K. Crawford, H. Wang, D. Schols, E. De Clercq, K. Kunstman, S. M. Wolinsky, and D. Gabuzda. 2001. Macrophage tropism of human immunodeficiency virus type 1 isolates from brain and lymphoid tissues predicts neurotropism independent of coreceptor specificity. J. Virol. 75:10073-10089.
Gorry, P. R., J. Taylor, G. H. Holm, A. Mehle, T. Morgan, M. Cayabyab, M. Farzan, H. Wang, J. E. Bell, K. Kunstman, J. P. Moore, S. M. Wolinsky, and D. Gabuzda. 2002. Increased CCR5 affinity and reduced CCR5/CD4 dependence of a neurovirulent primary human immunodeficiency virus type 1 isolate. J. Virol. 76:6277-6292.
Grant, I., J. H. Atkinson, J. R. Hesselink, C. J. Kennedy, D. D. Richman, S. A. Spector, and J. A. McCutchan. 1987. Evidence for early central nervous system involvement in the acquired immunodeficiency syndrome (AIDS) and other human immunodeficiency virus (HIV) infections. Ann. Intern. Med. 107:828-836.
Gunthard, H. F., S. D. Frost, A. J. Leigh-Brown, C. C. Ignacio, K. Kee, A. S. Perelson, C. A. Spina, D. V. Havlir, M. Hezareh, D. J. Looney, D. D. Richman, and J. K. Wong. 1999. Evolution of envelope sequences of human immunodeficiency virus type 1 in cellular reservoirs in the setting of potent antiviral therapy. J. Virol. 73:9404-9412.
Gunthard, H. F., D. V. Havlir, S. Fiscus, Z. Q. Zhang, J. Eron, J. Mellors, R. Gulick, S. D. F. Frost, A. J. Leigh-Brown, W. Schlief, F. Valentine, L. Jonas, A. Meibohm, C. Ignacio, R. Isaacs, R. Gamagami, E. Emini, A. T. Haase, D. D. Richman, and J. K. Wong. 2001. Residual HIV RNA and DNA in lymph node and HIV RNA in genital secretions and in CSF after two years of suppression of viremia in the Merck 035 cohort. J. Infect. Dis. 183:1318-1327.
Gunthard, H. F., J. K. Wong, C. C. Ignacio, J. C. Guatelli, N. L. Riggs, D. Havlir, and D. D. Richman. 1998. Human immunodeficiency virus replication and genotypic resistance in blood and lymph nodes after a year of potent antiretroviral therapy. J. Virol. 72:2422-2428.
Haggerty, S., and M. Stevenson. 1991. Predominance of distinct viral genotypes in brain and lymph node compartments of HIV-1-infected individuals. Viral Immunol. 4:123-131.
Heaton, R. K., I. Grant, N. Butters, D. A. White, D. Kirson, J. H. Atkinson, J. A. McCutchan, M. J. Taylor, M. D. Kelly, R. J. Ellis, et al. 1995. The HNRC 500—neuropsychology of HIV infection at different disease stages. J. Int. Neuropsychol. Soc. 1:231-251.
Ho, D. D., T. R. Rota, R. T. Schooley, J. C. Kaplan, J. D. Allan, J. E. Groopman, L. Resnick, D. Felsenstein, C. A. Andrews, and M. S. Hirsch. 1985. Isolation of HTLV-III from cerebrospinal fluid and neural tissues of patients with neurologic syndromes related to the acquired immunodeficiency syndrome. N. Engl. J. Med. 313:1493-1497.
Hudson, R. R., M. Slatkin, and W. P. Maddison. 1992. Estimation of levels of gene flow from DNA sequence data. Genetics 132:583-589.
Hughes, E. S., J. E. Bell, and P. Simmonds. 1997. Investigation of the dynamics of the spread of human immunodeficiency virus to brain and other tissues by evolutionary analysis of sequences from the p17gag and env genes. J. Virol. 71:1272-1280.
Jost, S., M. C. Bernard, L. Kaiser, S. Yerly, B. Hirschel, A. Samri, B. Autran, L. E. Goh, and L. Perrin. 2002. A patient with HIV-1 superinfection. N. Engl. J. Med. 347:731-736.
Kepler, T. B., and A. S. Perelson. 1998. Drug concentration heterogeneity facilitates the evolution of drug resistance. Proc. Natl. Acad. Sci. USA 95:11514-11519.
Keys, B., J. Karis, B. Fadeel, A. Valentin, G. Norkrans, L. Hagberg, and F. Chiodi. 1993. V3 sequences of paired HIV-1 isolates from blood and cerebrospinal fluid cluster according to host and show variation related to the clinical stage of disease. Virology 196:475-483.
Koelsch, K. K., D. M. Smith, S. J. Little, C. C. Ignacio, T. R. Macaranas, A. J. Brown, C. J. Petropoulos, D. D. Richman, and J. K. Wong. 2003. Clade B HIV-1 superinfection with wild-type virus after primary infection with drug-resistant clade B virus. AIDS 17:F11-F16.
Koenig, S., H. E. Gendelman, J. M. Orenstein, M. C. Dal Canto, G. H. Pezeshkpour, M. Yungbluth, F. Janotta, A. Aksamit, M. A. Martin, and A. S. Fauci. 1986. Detection of AIDS virus in macrophages in brain tissue from AIDS patients with encephalopathy. Science 233:1089-1093.
Korber, B. T., E. E. Allen, A. D. Farmer, and G. Myers. 1995. Heterogeneity of HIV-1 and HIV-2. AIDS 9(Suppl. A):S5-S18.
Korber, B. T., K. J. Kunstman, B. K. Patterson, M. Furtado, M. M. McEvilly, R. Levy, and S. M. Wolinsky. 1994. Genetic differences between blood- and brain-derived viral sequences from human immunodeficiency virus type 1-infected patients: evidence of conserved elements in the V3 region of the envelope protein of brain-derived sequences. J. Virol. 68:7467-7481.
Koup, R. A. 1994. Virus escape from CTL recognition. J. Exp. Med. 180:779-782.
Kuiken, C. L., J. Goudsmit, G. F. Weiller, J. S. Armstrong, S. Hartman, P. Portegies, J. Dekker, and M. Cornelissen. 1995. Differences in human immunodeficiency virus type 1 V3 sequences from patients with and without AIDS dementia complex. J. Gen. Virol. 76:175-180.
Kumar, S., K. Tamura, I. B. Jakobsen, and M. Nei. 2001. MEGA2: molecular evolutionary genetics analysis software. Bioinformatics 17:1244-1245.
Lafeuillade, A., C. Solas, P. Halfon, S. Chadapaud, G. Hittinger, and B. Lacarelle. 2002. Differences in the detection of three HIV-1 protease inhibitors in non-blood compartments: clinical correlations. HIV Clin. Trials 3:27-35.
Lane, T. E., M. J. Buchmeier, D. D. Watry, D. B. Jakubowski, and H. S. Fox. 1995. Serial passage of microglial SIV results in selection of homogeneous env quasispecies in the brain. Virology 212:458-465.
Letendre, S. L., E. V. Capparelli, R. J. Ellis, J. A. McCutchan, and the HIV Neurobehavioral Research Center Group. 2000. Indinavir population pharmacokinetics in plasma and cerebrospinal fluid. Antimicrob. Agents Chemother. 44:2173-2175.
Li, Y., J. C. Kappes, J. A. Conway, R. W. Price, G. M. Shaw, and B. H. Hahn. 1991. Molecular characterization of human immunodeficiency virus type 1 cloned directly from uncultured human brain tissue: identification of replication-competent and -defective viral genomes. J. Virol. 65:3973-3985.
Liu, Y., X. P. Tang, J. C. McArthur, J. Scott, and S. Gartner. 2000. Analysis of human immunodeficiency virus type 1 gp160 sequences from a patient with HIV dementia: evidence for monocyte trafficking into brain. J. Neurovirol. 6(Suppl. 1):S70-S81.
Marra, C. M., D. Lockhart, J. R. Zunt, M. Perrin, R. W. Coombs, and A. C. Collier. 2003. Changes in CSF and plasma HIV-1 RNA and cognition after starting potent antiretroviral therapy. Neurology 60:1388-1390.
Masliah, E., R. K. Heaton, T. D. Marcotte, R. J. Ellis, C. A. Wiley, M. Mallory, C. L. Achim, J. A. McCutchan, J. A. Nelson, J. H. Atkinson, I. Grant, et al. 1997. Dendritic injury is a pathological substrate for human immunodeficiency virus-related cognitive disorders. Ann. Neurol. 42:963-972.
McArthur, J. C., N. Haughey, S. Gartner, K. Conant, C. Pardo, A. Nath, and N. Sacktor. 2003. Human immunodeficiency virus-associated dementia: an evolving disease. J. Neurovirol. 9:205-221.
McArthur, J. C., D. R. McClernon, M. F. Cronin, T. E. Nance-Sproson, A. J. Saah, M. St. Clair, and E. R. Lanier. 1997. Relationship between human immunodeficiency virus-associated dementia and viral load in cerebrospinal fluid and brain. Ann. Neurol. 42:689-698.
McClernon, D. R., R. Lanier, S. Gartner, P. Feaser, C. A. Pardo, M. St. Clair, Q. Liao, and J. C. McArthur. 2001. HIV in the brain: RNA levels and patterns of zidovudine resistance. Neurology 57:1396-1401.
Messam, C. A., and E. O. Major. 2000. Stages of restricted HIV-1 infection in astrocyte cultures derived from human fetal brain tissue. J. Neurovirol. 6(Suppl. 1):S90-S94.
Michael, N. L., G. Chang, P. K. Ehrenberg, M. T. Vahey, and R. R. Redfield. 1993. HIV-1 proviral genotypes from the peripheral blood mononuclear cells of an infected patient are differentially represented in expressed sequences. J. Acquir. Immune Defic. Syndr. 6:1073-1085.
Morris, A., M. Marsden, K. Halcrow, E. S. Hughes, R. P. Brettle, J. E. Bell, and P. Simmonds. 1999. Mosaic structure of the human immunodeficiency virus type 1 genome infecting lymphoid cells and the brain: evidence for frequent in vivo recombination events in the evolution of regional populations. J. Virol. 73:8720-8731.
Nei, M., and S. Kumar. 2000. Molecular evolution and phylogenetics. Oxford University Press, New York, N.Y.
Neuenburg, J. K., H. R. Brodt, B. G. Herndier, M. Bickel, P. Bacchetti, R. W. Price, R. M. Grant, and W. Schlote. 2002. HIV-related neuropathology, 1985 to 1999: rising prevalence of HIV encephalopathy in the era of highly active antiretroviral therapy. J. Acquir. Immune Defic. Syndr. 31:171-177.
Nickle, D. C., D. Shriner, J. E. Mittler, L. M. Frenkel, and J. I. Mullins. 2003. Importance and detection of virus reservoirs and compartments of HIV infection. Curr. Opin. Microbiol. 6:410-416.
Ohagen, A., A. Devitt, K. J. Kunstman, P. R. Gorry, P. P. Rose, B. Korber, J. Taylor, R. Levy, R. L. Murphy, S. M. Wolinsky, and D. Gabuzda. 2003. Genetic and functional analysis of full-length human immunodeficiency virus type 1 env genes derived from brain and blood of patients with AIDS. J. Virol. 77:12336-12345.
Overholser, E. D., G. D. Coleman, J. L. Bennett, R. J. Casaday, M. C. Zink, S. A. Barber, and J. E. Clements. 2003. Expression of simian immunodeficiency virus (SIV) nef in astrocytes during acute and terminal infection and requirement of nef for optimal replication of neurovirulent SIV in vitro. J. Virol. 77:6855-6866.
Pang, S., Y. Koyanagi, S. Miles, C. A. Wiley, H. V. Vinters, and I. S. Y. Chen. 1990. High levels of unintegrated HIV-1 DNA in brain tissue of AIDS dementia patients. Nature 343:85-89.
Pang, S., H. V. Vinters, T. Akashi, W. A. O'Brien, and I. S. Y. Chen. 1991. HIV-1 env sequence variation in brain tissue of patients with AIDS-related neurologic disease. J. Acquir. Immune Defic. Syndr. 4:1082-1092.
Parikh, U., C. Calef, B. Larder, R. F. Schinazi, and J. Mellors. 2001. Mutations in retroviral genes associated with drug resistance, p. 191-278. In C. Kuiken, B. Foley, B. Hahn, P. Marx, F. McCutchan, J. Mellors, S. Wolinsky, and B. Korber (ed.), HIV sequence compendium 2001. Theoretical Biology and Biophysics Group, Los Alamos National Laboratory, Los Alamos, N.Mex.
Pierson, T., J. McArthur, and R. F. Siliciano. 2000. Reservoirs for HIV-1: mechanisms for viral persistence in the presence of antiviral immune responses and antiretroviral therapy. Annu. Rev. Immunol. 18:665-708.
Power, C., J. C. McArthur, R. T. Johnson, D. E. Griffin, J. D. Glass, R. Dewey, and B. Chesebro. 1995. Distinct HIV-1 env sequences are associated with neurotropism and neurovirulence. Curr. Top. Microbiol. Immunol. 202:89-104.
Power, C., J. C. McArthur, R. T. Johnson, D. E. Griffin, J. D. Glass, S. Perryman, and B. Chesebro. 1994. Demented and nondemented patients with AIDS differ in brain-derived human immunodeficiency virus type 1 envelope sequences. J. Virol. 68:4643-4649.
Price, R. W. 1996. Neurological complications of HIV infection. Lancet 348:445-452.
Price, R. W. 2000. The two faces of HIV infection of cerebrospinal fluid. Trends Microbiol. 8:387-391.
Price, R. W., and S. Staprans. 1997. Measuring the "viral load" in cerebrospinal fluid in human immunodeficiency virus infection: window into brain infection? Ann. Neurol. 42:675-678.
Pulliam, L., R. Gascon, M. Stubblebine, D. McGuire, and M. S. McGrath. 1997. Unique monocyte subset in patients with AIDS dementia. Lancet 349:692-695.
Pulliam, L., B. G. Herndier, N. M. Tang, and M. S. McGrath. 1991. Human immunodeficiency virus-infected macrophages produce soluble factors that cause histological and neurochemical alterations in cultured human brains. J. Clin. Investig. 87:503-512.
Reddy, R. T., A. D. Alexa, A. Sirko, S. Tehranchi, F. G. Kraus, F. Wong-Staal, C. A. Wiley, et al. 1996. Sequence analysis of the V3 loop in brain and spleen of patients with HIV encephalitis. AIDS Res. Hum. Retrovir. 12:477-482.
Richman, D. D., T. Wrin, S. J. Little, and C. J. Petropoulos. 2003. Rapid evolution of the neutralizing antibody response to HIV type 1 infection. Proc. Natl. Acad. Sci. USA 100:4144-4149.
Ross, H. L., S. Gartner, J. C. McArthur, J. R. Corboy, J. J. McAllister, S. Millhouse, and B. Wigdahl. 2001. HIV-1 LTR C/EBP binding site sequence configurations preferentially encountered in brain lead to enhanced C/EBP factor binding and increased LTR-specific activity. J. Neurovirol. 7:235-249.
Sacktor, N., P. M. Tarwater, R. L. Skolasky, J. C. McArthur, O. A. Selnes, J. Becker, B. Cohen, and E. N. Miller. 2001. CSF antiretroviral drug penetrance and the treatment of HIV-associated psychomotor slowing. Neurology 57:542-544.
Saksena, N. K., B. Wang, Y. C. Ge, J. Chang, D. E. Dwyer, S. H. Xiang, D. R. Packham, C. Randle, and A. L. Cunningham. 1997. Region-specific changes, gene duplications, and random deletions in the nef gene from HIV type 1-infected brain tissues and blood of a demented patient. AIDS Res. Hum. Retrovir. 13:111-116.
Sei, S., S. K. Stewart, M. Farley, B. U. Mueller, J. R. Lane, M. L. Robb, P. Brouwers, and P. A. Pizzo. 1996. Evaluation of human immunodeficiency virus (HIV) type 1 RNA levels in cerebrospinal fluid and viral resistance to zidovudine in children with HIV encephalopathy. J. Infect. Dis. 174:1200-1206.
Shioda, T., J. A. Levy, and C. Cheng-Mayer. 1992. Small amino acid changes in the V3 hypervariable region of gp120 can affect the T-cell-line and macrophage tropism of human immunodeficiency virus type 1. Proc. Natl. Acad. Sci. USA 89:9434-9438.
Slatkin, M., and W. P. Maddison. 1989. A cladistic measure of gene flow inferred from the phylogenies of alleles. Genetics 123:603-613.
Staprans, S., N. Marlowe, D. Glidden, T. Novakovic-Agopian, R. M. Grant, M. Heyes, F. Aweeka, S. Deeks, and R. W. Price. 1999. Time course of cerebrospinal fluid responses to antiretroviral therapy: evidence for variable compartmentalization of infection. AIDS 13:1051-1061.
Strain, M., H. Gunthard, D. Havlir, C. Ignacio, D. M. Smith, A. J. Leigh Brown, T. Macaranas, R. Y. Lam, O. A. Daly, M. Fischer, M. Opravil, H. Levine, L. Bacheler, C. Spina, D. Richman, and J. K. Wong. 2003. Heterogeneous clearance rates of long-lived lymphocytes infected with HIV: intrinsic stability predicts lifelong persistence. Proc. Natl. Acad. Sci. USA 100:4819-4824.
Strelow, L. I., D. D. Watry, H. S. Fox, and J. A. Nelson. 1998. Efficient infection of brain microvascular endothelial cells by an in vivo-selected neuroinvasive SIVmac variant. J. Neurovirol. 4:269-280.
Tyler, K. L., and J. C. McArthur. 2002. Through a glass, darkly: cerebrospinal fluid viral load measurements and the pathogenesis of human immunodeficiency virus infection of the central nervous system. Arch. Neurol. 59:909-912.
Venturi, G., M. Catucci, L. Romano, P. Corsi, F. Leoncini, P. E. Valensin, and M. Zazzi. 2000. Antiretroviral resistance mutations in HIV-1 RT and protease from paired CSF and plasma samples. J. Infect. Dis. 181:740-745.
Watry, D., T. E. Lane, M. Streb, and H. S. Fox. 1995. Transfer of neuropathogenic simian immunodeficiency virus with naturally infected microglia. Am. J. Pathol. 146:914-923.
Wei, X., J. M. Decker, S. Wang, H. Hui, J. C. Kappes, X. Wu, J. F. Salazar-Gonzalez, M. G. Salazar, J. M. Kilby, M. S. Saag, N. L. Komarova, M. A. Nowak, B. H. Hahn, P. D. Kwong, and G. M. Shaw. 2003. Antibody neutralization and escape by HIV-1. Nature 422:307-312.
Wildemann, B., J. Haas, K. Ehrhart, M. Hahn, H. Wagner, N. Lynen, and B. Storch-Hagenlocher. 1993. In vivo comparison of zidovudine resistance mutations in blood and CSF of HIV-1-infected patients. Neurology 43:2659-2663.
Wiley, C. A., R. D. Schrier, J. A. Nelson, P. W. Lampert, and M. B. Oldstone. 1986. Cellular localization of human immunodeficiency virus infection within the brains of acquired immune deficiency syndrome patients. Proc. Natl. Acad. Sci. USA 83:7089-7093.
Wiley, C. A., V. Soontornniyomkij, L. Radhakrishnan, E. Masliah, J. Mellors, S. A. Hermann, P. Dailey, and C. L. Achim. 1998. Distribution of brain HIV load in AIDS. Brain Pathol. 8:277-284.
Willey, S. J., J. D. Reeves, R. Hudson, K. Miyake, N. Dejucq, D. Schols, E. De Clercq, J. Bell, A. McKnight, and P. R. Clapham. 2003. Identification of a subset of human immunodeficiency virus type 1 (HIV-1), HIV-2, and simian immunodeficiency virus strains able to exploit an alternative coreceptor on untransformed human brain and lymphoid cells. J. Virol. 77:6138-6152.
Wolinsky, S. M., B. T. Korber, A. U. Neumann, M. Daniels, K. J. Kunstman, A. J. Whetsell, M. R. Furtado, Y. Cao, D. D. Ho, and J. T. Safrit. 1996. Adaptive evolution of human immunodeficiency virus-type 1 during the natural course of infection. Science 272:537-542.
Wong, J. K., C. C. Ignacio, F. Torriani, D. Havlir, N. J. S. Fitch, and D. D. Richman. 1997. In vivo compartmentalization of HIV: evidence from the examination of pol sequences from autopsy tissues. J. Virol. 70:2059-2071.
Zhang, K., F. Rana, C. Silva, J. Ethier, K. Wehrly, B. Chesebro, and C. Power. 2003. Human immunodeficiency virus type 1 envelope-mediated neuronal death: uncoupling of viral replication and neurotoxicity. J. Virol. 77:6899-6912.
Zhang, Z., T. Schuler, M. Zupancic, S. Wietgrefe, K. A. Staskus, K. A. Reimann, T. A. Reinhart, M. Rogan, W. Cavert, C. J. Miller, R. S. Veazey, D. Notermans, S. Little, S. A. Danner, D. D. Richman, D. Havlir, J. Wong, H. L. Jordan, T. W. Schacker, P. Racz, K. Tenner-Racz, N. L. Letvin, S. Wolinsky, and A. T. Haase. 1999. Sexual transmission and propagation of SIV and HIV in resting and activated CD4+ T cells. Science 286:1353-1357.
Zink, M. C., K. Suryanarayana, J. L. Mankowski, A. Shen, M. Piatak, Jr., J. P. Spelman, D. L. Carter, R. J. Adams, J. D. Lifson, and J. E. Clements. 1999. High viral load in the cerebrospinal fluid and brain correlates with severity of simian immunodeficiency virus encephalitis. J. Virol. 73:10480-10488.
Zink, W. E., J. Zheng, Y. Persidsky, L. Poluektova, and H. E. Gendelman. 1999. The neuropathogenesis of HIV-1 infection. FEMS Immunol. Med. Microbiol. 26:233-241.(M. C. Strain, S. Letendre)
HIV Neurobehavioral Research Center
Veterans Affairs San Diego Healthcare System, San Diego
Applied Biosystems, Foster City, California
Division of Infectious Diseases and Hospital Epidemiology, University Hospital of Zurich, Zurich, Switzerland
Northwestern University, Chicago, Illinois
ABSTRACT
Human immunodeficiency virus (HIV) infection of the central nervous system (CNS) is a significant cause of morbidity. The requirements for HIV adaptation to the CNS for neuropathogenesis and the value of CSF virus as a surrogate for virus activity in brain parenchyma are not well established. We studied 18 HIV-infected subjects, most with advanced immunodeficiency and some neurocognitive impairment but none with evidence of opportunistic infection or malignancy of the CNS. Clonal sequences of C2-V3 env and population sequences of pol from HIV RNA in cerebrospinal fluid (CSF) and plasma were correlated with clinical and virologic variables. Most (14 of 18) subjects had partitioning of C2-V3 sequences according to compartment, and 9 of 13 subjects with drug resistance exhibited discordant resistance patterns between the two compartments. Regression analyses identified three to seven positions in C2-V3 that discriminated CSF from plasma HIV. The presence of compartmental differences at one or more of the identified positions in C2-V3 was highly associated with the presence of discordant resistance (P = 0.007), reflecting the autonomous replication of HIV and the independent evolution of drug resistance in the CNS. Discordance of resistance was associated with severity of neurocognitive deficits (P = 0.07), while low nadir CD4 counts were linked both to the severity of neurocognitive deficits and to discordant resistance patterns (P = 0.05 and 0.09, respectively). These observations support the study of CSF HIV as an accessible surrogate for HIV virions in the brain, confirm the high frequency of discordant resistance in subjects with advanced disease in the absence of opportunistic infection or malignancy of the CNS, and begin to identify genetic patterns in HIV env associated with adaptation to the CNS.
INTRODUCTION
The primate lentiviruses human immunodeficiency virus type 1 (HIV-1) and the simian immunodeficiency virus SIVmac invade the central nervous system (CNS) within days to weeks of primary infection (11, 37, 102), and in the case of HIV result in neurocognitive dysfunction in up to 50% of untreated individuals in late disease (31, 36, 57, 71, 74). Such neurocognitive changes have been linked to postmortem neuropathologic findings of HIV encephalitis (10) and injury to the synaptodendritic apparatus (24, 56). This neuropathology occurs independently of known opportunistic infection (OI) or neurologic malignancy and requires both productive HIV infection and its immunologic sequelae, in particular, the activation and recruitment of macrophages into the CNS (27, 77, 78, 103).
Vigorous but ultimately ineffective host immune responses impel the rapid evolution of HIV-1 in blood and lymphatic tissues (6, 47, 80, 93, 98), but whether similar forces are in play in the CNS is not known. Indeed, the CNS constitutes an immunologically distinct and possibly privileged site that, in the case of other viral pathogens, serves as a chronic sanctuary (12). In addition, HIV predominantly infects microglial cells and perivascular macrophages in the CNS (44, 95), while the principal targets are activated CD4+ lymphocytes in blood and the lymphoid tissues (101). Selective adaptation of replicating HIV populations in the CNS to these cell types might therefore be expected. Thus, findings from earlier studies indicating that HIV populations in the CNS are genetically distinct from virus present in blood or lymphoid tissues (23, 29, 35, 46, 66, 69, 72, 73) are not surprising and may reflect either differences of the principal cellular targets or differences of immune selection. Whether particular sequence motifs define either neurotropism or neurovirulence remains unresolved (29, 46, 66, 73, 79). Observed sequence differences may arise due to genetic drift rather than to increased replicative efficiency or immune selection in the CNS (42).
The potential for differential effects of antiretroviral treatment on CNS and systemic HIV populations both has clinical implications and provides opportunities for the investigation of HIV genetics in the CNS (20, 21, 59, 75, 87, 90). Although many studies using CSF virus as a surrogate for virus in brain parenchyma suggest that highly active antiretroviral therapy (ART) is usually effective over short periods in the CNS (26, 33), the long-term durability of these responses remains unaddressed. Differences in baseline HIV genetics, the longevity of infected microglia, and limited drug penetration justify concern for long-term outcomes. Indeed, studies suggest that neurological manifestations of HIV infection constitute an increasing proportion of AIDS-defining illness in some settings (17, 18, 64), and when drug resistance develops, discordant resistance patterns between CNS and blood virus have been observed in many though not all studies of HIV in brain (59, 99) or cerebrospinal fluid (CSF) (14, 15, 84, 91, 94).
In this study, we examined the genetics of HIV-1 in subjects either failing therapy or not on treatment who underwent paired blood and CSF sampling in the absence of opportunistic neurological disease. Population sequences of pol and clonal sequences of C2-V3 env from these subjects revealed genetically distinct populations in the CSF of most subjects irrespective of antiviral therapy. Analysis by logistic regression identified three to seven codons in C2-V3 that are important as tissue-specific genetic patterns. Subjects harboring these signature patterns in CSF virus were very likely to exhibit discordant patterns of drug resistance in CSF and plasma. These results provide evidence that HIV sequence evolution in the CNS is shaped by compartment-specific selective pressures reflected, in a majority of cases, by genetically discrete CSF HIV populations.
(Presented in part at the 10th Conference on Retroviruses and Opportunistic Infections, Boston, Mass., February 2003, and the Workshop on HIV Genetics and NeuroAIDS, Washington, D.C., December 2002.)
MATERIALS AND METHODS
Subjects. Twenty-one individuals enrolled in longitudinal clinical studies at the HIV Neurobehavioral Research Center between years 1998 and 2002 were initially studied. All subjects had stable or no antiviral therapy for at least 2 months prior to study, had plasma and CSF HIV RNA of >500 copies/ml, and had no evidence of systemic or CNS OIs or malignancy based on clinical, laboratory, and neuro-imaging studies. Data were available on past and present therapy, current HIV RNA and CD4 counts, nadir CD4 counts, and CSF cell counts. All studies were conducted in compliance with local Institutional Review Board guidelines and with subjects' written informed consent.
Specimen processing. Paired blood from peripheral venipuncture in acid citrate dextrose tubes and CSF from lumbar punctures were collected (typically collected within 1 h of each other) and processed within 2 h of collection. Plasma and cell-free CSF were aliquoted, frozen, and stored at –70°C until processing. All subsequent plasma and CSF processing was performed separately to minimize the within-subject cross-contamination of samples.
HIV RNA quantitation. Plasma and CSF HIV RNA levels were measured by the Roche Amplicor Monitor assay (lower limit of detection of 400 copies/ml) or the Ultrasensitive assay (lower limit of detection of 50 copies/ml) with modifications as previously described (21).
NP testing. Neuropsychological (NP) testing was performed within 2 days of CSF sampling for 17 of 18 evaluable subjects except for subject N, whose testing was performed 17 months earlier. Overall NP performance was summarized by calculating global deficit scores (GDS). In brief, the GDS approach quantifies the number and degree of impaired NP test scores across the battery while attaching relatively less significance to performances that are within (or above) normal limits. For each of the NP test variables, raw scores were converted to T-scores (mean, 50; standard deviation, 10) using published demographically corrected normative data. The T-scores were then transformed into deficit scores using the following conversions: >40T = 0; 39T to 35T = 1; 34T to 30T = 2; 29T to 25T = 3; 24T to 20T = 4; and <19T = 5. The deficit scores from each NP test were then averaged to derive a GDS for each participant (range, 0 to 5). Prior research supports the construct validity of the GDS approach as an indicator of global NP functioning in persons with HIV infection and indicates that the GDS is a reliable and valid alternative to clinical ratings and, more importantly, can be generalized across different NP test batteries (9, 36). GDS above 0.5 were interpreted as signifying NP impairment.
Nucleotide sequencing. (i) pol sequencing. pol sequencing was performed using the Viroseq version 2.0 system (Applied Biosystems, Alameda, Calif.) on an ABI 3100 genetic analyzer according to the manufacturer's protocol. In brief, 0.5 to 2.0 ml of cell-free CSF or plasma was thawed and centrifuged at 23,900 x g and 4°C for 1 h to pellet virus. We chose to concentrate virus even though most samples contained sufficient HIV copy numbers for assays and successful amplification without further concentration in order to improved representative sampling. RNA was extracted and resuspended in 50 to 100 μl of diluent according to the protocol. A 10-μl aliquot of resuspended RNA was used in pol sequencing; the remaining 40 μl was used for clonal C2-V3 sequence studies. Raw data from pol sequencing were edited and interpreted using Viroseq HIV genotyping system software version 2.5 (Applied Biosystems, Foster City, Calif.) to generate a genotypic drug resistance profile. The default base-caller for this software identifies a minority HIV population when it is present in at least a 30% proportion.
(ii) env sequencing. Reverse transcription and PCR amplification of C2-V3 env for each sample was performed in triplicate or quadruplicate using the Finnzyme one-step reverse transcription-PCR (RT-PCR) kit (MJ Research, Waltham, Mass.) and primers V3Fout and V3Bout as previously described (32) in a 25-μl reaction volume. A 2.5-μl aliquot of first-step RT-PCR product was used in the second, nested PCR with primers V3Fin and V3Bin (32) and Tris (pH 8.0), 20 pmol of each primer, 2.5 mM MgCl2, 200 μM deoxynucleoside triphosphates, and bovine serum albumin (10 mg/ml) in a total volume of 50 μl. Cycling parameters were 94°C for 2 min; 94°C for 30s, 54°C for 30 s, and 72°C for 1 min for 35 cycles; and 72°C for 10 min. All assays were conducted under conditions to minimize the potential for PCR contamination by utilizing aerosol-resistant pipette tips, dedicated PCR reagents, and laminar flow hoods. All assays included negative controls. Nested PCR products were visualized by agarose gel electrophoresis and ethidium staining. Replicate PCR products were proportionately pooled and cloned using the TOPO-TA cloning system (Invitrogen, Carlsbad, Calif.). Clones were selected by blue-white screening and expanded in 4.0-ml broth cultures. Plasmid clones were purified using QIAGEN Mini-prep kits (QIAGEN). Purified plasmids were sequenced in both directions with –20M13 primer (5'-GTAAAACGACGGCCAG-3') and Topo forward primer (5'-TGGATATCTGCAGAATTCG-3') using Prism Dye terminator kits (ABI) on an ABI 3100 genetic analyzer. Sequences were compiled, aligned, and edited using Sequencher 4.0 (Genecodes, Ann Arbor, Mich.) and Clustal (version 1.81).
Genetic analysis. Phylogenetic analysis was performed using DNAML (Phylip 3.6a; J. Felsenstein) (T/T ratio of 2.0, gamma-distributed substitution rate) on clonal sequences for individual subjects and DNADIST, NEIGHBOR, and CONSENSE for 100 bootstrapped data sets from all sequences (Phylip 3.5; J. Felsenstein). Trees were displayed using Treeview (Roderic DM Page). Assessment of degree of intercompartment segregation was performed by testing for panmixis using gene phylogenies (38, 86) as implemented in MacClade (Sinauer, Sunderland, Mass.). In brief, from the maximum likelihood trees for each individual subject's C2-V3 sequences and their characterization according to compartment of origin, the minimum number of intercompartment migration events allowed by the tree was tallied. This result was compared to the distribution of migration events for 1,000 randomly generated trees. Evidence of restricted gene flow (compartmentalization) was documented when <1% (or between 1 and 5%) of the random trees required the same or a fewer number of migration events as for the sample data (65, 86). Assessment of the role of selection on HIV genetic composition was performed using a Z test for selection with a modified Nei-Gojobori method and p-distances (63) as implemented in MEGA (49). Tests for positive selection, neutral evolution, and purifying selection were performed at sites identified as important for distinguishing CSF from plasma virus.
Analysis of sequence differences between CSF and plasma virus. Prediction of coreceptor usage was based on composition at residues 11 and 25 of the V3 loop (codons 306 and 322 of HXB2 gp120) (25, 85). Residues at which amino acid (AA) frequencies varied significantly between compartments were identified by Monte Carlo approximation of the Mantel-Haenszel and likelihood ratio tests, based on identity of individual AA or on chemical class of AA (positive, negative, neutral, or hydrophobic), with a P value of <0.05 considered significant. Positions with intercompartmental variation by both of these univariate measures were included in a stepwise logistic regression (with entry cutoff P value of <0.10 and retention cutoff P value of <0.05). Forward logistic regression was performed with the entry criterion P value of <0.05. Analyses were performed on the entire data set as well as on "pruned" data, including only those subjects exhibiting phylogenetic segregation of CSF and plasma sequences verified to be significant by the Maddison-Slatkin test (P < 0.01) (86).
Statistical analysis of group characteristics. Fisher's exact test was used to compare proportions between groups. In some cases as noted, parameters were logistically fitted and a whole-model test of significance was applied. Statistical analyses were performed using SAS V8 and JMP IN (SAS Institute, Cary, N.C.).
Nucleotide sequence accession number. The pol and env sequences have been to GenBank and assigned accession numbers AY859056 to AY859091.
RESULTS
PCR amplification for both C2-V3 env and resistance genotyping was successful in 18 of 21 subjects but failed in 3 subjects. HIV RNA concentrations did not appear to predict failure for PCR amplification of either pol or env sequences. The median levels of HIV RNA (log10 copies per milliliter) for subjects whose amplifications failed versus succeeded were similar in plasma (5.29 versus 5.35) and CSF (4.08 versus 4.15). Subject characteristics, treatment history, and NP test scores for the remaining 18 subjects are shown in Table 1. Subjects had relatively advanced disease (median CD4, 128 cells/mm3) and the majority (13 of 18) were neurocognitively impaired as defined by a GDS of >0.5.
C2-V3 env sequences. Most subjects (13 of 18) exhibited partitioning of CSF from plasma C2-V3 sequences (Fig. 1). For 11 subjects, the Maddison-Slatkin test for compartment-based partitioning was highly significant (P < 0.01), and all exhibited clear segregation of the majority of the CSF from plasma sequences on visual inspection of phylogenetic trees. Two other subjects had evidence of partitioning (0.01 < P < 0.05) but appeared to have a greater degree of intermingling of CSF and plasma sequences with only subsets of sequences that were segregated. In 17 of 18 subjects, sequences from CSF and plasma from the same subject clustered more closely to one another than did either CSF-CSF or plasma-plasma sequences between subjects, reflecting the impact of common ancestry of within-subject virus populations on the phylogenetic reconstruction (Fig. 2). In subject Q, the sequences formed two divergent but related clusters. While these sequences were more closely related to one another than to sequences from other subjects, the bootstrap value was marginal (43). Sequences from subject B comprised two very divergent, seemingly unrelated clusters, one having both CSF and plasma sequences and one with only plasma sequences (Fig. 1 and 2). Genetic distance between sequences from these two clusters was approximately 10%. These sequences did not appear to cluster with sequences from any other subjects or with lab strains. Of note, pol sequences from plasma virus in this subject also demonstrated genotypic mixtures at ratios of approximately 50:50 based on peak heights of chromatograms (data not shown). These two cases appear consistent with cocirculation of markedly divergent HIV variants reflecting either replication under very different selective environments or, in the case of subject B, coinfection or superinfection with a second HIV variant (2, 40, 43).
Analysis of differences in env sequence between CSF and plasma virus. Predicted coreceptor usage did not distinguish CSF from plasma viruses in the clonal sequences examined for most subjects. With few exceptions the majority of clones from both CSF and plasma were predicted to be from virus using CCR5 rather than CXCR4, based on the composition of residues 11 and 25 of the V3 loop (codons 306 and 322, respectively, of env) (25, 85) (Fig. 3). Basic residues at position 11 or 25 were found in 3 of 14 plasma and 1 of 13 CSF clones from subject I, 5 of 15 plasma and 0 of 14 CSF clones from subject M, and 11 of 15 plasma and 0 of 14 CSF clones from subject N. Notably, all CSF clones from subject O exhibited an arginine at position 25, but this position was deleted from all 10 plasma clones sequenced. Thus, while we encountered clones predicted to utilize CXCR4 more frequently in plasma virus, this type occurred in only a small number of subjects. CXCR4 using HIV was found occasionally in CSF, consistent with recent reports of its identification from brain parenchyma (66).
Regression analyses were used to verify positions that distinguished CSF from plasma clones for each subject and for all the sequenced clones. Residues 293, 308, and 341 were associated with CSF compartmentalization in the entire sequence set (data not shown). The set of unique sequences in the subset of subjects exhibiting clear partitioning of CSF and plasma sequences (n = 11, pruned subset) were separately analyzed. The P value for evidence of partitioning was <0.01 by both 2 analysis and stepwise logistic regression (Table 2; Fig. 4a). Categorization of amino acids by charge and hydrophobicity (as positive, negative, neutral, or hydrophobic) improved the predictive value of positions 293, 308, and 341 (Table 3; Fig. 4b).
In order to assess whether the sequence differences between plasma and CSF virus were due to genetic drift of isolated HIV populations or resulted from selection, a Z-test using a modified Nei-Gojobori method (p distances) for comparing synonymous and nonsynonymous substitutions at codons 293, 308, and 341 was performed for all CSF sequences and for all plasma sequences. For both CSF and plasma virus, the P values strongly supported positive selection at these sites (P < 0.001 for both). When all seven sites were included, the P value for positive selection for CSF virus increased to 0.07.
Drug resistance. Fourteen of 18 subjects exhibited mutations conferring drug resistance (70), 8 of 9 receiving antiretrovirals and 6 of 9 not on treatment at the time of study (Table 4). Among the eight subjects with drug resistance on ART, five had discordance in resistance patterns between CSF and plasma HIV. Similarly, three of four subjects with past but not current ART had discordance in resistance patterns between CSF and plasma HIV. Two subjects without a known history of drug treatment exhibited drug resistance. Overall, 50% of subjects demonstrated discordant resistance mutations, a proportion comparable to that in previous reports (14, 91).
Twelve of 14 subjects with resistance had mutations associated with resistance to more than one class of antiretroviral drug. Phylogenetic reconstruction of pol sequences showed that CSF and plasma sequences from the same subject were most closely related to one another, which excluded contaminating sequences as an explanation for discordant resistance patterns (data not shown).
Of the nine subjects with discordant drug resistance, resistance mutations were noted in plasma but not in CSF for eight, while only one subject (A) exhibited more resistance in virus from CSF than from plasma. This pattern suggests lesser selective pressure by ART or a smaller population of preexisting HIV with mutations for resistance in the CNS compartment. In several cases of concordant multidrug resistance, the patterns of resistance were extremely complex but were nonetheless identical between CSF and plasma virus, arguing for the emergence of resistance as single rather than parallel events in these cases.
Poor penetration of individual antiviral drugs into the CNS could favor the development of discordant resistance for some drugs. We could not systematically address this issue, since the study subjects were treated with a variety of different antiretroviral regimens. Nevertheless, the frequency of discordance for resistance mutations to nucleoside reverse transcriptase inhibitors, a class with good CNS penetration (26, 82), was no less than for resistance mutations to protease inhibitors, which generally penetrate poorly (3, 28, 50, 52) (Table 4).
Correlates of drug resistance patterns and partitioning of C2-V3 sequence. In subjects with drug-resistant virus, concordant resistance patterns were associated with lack of partitioning between CSF and plasma env sequences, reflecting similar HIV populations in both compartments (P = 0.035) (Table 5). This observation is compatible with selection of drug resistance in the HIV population in one compartment, which then spreads to the other perhaps because of mixing of the two populations. Conversely, discordance of resistance patterns correlated with envelope partitioning (P = 0.035) and with the presence of differences at signature positions in env as summarized in Table 2 (P = 0.007), consistent with the independent evolution of resistance in segregated and autonomously replicating HIV populations. Discordant resistance tended to be more common in subjects with lower nadir CD4 counts, as suggested by Cunningham and colleagues (14).
Treatment history tended to correlate with segregation of CSF and plasma env sequence clusters (P = 0.09) (Table 5), suggesting an effect of treatment on the composition of HIV env genotypes; however, there were relatively few subjects who were drug na?ve (n = 5) (Table 1). The apparent association of treatment and env sequence partitioning may have reflected a prestudy bias towards treating subjects with more advanced disease. In this small study population, CSF pleiocytosis was not significantly associated with resistance patterns or with the observed partitioning of C2-V3 sequences (data not shown).
Correlates of neurocognitive dysfunction. Nadir but not current CD4 count was correlated with severity of neurocognitive impairment (P = 0.05) (Table 5). A whole-model test for the logistic fit of nadir CD4 and presence or absence of neurocognitive impairment also demonstrated a significant association (P = 0.04). These correlations are consistent with the increasing prevalence of neurocognitive dysfunction with advancing disease. Of note, partitioning of blood and CSF C2-V3 sequences was not significantly associated with neurocognitive abnormalities. However, subjects with concordant resistance or no resistance had a lower GDS than those with discordant resistance (median GDS = 0.8 for concordant or no resistance, median GDS = 1.6 for discordant resistance; P = 0.07). Thus, more advanced disease and the presence of discordant resistance were positively associated with the presence and severity of neurocognitive dysfunction.
DISCUSSION
Infection of the CNS is an important contributor to the morbidity of HIV disease, but study of the in vivo biology of CNS infection by HIV is complicated by the relative inaccessibility of tissues of the CNS. Because of the impracticality of sampling the brain for the antemortem study of HIV genetics, many investigators have examined virus in CSF as a surrogate marker for parenchymal disease (8, 22, 55, 58, 76, 87). Whether HIV RNA in CSF accurately reflects the predominant HIV species replicating in brain parenchyma is unknown, but HIV RNA concentrations in brain and CSF appear to be correlated (58, 96) and the predictive value of CSF HIV RNA but not HIV RNA in blood for neurocognitive impairment suggests a significant relationship (8, 21, 58). In addition to the relative ease of sampling CSF, the study of CSF virus provides additional theoretical advantages. Whereas parenchymal DNA can disproportionately reflect archival and in some cases replication-defective virus, RNA in CSF has been shown to turn over relatively rapidly (20, 87), reflecting the kinetics of recently formed virions, in a manner analogous to the differences observed between plasma virus and HIV DNA in peripheral blood mononuclear cells (34, 61, 88). Relatively few HIV RNA genetic studies of postmortem brain specimens have been reported, possibly due to the lability of RNA and their degradation during the postmortem period. However, in the SIV-infected macaque model, DNA and RNA comprised distinct sequence variants when recovered from brain and isolated microglia of freshly euthanized animals (4). Underscoring this particular advantage of CSF RNA sequences, only 3 of the 225 CSF RNA clones we sequenced contained frameshifts or nonsense mutations compared with high frequencies of defective forms reported from prior studies of HIV DNA sequence in brain parenchyma (53, 54, 68). Thus, genetic studies of CSF HIV in subjects with relatively advanced disease complement HIV studies using brain samples from autopsies or biopsies (29, 39, 62) by providing only recently produced virions. We studied cell-free CSF samples to minimize effects from sampling of HIV sequences from lymphocytes that may traffic into the CSF. This approach may account in part for the relatively high proportion of samples demonstrating segregation of HIV populations.
Genetic studies comparing HIV from the CNS with that in blood or lymphoid organs have identified sequence differences in env, pol, gag, nef, tat, and the long terminal repeat (1, 5, 7, 13, 42, 46, 62, 66, 72, 81, 83, 99). The greatest attention has been paid to env sequences, in particular those corresponding to the V3 loop, an important (though not exclusive) determinant of both cell tropism and neutralization susceptibility by antibody (45). The reported differences could result from genetic drift of segregated populations or from selection by immune or other factors within the CNS. Our findings that C2-V3 sequences in CSF differed from those in plasma therefore corroborate many earlier studies that have catalogued genetic differences between virus in CNS and virus in blood or lymphoid tissues. However, while some earlier studies have suggested the existence of env genotypes associated with neurotropism and neurovirulence (46, 48, 73), others have disputed this (16, 66, 79). An aim of the present study was to revisit this question.
Recently proposed models for the origin of CSF virus implicate an increasing contribution of parenchymal CNS sources over peripheral lymphoid sources with advancing disease (20, 21, 87, 90). Our results are consistent with such a view, since we found a trend towards more segregation between CSF and plasma sequences in individuals with advanced disease. Discordant resistance was encountered more frequently for individuals with a nadir CD4 of less than 200/mm3, although this association did not reach statistical significance, owing perhaps to the small number of individuals studied with CD4 counts of >200 mm3.
Anticipating that HIV from both systemic and CNS sources might be sampled from CSF in any given subject, we sought a CNS signature sequence in C2-V3 from both the entire data set as well as from a pruned data set by using phylogenetic criteria to exclude those individuals likely to have CSF virus produced by systemic sources. We did not find a single exclusive signature pattern that predicted the CNS origin of a virus. However, regression analyses identified a small number of positions that appeared to be important for discriminating HIV populations from the two compartments. In related work, construction of a decision tree based on the physico-chemical characteristics of AA at six positions permitted the proper classification of up to 90% of sequences when evaluating only the pruned data (M. C. Strain, J. K. Wong and S. Pillai, unpublished data). Of note, position 308 is identical to one of two positions implicated by Power and colleagues (referred to as position 305 of the macrophage-tropic HIV consensus) to be important in determining neurovirulence (73) and to one of several positions in the V3 loop identified by Korber and colleagues as possible CNS signatures (46). Together, these findings suggest that the HIV phenotype(s) that confers neurotropism is genetically complex, and although the phenotype(s) may not be described by simple signature AA substitutions, it is nevertheless associated with discernible and predictable covariation. Further, the finding that codons 293, 308, and 341 are under positive selection and can be used to classify sequences by anatomic source based either on specific AA or physico-chemical characteristics of the AA provides compelling evidence for effects of selective forces generic to the CNS acting on these sites in the HIV envelope.
Despite the identification of residue 308 in this study as an important CNS-defining position and the implication of the same position by Powers and colleagues as a position conferring neurovirulence, we did not observe any evidence that CNS signature mutations correlated with cognitive abnormalities. Thus, we interpret these patterns as indicative of HIV neurotropism but not necessarily neurovirulence. However, such determinants of neurovirulence might be outside of the C2-V3 domain (66, 100) or might be found only in minority HIV populations that infect small but functionally critical cellular targets, such as astrocytes (60, 67, 97) and microvascular endothelial cells (89, 97). In the macaque model of neuro-AIDS, adaptation of SIV by serial animal passage in microglial cells appears to confer a phenotype that reliably produces SIV encephalitis (51, 92), providing further evidence that either or both neurotropism and neurovirulence of primate lentiviruses are selectable traits.
The seven signature positions identified in this study do not include either of the canonical positions in the V3 loop that help to define coreceptor usage (25, 85). This observation is consistent with others indicating that coreceptor usage does not discriminate HIV populations in the CNS from systemic HIV (30, 66). Phenotypic studies were not performed as a part of the present study, and such studies will be necessary to define the biologic requirements for neurotropism. Potential phenotypic characteristics for CNS adaptation might include such features as a lower requirement for primary and coreceptor density, as supported by in vitro models (19) and in vivo studies on small numbers of subjects (30). While one or more of the residues that we identified here might confer such a phenotype, they may also interact with sites outside of V3 and play indirect roles.
Segregated, CNS-adapted HIV populations in the CSF exhibited distinct resistance patterns. The high frequency of discordant resistance in our subjects, all of whom had no concurrent OIs, confirms recent results from subjects who underwent CSF sampling because of suspected and documented CNS OIs (14, 91). Concordance of resistance patterns was associated with similar V3 sequences in CSF and plasma virus, suggesting either that CNS and systemic HIV populations were not segregated or a resistant HIV variant had replaced wild-type virus in both compartments in the absence of recombination between env and pol coding regions. In contrast, we observed both discordant resistance patterns and the simultaneous partitioning of V3 sequences in the majority of cases. These cases are consistent with either the independent emergence of resistance in isolated HIV populations or the emergence of resistance in one compartment that then spreads to the other with the subsequent divergence of HIV sequences (in such cases, possibly aided by recombination). However, without longitudinal data, these possibilities cannot be distinguished.
We encountered discordant resistance both in individuals receiving failing ART at the time of sampling as well as in individuals off therapy at the time of study. This suggests that HIV populations may evolve independently during both the selection of resistance on therapy and reversion to wild-type off therapy. One subject harbored a substantially more resistant virus in CSF off therapy, while the plasma virus had apparently reverted to a less resistant sequence type. In such a scenario, treatment selection based on resistance testing of plasma virus alone would have failed to account for this highly resistant, actively replicating HIV population.
This and other recent studies of HIV in CSF reveal a high frequency of discordant resistance compared with earlier studies (84, 94). Possible explanations for these discrepant findings are differences in the complexity and composition of antiretroviral regimens the study subjects received. Whereas many earlier studies had focused on resistance to zidovudine, a drug with reasonably good CSF penetration, many contemporary regimens include drugs with marginal CSF penetration, such as the protease inhibitors (3, 28, 50, 52). Heterogeneity of drug concentrations has been predicted to promote the emergence of resistance (41) and may also explain the maintenance of populations of virus with differing resistance patterns in this and other recent studies (14, 91). Finally, the presence of discordant drug resistance was associated with more severe neurocognitive dysfunction as reflected by higher GDS. This trend demonstrated only borderline statistical significance due to the relatively small sample size but is important to note, since subjects with more severe neurocognitive deficits for whom control of virus replication in CNS would be more clinically important appear more likely to harbor virus in the CNS that has a resistance profile discordant from that of blood virus.
In summary, our results demonstrate that CSF HIV appears genetically distinct from contemporaneous plasma HIV in a majority of individuals with advanced disease, confirming both compartmental segregation and autonomous replication within the CNS. Furthermore, a small number of signature positions in and near the V3 loop may reflect genetic adaptation to replication in the CNS. An important correlate of independent V3 genotypes was the frequent observation of discordant resistance patterns between virus in plasma and CSF. These findings may have important implications for drug treatment and subject monitoring. For example, drug failure based on plasma HIV resistance pattern might not mean that CSF HIV has also developed resistance. Continuation of drugs that were previously effective may be indicated for CNS protection, even as new agents are added to address drug-resistant plasma virus. Clearly, establishment of the clinical significance of such discordant resistance patterns will need to await results of detailed longitudinal studies. Nevertheless, these results help to substantiate the value of CSF sampling in studies aimed at elucidating HIV neuropathogenesis and guiding neuro-AIDS treatment.
ACKNOWLEDGMENTS
We are grateful to Sharon Wilcox, Roma Sysyn, Melanie Sherman, and Darica Smith for administrative and technical support, to Clayton Wiley for helpful comments on the manuscript, and to the participating subjects from the UCSD HNRC.
This study was supported by a VA Research Merit award and NIH grants AI 43752 (J.K.W.); MH 58076 (J.K.W., S.L., R.E., I.G., and J.A.M.); MH 62512 (I.G., R.E., T.M., S.L., J.A.M., J.M.-B., and J.D.); MH 59745 (I.G., R.E., T.M., and J.A.M.); DA12065 (I.G., R.E., T.M., and J.A.M.); AI 27670, 38858, 29164, 47745, and 36214 (D.D.R.); an NSF grant (M.S.); and Swiss National Science Foundation grant 3345-65168 (H.F.G.).
Supplemental material for this article may be found at http://jvi.asm.org/.
REFERENCES
Ait-Khaled, M., J. E. McLaughlin, M. A. Johnson, and V. C. Emery. 1995. Distinct HIV-1 long terminal repeat quasispecies present in nervous tissues compared to that in lung, blood and lymphoid tissues of an AIDS patient. AIDS 9:675-683.
Altfeld, M., T. M. Allen, X. G. Yu, M. N. Johnston, D. Agrawal, B. T. Korber, D. C. Montefiori, D. H. O'Connor, B. T. Davis, P. K. Lee, E. L. Maier, J. Harlow, P. J. Goulder, C. Brander, E. S. Rosenberg, and B. D. Walker. 2002. HIV-1 superinfection despite broad CD8+ T-cell responses containing replication of the primary virus. Nature 420:434-439.
Aweeka, F., A. Jayewardene, S. Staprans, S. E. Bellibas, B. Kearney, P. Lizak, T. Novakovic-Agopian, and R. W. Price. 1999. Failure to detect nelfinavir in the cerebrospinal fluid of HIV-1-infected patients with and without AIDS dementia complex. J. Acquir. Immune Defic. Syndr. Hum. Retrovir. 20:39-43.
Babas, T., D. Munoz, J. L. Mankowski, P. M. Tarwater, J. E. Clements, and M. C. Zink. 2003. Role of microglial cells in selective replication of simian immunodeficiency virus genotypes in the brain. J. Virol. 77:208-216.
Blumberg, B. M., L. G. Epstein, Y. Saito, D. Chen, L. R. Sharer, and R. Anand. 1992. Human immunodeficiency virus type 1 nef quasispecies in pathological tissue. J. Virol. 66:5256-5264.
Borrow, P., H. Lewicki, X. Wei, M. S. Horwitz, N. Peffer, H. Meyers, J. A. Nelson, J. E. Gairin, B. H. Hahn, M. B. Oldstone, and G. M. Shaw. 1997. Antiviral pressure exerted by HIV-1-specific cytotoxic T lymphocytes (CTLs) during primary infection demonstrated by rapid selection of CTL escape virus. Nat. Med. 3:205-211.
Bratanich, A. C., C. Liu, J. C. McArthur, T. Fudyk, J. D. Glass, S. Mittoo, G. A. Klassen, and C. Power. 1998. Brain-derived HIV-1 tat sequences from AIDS patients with dementia show increased molecular heterogeneity. J. Neurovirol. 4:387-393.
Brew, B. J., L. Pemberton, P. Cunningham, and M. G. Law. 1997. Levels of human immunodeficiency virus type 1 RNA in cerebrospinal fluid correlate with AIDS dementia stage. J. Infect. Dis. 175:963-966.
Carey, C. L., S. P. Woods, R. Gonzales, E. Conover, T. D. Marcotte, I. Grant, R. K. Heaton, et al. 2004. Predictive validity of global deficit scores in detecting neuropsychological impairment in HIV infection. J. Clin. Exp. Neuropsychol. 26:307-319.
Cherner, M., E. Masliah, R. J. Ellis, T. D. Marcotte, D. J. Moore, I. Grant, and R. K. Heaton. 2002. Neurocognitive dysfunction predicts postmortem findings of HIV encephalitis. Neurology 59:1563-1567.
Clements, J. E., T. Babas, J. L. Mankowski, K. Suryanarayana, M. Piatak, Jr., P. M. Tarwater, J. D. Lifson, and M. C. Zink. 2002. The central nervous system as a reservoir for simian immunodeficiency virus (SIV): steady-state levels of SIV DNA in brain from acute through asymptomatic infection. J. Infect. Dis. 186:905-913.
Cook, M. L., V. B. Bastone, and J. G. Stevens. 1974. Evidence that neurons harbor latent herpes simplex virus. Infect. Immun. 9:946-951.
Corboy, J. R., and P. J. Garl. 1997. HIV-1 LTR DNA sequence variation in brain-derived isolates. J. Neurovirol. 3:331-341.
Cunningham, P. H., D. G. Smith, C. Satchell, D. A. Cooper, and B. Brew. 2000. Evidence for independent development of resistance to HIV-1 reverse transcriptase inhibitors in the CSF. AIDS 14:1949-1954.
Di Stefano, M., F. Sabri, T. Leitner, B. Svennerholm, L. Hagberg, G. Norkrans, and F. Chiodi. 1995. Reverse transcriptase sequence of paired isolates of cerebrospinal fluid and blood from patients infected with human immunodeficiency virus type 1 during zidovudine treatment. J. Clin. Microbiol. 33:352-355.
Di Stefano, M., S. Wilt, F. Gray, M. Dubois-Dalcq, and F. Chiodi. 1996. HIV type 1 V3 sequences and the development of dementia during AIDS. AIDS Res. Hum. Retrovir. 12:471-476.
Dore, G. J., P. K. Correll, Y. Li, J. M. Kaldor, D. A. Cooper, and B. J. Brew. 1999. Changes to AIDS dementia complex in the era of highly active antiretroviral therapy. AIDS 13:1249-1253.
Dore, G. J., A. McDonald, Y. Li, J. M. Kaldor, and B. J. Brew. 2003. Marked improvement in survival following AIDS dementia complex in the era of highly active antiretroviral therapy. AIDS 17:1539-1545.
Edinger, A. L., J. L. Mankowski, B. J. Doranz, B. J. Margulies, B. Lee, J. Rucker, M. Sharron, T. L. Hoffman, J. F. Berson, M. C. Zink, V. M. Hirsch, J. E. Clements, and R. W. Doms. 1997. CD4-independent, CCR5-dependent infection of brain capillary endothelial cells by a neurovirulent simian immunodeficiency virus strain. Proc. Natl. Acad. Sci. USA 94:14742-14747.
Ellis, R. J., A. C. Gamst, E. Capparelli, S. A. Spector, K. Hsia, T. Wolfson, I. Abramson, I. Grant, and J. A. McCutchan. 2000. Cerebrospinal fluid HIV RNA originates from both local CNS and systemic sources. Neurology 54:927-936.
Ellis, R. J., K. Hsia, S. A. Spector, J. A. Nelson, R. K. Heaton, M. R. Wallace, I. Abramson, J. H. Atkinson, I. Grant, J. A. McCutchan, et al. 1997. Cerebrospinal fluid human immunodeficiency virus type 1 RNA levels are elevated in neurocognitively impaired individuals with acquired immunodeficiency syndrome. Ann. Neurol. 42:679-688.
Ellis, R. J., D. J. Moore, M. E. Childers, S. Letendre, J. A. McCutchan, T. Wolfson, S. A. Spector, K. Hsia, R. K. Heaton, and I. Grant. 2002. Progression to neuropsychological impairment in human immunodeficiency virus infection predicted by elevated cerebrospinal fluid levels of human immunodeficiency virus RNA. Arch. Neurol. 59:923-928.
Epstein, L. G., C. Kuiken, B. M. Blumberg, S. Hartman, L. R. Sharer, M. Clement, and J. Goudsmit. 1991. HIV-1 V3 domain variation in brain and spleen of children with AIDS: tissue-specific evolution within host-determined quasispecies. Virology 180:583-590.
Everall, I. P., R. K. Heaton, T. D. Marcotte, R. J. Ellis, J. A. McCutchan, J. H. Atkinson, I. Grant, M. Mallory, E. Masliah, et al. 1999. Cortical synaptic density is reduced in mild to moderate human immunodeficiency virus neurocognitive disorder. Brain Pathol. 9:209-217.
Fouchier, R. A. M., M. Groenink, N. A. Kootstra, M. Tersmette, H. G. Huisman, F. Miedema, and H. Schuitemaker. 1992. Phenotype-associated sequence variation in the third variable domain of the human immunodeficiency virus type 1 gp120 molecule. J. Virol. 66:3183-3187.
Foudraine, N. A., R. M. Hoetelmans, J. M. Lange, F. De Wolf, B. H. van Benthem, J. J. Maas, I. P. Keet, and P. Portegies. 1998. Cerebrospinal-fluid HIV-1 RNA and drug concentrations after treatment with lamivudine plus zidovudine or stavudine. Lancet 351:1547-1551.
Gendelman, H. E., S. A. Lipton, M. Tardieu, M. I. Bukrinsky, and H. S. L. M. Nottet. 1994. The neuropathogenesis of HIV-1 infection. J. Leukoc. Biol. 56:389-398.
Gisolf, E. H., R. H. Enting, S. Jurriaans, F. de Wolf, M. E. van der Ende, R. M. Hoetelmans, P. Portegies, and S. A. Danner. 2000. Cerebrospinal fluid HIV-1 RNA during treatment with ritonavir/saquinavir or ritonavir/saquinavir/stavudine. AIDS 14:1583-1589.
Gorry, P. R., G. Bristol, J. A. Zack, K. Ritola, R. Swanstrom, C. J. Birch, J. E. Bell, N. Bannert, K. Crawford, H. Wang, D. Schols, E. De Clercq, K. Kunstman, S. M. Wolinsky, and D. Gabuzda. 2001. Macrophage tropism of human immunodeficiency virus type 1 isolates from brain and lymphoid tissues predicts neurotropism independent of coreceptor specificity. J. Virol. 75:10073-10089.
Gorry, P. R., J. Taylor, G. H. Holm, A. Mehle, T. Morgan, M. Cayabyab, M. Farzan, H. Wang, J. E. Bell, K. Kunstman, J. P. Moore, S. M. Wolinsky, and D. Gabuzda. 2002. Increased CCR5 affinity and reduced CCR5/CD4 dependence of a neurovirulent primary human immunodeficiency virus type 1 isolate. J. Virol. 76:6277-6292.
Grant, I., J. H. Atkinson, J. R. Hesselink, C. J. Kennedy, D. D. Richman, S. A. Spector, and J. A. McCutchan. 1987. Evidence for early central nervous system involvement in the acquired immunodeficiency syndrome (AIDS) and other human immunodeficiency virus (HIV) infections. Ann. Intern. Med. 107:828-836.
Gunthard, H. F., S. D. Frost, A. J. Leigh-Brown, C. C. Ignacio, K. Kee, A. S. Perelson, C. A. Spina, D. V. Havlir, M. Hezareh, D. J. Looney, D. D. Richman, and J. K. Wong. 1999. Evolution of envelope sequences of human immunodeficiency virus type 1 in cellular reservoirs in the setting of potent antiviral therapy. J. Virol. 73:9404-9412.
Gunthard, H. F., D. V. Havlir, S. Fiscus, Z. Q. Zhang, J. Eron, J. Mellors, R. Gulick, S. D. F. Frost, A. J. Leigh-Brown, W. Schlief, F. Valentine, L. Jonas, A. Meibohm, C. Ignacio, R. Isaacs, R. Gamagami, E. Emini, A. T. Haase, D. D. Richman, and J. K. Wong. 2001. Residual HIV RNA and DNA in lymph node and HIV RNA in genital secretions and in CSF after two years of suppression of viremia in the Merck 035 cohort. J. Infect. Dis. 183:1318-1327.
Gunthard, H. F., J. K. Wong, C. C. Ignacio, J. C. Guatelli, N. L. Riggs, D. Havlir, and D. D. Richman. 1998. Human immunodeficiency virus replication and genotypic resistance in blood and lymph nodes after a year of potent antiretroviral therapy. J. Virol. 72:2422-2428.
Haggerty, S., and M. Stevenson. 1991. Predominance of distinct viral genotypes in brain and lymph node compartments of HIV-1-infected individuals. Viral Immunol. 4:123-131.
Heaton, R. K., I. Grant, N. Butters, D. A. White, D. Kirson, J. H. Atkinson, J. A. McCutchan, M. J. Taylor, M. D. Kelly, R. J. Ellis, et al. 1995. The HNRC 500—neuropsychology of HIV infection at different disease stages. J. Int. Neuropsychol. Soc. 1:231-251.
Ho, D. D., T. R. Rota, R. T. Schooley, J. C. Kaplan, J. D. Allan, J. E. Groopman, L. Resnick, D. Felsenstein, C. A. Andrews, and M. S. Hirsch. 1985. Isolation of HTLV-III from cerebrospinal fluid and neural tissues of patients with neurologic syndromes related to the acquired immunodeficiency syndrome. N. Engl. J. Med. 313:1493-1497.
Hudson, R. R., M. Slatkin, and W. P. Maddison. 1992. Estimation of levels of gene flow from DNA sequence data. Genetics 132:583-589.
Hughes, E. S., J. E. Bell, and P. Simmonds. 1997. Investigation of the dynamics of the spread of human immunodeficiency virus to brain and other tissues by evolutionary analysis of sequences from the p17gag and env genes. J. Virol. 71:1272-1280.
Jost, S., M. C. Bernard, L. Kaiser, S. Yerly, B. Hirschel, A. Samri, B. Autran, L. E. Goh, and L. Perrin. 2002. A patient with HIV-1 superinfection. N. Engl. J. Med. 347:731-736.
Kepler, T. B., and A. S. Perelson. 1998. Drug concentration heterogeneity facilitates the evolution of drug resistance. Proc. Natl. Acad. Sci. USA 95:11514-11519.
Keys, B., J. Karis, B. Fadeel, A. Valentin, G. Norkrans, L. Hagberg, and F. Chiodi. 1993. V3 sequences of paired HIV-1 isolates from blood and cerebrospinal fluid cluster according to host and show variation related to the clinical stage of disease. Virology 196:475-483.
Koelsch, K. K., D. M. Smith, S. J. Little, C. C. Ignacio, T. R. Macaranas, A. J. Brown, C. J. Petropoulos, D. D. Richman, and J. K. Wong. 2003. Clade B HIV-1 superinfection with wild-type virus after primary infection with drug-resistant clade B virus. AIDS 17:F11-F16.
Koenig, S., H. E. Gendelman, J. M. Orenstein, M. C. Dal Canto, G. H. Pezeshkpour, M. Yungbluth, F. Janotta, A. Aksamit, M. A. Martin, and A. S. Fauci. 1986. Detection of AIDS virus in macrophages in brain tissue from AIDS patients with encephalopathy. Science 233:1089-1093.
Korber, B. T., E. E. Allen, A. D. Farmer, and G. Myers. 1995. Heterogeneity of HIV-1 and HIV-2. AIDS 9(Suppl. A):S5-S18.
Korber, B. T., K. J. Kunstman, B. K. Patterson, M. Furtado, M. M. McEvilly, R. Levy, and S. M. Wolinsky. 1994. Genetic differences between blood- and brain-derived viral sequences from human immunodeficiency virus type 1-infected patients: evidence of conserved elements in the V3 region of the envelope protein of brain-derived sequences. J. Virol. 68:7467-7481.
Koup, R. A. 1994. Virus escape from CTL recognition. J. Exp. Med. 180:779-782.
Kuiken, C. L., J. Goudsmit, G. F. Weiller, J. S. Armstrong, S. Hartman, P. Portegies, J. Dekker, and M. Cornelissen. 1995. Differences in human immunodeficiency virus type 1 V3 sequences from patients with and without AIDS dementia complex. J. Gen. Virol. 76:175-180.
Kumar, S., K. Tamura, I. B. Jakobsen, and M. Nei. 2001. MEGA2: molecular evolutionary genetics analysis software. Bioinformatics 17:1244-1245.
Lafeuillade, A., C. Solas, P. Halfon, S. Chadapaud, G. Hittinger, and B. Lacarelle. 2002. Differences in the detection of three HIV-1 protease inhibitors in non-blood compartments: clinical correlations. HIV Clin. Trials 3:27-35.
Lane, T. E., M. J. Buchmeier, D. D. Watry, D. B. Jakubowski, and H. S. Fox. 1995. Serial passage of microglial SIV results in selection of homogeneous env quasispecies in the brain. Virology 212:458-465.
Letendre, S. L., E. V. Capparelli, R. J. Ellis, J. A. McCutchan, and the HIV Neurobehavioral Research Center Group. 2000. Indinavir population pharmacokinetics in plasma and cerebrospinal fluid. Antimicrob. Agents Chemother. 44:2173-2175.
Li, Y., J. C. Kappes, J. A. Conway, R. W. Price, G. M. Shaw, and B. H. Hahn. 1991. Molecular characterization of human immunodeficiency virus type 1 cloned directly from uncultured human brain tissue: identification of replication-competent and -defective viral genomes. J. Virol. 65:3973-3985.
Liu, Y., X. P. Tang, J. C. McArthur, J. Scott, and S. Gartner. 2000. Analysis of human immunodeficiency virus type 1 gp160 sequences from a patient with HIV dementia: evidence for monocyte trafficking into brain. J. Neurovirol. 6(Suppl. 1):S70-S81.
Marra, C. M., D. Lockhart, J. R. Zunt, M. Perrin, R. W. Coombs, and A. C. Collier. 2003. Changes in CSF and plasma HIV-1 RNA and cognition after starting potent antiretroviral therapy. Neurology 60:1388-1390.
Masliah, E., R. K. Heaton, T. D. Marcotte, R. J. Ellis, C. A. Wiley, M. Mallory, C. L. Achim, J. A. McCutchan, J. A. Nelson, J. H. Atkinson, I. Grant, et al. 1997. Dendritic injury is a pathological substrate for human immunodeficiency virus-related cognitive disorders. Ann. Neurol. 42:963-972.
McArthur, J. C., N. Haughey, S. Gartner, K. Conant, C. Pardo, A. Nath, and N. Sacktor. 2003. Human immunodeficiency virus-associated dementia: an evolving disease. J. Neurovirol. 9:205-221.
McArthur, J. C., D. R. McClernon, M. F. Cronin, T. E. Nance-Sproson, A. J. Saah, M. St. Clair, and E. R. Lanier. 1997. Relationship between human immunodeficiency virus-associated dementia and viral load in cerebrospinal fluid and brain. Ann. Neurol. 42:689-698.
McClernon, D. R., R. Lanier, S. Gartner, P. Feaser, C. A. Pardo, M. St. Clair, Q. Liao, and J. C. McArthur. 2001. HIV in the brain: RNA levels and patterns of zidovudine resistance. Neurology 57:1396-1401.
Messam, C. A., and E. O. Major. 2000. Stages of restricted HIV-1 infection in astrocyte cultures derived from human fetal brain tissue. J. Neurovirol. 6(Suppl. 1):S90-S94.
Michael, N. L., G. Chang, P. K. Ehrenberg, M. T. Vahey, and R. R. Redfield. 1993. HIV-1 proviral genotypes from the peripheral blood mononuclear cells of an infected patient are differentially represented in expressed sequences. J. Acquir. Immune Defic. Syndr. 6:1073-1085.
Morris, A., M. Marsden, K. Halcrow, E. S. Hughes, R. P. Brettle, J. E. Bell, and P. Simmonds. 1999. Mosaic structure of the human immunodeficiency virus type 1 genome infecting lymphoid cells and the brain: evidence for frequent in vivo recombination events in the evolution of regional populations. J. Virol. 73:8720-8731.
Nei, M., and S. Kumar. 2000. Molecular evolution and phylogenetics. Oxford University Press, New York, N.Y.
Neuenburg, J. K., H. R. Brodt, B. G. Herndier, M. Bickel, P. Bacchetti, R. W. Price, R. M. Grant, and W. Schlote. 2002. HIV-related neuropathology, 1985 to 1999: rising prevalence of HIV encephalopathy in the era of highly active antiretroviral therapy. J. Acquir. Immune Defic. Syndr. 31:171-177.
Nickle, D. C., D. Shriner, J. E. Mittler, L. M. Frenkel, and J. I. Mullins. 2003. Importance and detection of virus reservoirs and compartments of HIV infection. Curr. Opin. Microbiol. 6:410-416.
Ohagen, A., A. Devitt, K. J. Kunstman, P. R. Gorry, P. P. Rose, B. Korber, J. Taylor, R. Levy, R. L. Murphy, S. M. Wolinsky, and D. Gabuzda. 2003. Genetic and functional analysis of full-length human immunodeficiency virus type 1 env genes derived from brain and blood of patients with AIDS. J. Virol. 77:12336-12345.
Overholser, E. D., G. D. Coleman, J. L. Bennett, R. J. Casaday, M. C. Zink, S. A. Barber, and J. E. Clements. 2003. Expression of simian immunodeficiency virus (SIV) nef in astrocytes during acute and terminal infection and requirement of nef for optimal replication of neurovirulent SIV in vitro. J. Virol. 77:6855-6866.
Pang, S., Y. Koyanagi, S. Miles, C. A. Wiley, H. V. Vinters, and I. S. Y. Chen. 1990. High levels of unintegrated HIV-1 DNA in brain tissue of AIDS dementia patients. Nature 343:85-89.
Pang, S., H. V. Vinters, T. Akashi, W. A. O'Brien, and I. S. Y. Chen. 1991. HIV-1 env sequence variation in brain tissue of patients with AIDS-related neurologic disease. J. Acquir. Immune Defic. Syndr. 4:1082-1092.
Parikh, U., C. Calef, B. Larder, R. F. Schinazi, and J. Mellors. 2001. Mutations in retroviral genes associated with drug resistance, p. 191-278. In C. Kuiken, B. Foley, B. Hahn, P. Marx, F. McCutchan, J. Mellors, S. Wolinsky, and B. Korber (ed.), HIV sequence compendium 2001. Theoretical Biology and Biophysics Group, Los Alamos National Laboratory, Los Alamos, N.Mex.
Pierson, T., J. McArthur, and R. F. Siliciano. 2000. Reservoirs for HIV-1: mechanisms for viral persistence in the presence of antiviral immune responses and antiretroviral therapy. Annu. Rev. Immunol. 18:665-708.
Power, C., J. C. McArthur, R. T. Johnson, D. E. Griffin, J. D. Glass, R. Dewey, and B. Chesebro. 1995. Distinct HIV-1 env sequences are associated with neurotropism and neurovirulence. Curr. Top. Microbiol. Immunol. 202:89-104.
Power, C., J. C. McArthur, R. T. Johnson, D. E. Griffin, J. D. Glass, S. Perryman, and B. Chesebro. 1994. Demented and nondemented patients with AIDS differ in brain-derived human immunodeficiency virus type 1 envelope sequences. J. Virol. 68:4643-4649.
Price, R. W. 1996. Neurological complications of HIV infection. Lancet 348:445-452.
Price, R. W. 2000. The two faces of HIV infection of cerebrospinal fluid. Trends Microbiol. 8:387-391.
Price, R. W., and S. Staprans. 1997. Measuring the "viral load" in cerebrospinal fluid in human immunodeficiency virus infection: window into brain infection? Ann. Neurol. 42:675-678.
Pulliam, L., R. Gascon, M. Stubblebine, D. McGuire, and M. S. McGrath. 1997. Unique monocyte subset in patients with AIDS dementia. Lancet 349:692-695.
Pulliam, L., B. G. Herndier, N. M. Tang, and M. S. McGrath. 1991. Human immunodeficiency virus-infected macrophages produce soluble factors that cause histological and neurochemical alterations in cultured human brains. J. Clin. Investig. 87:503-512.
Reddy, R. T., A. D. Alexa, A. Sirko, S. Tehranchi, F. G. Kraus, F. Wong-Staal, C. A. Wiley, et al. 1996. Sequence analysis of the V3 loop in brain and spleen of patients with HIV encephalitis. AIDS Res. Hum. Retrovir. 12:477-482.
Richman, D. D., T. Wrin, S. J. Little, and C. J. Petropoulos. 2003. Rapid evolution of the neutralizing antibody response to HIV type 1 infection. Proc. Natl. Acad. Sci. USA 100:4144-4149.
Ross, H. L., S. Gartner, J. C. McArthur, J. R. Corboy, J. J. McAllister, S. Millhouse, and B. Wigdahl. 2001. HIV-1 LTR C/EBP binding site sequence configurations preferentially encountered in brain lead to enhanced C/EBP factor binding and increased LTR-specific activity. J. Neurovirol. 7:235-249.
Sacktor, N., P. M. Tarwater, R. L. Skolasky, J. C. McArthur, O. A. Selnes, J. Becker, B. Cohen, and E. N. Miller. 2001. CSF antiretroviral drug penetrance and the treatment of HIV-associated psychomotor slowing. Neurology 57:542-544.
Saksena, N. K., B. Wang, Y. C. Ge, J. Chang, D. E. Dwyer, S. H. Xiang, D. R. Packham, C. Randle, and A. L. Cunningham. 1997. Region-specific changes, gene duplications, and random deletions in the nef gene from HIV type 1-infected brain tissues and blood of a demented patient. AIDS Res. Hum. Retrovir. 13:111-116.
Sei, S., S. K. Stewart, M. Farley, B. U. Mueller, J. R. Lane, M. L. Robb, P. Brouwers, and P. A. Pizzo. 1996. Evaluation of human immunodeficiency virus (HIV) type 1 RNA levels in cerebrospinal fluid and viral resistance to zidovudine in children with HIV encephalopathy. J. Infect. Dis. 174:1200-1206.
Shioda, T., J. A. Levy, and C. Cheng-Mayer. 1992. Small amino acid changes in the V3 hypervariable region of gp120 can affect the T-cell-line and macrophage tropism of human immunodeficiency virus type 1. Proc. Natl. Acad. Sci. USA 89:9434-9438.
Slatkin, M., and W. P. Maddison. 1989. A cladistic measure of gene flow inferred from the phylogenies of alleles. Genetics 123:603-613.
Staprans, S., N. Marlowe, D. Glidden, T. Novakovic-Agopian, R. M. Grant, M. Heyes, F. Aweeka, S. Deeks, and R. W. Price. 1999. Time course of cerebrospinal fluid responses to antiretroviral therapy: evidence for variable compartmentalization of infection. AIDS 13:1051-1061.
Strain, M., H. Gunthard, D. Havlir, C. Ignacio, D. M. Smith, A. J. Leigh Brown, T. Macaranas, R. Y. Lam, O. A. Daly, M. Fischer, M. Opravil, H. Levine, L. Bacheler, C. Spina, D. Richman, and J. K. Wong. 2003. Heterogeneous clearance rates of long-lived lymphocytes infected with HIV: intrinsic stability predicts lifelong persistence. Proc. Natl. Acad. Sci. USA 100:4819-4824.
Strelow, L. I., D. D. Watry, H. S. Fox, and J. A. Nelson. 1998. Efficient infection of brain microvascular endothelial cells by an in vivo-selected neuroinvasive SIVmac variant. J. Neurovirol. 4:269-280.
Tyler, K. L., and J. C. McArthur. 2002. Through a glass, darkly: cerebrospinal fluid viral load measurements and the pathogenesis of human immunodeficiency virus infection of the central nervous system. Arch. Neurol. 59:909-912.
Venturi, G., M. Catucci, L. Romano, P. Corsi, F. Leoncini, P. E. Valensin, and M. Zazzi. 2000. Antiretroviral resistance mutations in HIV-1 RT and protease from paired CSF and plasma samples. J. Infect. Dis. 181:740-745.
Watry, D., T. E. Lane, M. Streb, and H. S. Fox. 1995. Transfer of neuropathogenic simian immunodeficiency virus with naturally infected microglia. Am. J. Pathol. 146:914-923.
Wei, X., J. M. Decker, S. Wang, H. Hui, J. C. Kappes, X. Wu, J. F. Salazar-Gonzalez, M. G. Salazar, J. M. Kilby, M. S. Saag, N. L. Komarova, M. A. Nowak, B. H. Hahn, P. D. Kwong, and G. M. Shaw. 2003. Antibody neutralization and escape by HIV-1. Nature 422:307-312.
Wildemann, B., J. Haas, K. Ehrhart, M. Hahn, H. Wagner, N. Lynen, and B. Storch-Hagenlocher. 1993. In vivo comparison of zidovudine resistance mutations in blood and CSF of HIV-1-infected patients. Neurology 43:2659-2663.
Wiley, C. A., R. D. Schrier, J. A. Nelson, P. W. Lampert, and M. B. Oldstone. 1986. Cellular localization of human immunodeficiency virus infection within the brains of acquired immune deficiency syndrome patients. Proc. Natl. Acad. Sci. USA 83:7089-7093.
Wiley, C. A., V. Soontornniyomkij, L. Radhakrishnan, E. Masliah, J. Mellors, S. A. Hermann, P. Dailey, and C. L. Achim. 1998. Distribution of brain HIV load in AIDS. Brain Pathol. 8:277-284.
Willey, S. J., J. D. Reeves, R. Hudson, K. Miyake, N. Dejucq, D. Schols, E. De Clercq, J. Bell, A. McKnight, and P. R. Clapham. 2003. Identification of a subset of human immunodeficiency virus type 1 (HIV-1), HIV-2, and simian immunodeficiency virus strains able to exploit an alternative coreceptor on untransformed human brain and lymphoid cells. J. Virol. 77:6138-6152.
Wolinsky, S. M., B. T. Korber, A. U. Neumann, M. Daniels, K. J. Kunstman, A. J. Whetsell, M. R. Furtado, Y. Cao, D. D. Ho, and J. T. Safrit. 1996. Adaptive evolution of human immunodeficiency virus-type 1 during the natural course of infection. Science 272:537-542.
Wong, J. K., C. C. Ignacio, F. Torriani, D. Havlir, N. J. S. Fitch, and D. D. Richman. 1997. In vivo compartmentalization of HIV: evidence from the examination of pol sequences from autopsy tissues. J. Virol. 70:2059-2071.
Zhang, K., F. Rana, C. Silva, J. Ethier, K. Wehrly, B. Chesebro, and C. Power. 2003. Human immunodeficiency virus type 1 envelope-mediated neuronal death: uncoupling of viral replication and neurotoxicity. J. Virol. 77:6899-6912.
Zhang, Z., T. Schuler, M. Zupancic, S. Wietgrefe, K. A. Staskus, K. A. Reimann, T. A. Reinhart, M. Rogan, W. Cavert, C. J. Miller, R. S. Veazey, D. Notermans, S. Little, S. A. Danner, D. D. Richman, D. Havlir, J. Wong, H. L. Jordan, T. W. Schacker, P. Racz, K. Tenner-Racz, N. L. Letvin, S. Wolinsky, and A. T. Haase. 1999. Sexual transmission and propagation of SIV and HIV in resting and activated CD4+ T cells. Science 286:1353-1357.
Zink, M. C., K. Suryanarayana, J. L. Mankowski, A. Shen, M. Piatak, Jr., J. P. Spelman, D. L. Carter, R. J. Adams, J. D. Lifson, and J. E. Clements. 1999. High viral load in the cerebrospinal fluid and brain correlates with severity of simian immunodeficiency virus encephalitis. J. Virol. 73:10480-10488.
Zink, W. E., J. Zheng, Y. Persidsky, L. Poluektova, and H. E. Gendelman. 1999. The neuropathogenesis of HIV-1 infection. FEMS Immunol. Med. Microbiol. 26:233-241.(M. C. Strain, S. Letendre)