Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Speelmon, E. C. Articles by McElrath, M. J. Search for Related Content PubMed PubMed Citation Articles by Speelmon, E. C. Articles by McElrath, M. J.

 Previous Article  |  Next Article 

Journal of Virology, March 2006, p. 2463-2471, Vol. 80, No. 5
0022-538X/06/$08.00+0     doi:10.1128/JVI.80.5.2463-2471.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Genetic Association of the Antiviral Restriction Factor TRIM5 with Human Immunodeficiency Virus Type 1 Infection

Emily C. Speelmon,^1,2,3 Devon Livingston-Rosanoff,³ Shuying Sue Li,⁴ Quyen Vu,³ John Bui,³ Daniel E. Geraghty,³ Lue Ping Zhao,⁴ and M. Juliana McElrath^3,5,6*

Medical Scientist Training Program,¹ Molecular and Cellular Biology Program, University of Washington, Seattle, Washington,² Clinical Research,³ Public Health Sciences Divisions, Fred Hutchinson Cancer Research Center, Seattle, Washington,⁴ Departments of Medicine,⁵ Laboratory Medicine, University of Washington School of Medicine, Seattle, Washington⁶

Received 20 July 2005/ Accepted 2 December 2005

Top Abstract Introduction Materials and Methods Results Discussion References

 
The innate antiviral factor TRIM5 restricts the replication of some retroviruses through its interaction with the viral capsid protein, leading to abortive infection. While overexpression of human TRIM5 results in modest restriction of human immunodeficiency virus type 1 (HIV-1), this inhibition is insufficient to block productive infection of human cells. We hypothesized that polymorphisms within TRIM5 may result in increased restriction of HIV-1 infection. We sequenced the TRIM5 gene (excluding exon 5) and the 4.8-kb 5' putative regulatory region in genomic DNA from 110 HIV-1-infected subjects and 96 exposed seronegative persons, along with targeted gene sequencing in a further 30 HIV-1-infected individuals. Forty-eight single nucleotide polymorphisms (SNPs), including 20 with allele frequencies of >1.0%, were identified. Among these were two synonymous and eight nonsynonymous coding polymorphisms. We observed no association between TRIM5 polymorphism in HIV-1-infected subjects and their set-point viral load after acute infection, although one TRIM5 haplotype was weakly associated with more rapid CD4⁺ T-cell loss. Importantly, a TRIM5 haplotype containing the nonsynonymous SNP R136Q showed increased frequency among HIV-1-infected subjects relative to exposed seronegative persons, with an odds ratio of 5.49 (95% confidence interval = 1.83 to 16.45; P = 0.002). Nonetheless, we observed no effect of individual TRIM5 nonsynonymous mutations on the in vitro HIV-1 susceptibility of CD4⁺ T cells. Therefore, any effect of TRIM5 polymorphism on HIV-1 infection in primary lymphocytes may depend on combinations of SNPs or on DNA sequences in linkage disequilibrium with the TRIM5 coding sequence.

Top Abstract Introduction Materials and Methods Results Discussion References

 
Humans demonstrate significant variability in their susceptibility to human immunodeficiency virus type 1 (HIV-1) infection (5, 11, 12, 14, 37) and to disease progression after seroconversion (7, 19, 24, 27, 28). Variation in human genes having innate antiviral function may contribute to these differences through subtle, but significant alterations in gene expression or protein function. Understanding and exploiting the effects of polymorphisms in innate antiviral factors on HIV-1 susceptibility and disease progression may provide essential insights for the development of effective vaccine strategies and novel treatment modalities.

The innate antiviral restriction factor TRIM5 exerts a saturable postentry block to retroviral infection (45). TRIM5 is a member of the tripartite motif family of proteins (36), which is characterized by a RING domain, one or two B-boxes, and a coiled-coil domain. The longest splice isoform of TRIM5, TRIM5, also contains a carboxy-terminal SPRY (B30.2) domain, which largely defines its antiviral activity (39, 46, 51). TRIM5 restricts viral infection in a capsid protein (CA)-dependent manner (17, 29, 48), possibly resulting in intracellular sequestration of the viral core, viral genome degradation or disruption of the uncoating process. In Old World monkey cells, TRIM5 produces a robust replicative block to HIV-1 infection (4, 8, 16, 20, 29, 45). Whereas human TRIM5 (TRIM5_hu) enacts a similarly vigorous postentry restriction against N-tropic murine leukemia virus and equine infectious anemia virus (16, 34, 50), its ability to restrict viruses having HIV-1 CA is insufficient to block infection (16, 45). Furthermore, TRIM5_hu-mediated restriction of HIV-1 infection generally has been characterized in cell lines that overexpress the protein (16, 18, 34, 39, 44, 45). What role, if any, TRIM5_hu might play under normal expression conditions in relevant HIV-1 target cells is not understood.

Given that HIV-1 readily infects most human target cells (CD4⁺ T cells), it is possible that TRIM5_hu does not appreciably restrict HIV-1 under normal conditions of expression. However, rare TRIM5_hu variants with greater anti-HIV-1 activity conceivably may exist. We hypothesized that these TRIM5 variants may be enriched among persons who remain seronegative despite frequent high-risk sexual exposures to HIV-1. Alternatively, the weak restriction of HIV-1 exerted by TRIM5_hu may play a role in determining disease progression or set-point viral load. To define the role of TRIM5_hu in HIV-1 infection, we investigated the hypothesis that genetic variation in the TRIM5 coding sequence or putative regulatory region is associated with altered clinical outcomes such as acquisition of HIV-1 infection, set-point viral load or CD4⁺ T-cell loss after infection, or in vitro indicators such as CD4⁺ T-cell susceptibility to HIV-1 infection. These studies were conducted in primary lymphocytes obtained from 96 exposed seronegative and 140 seropositive volunteers.

Top Abstract Introduction Materials and Methods Results Discussion References

 
Study population. We recruited and enrolled 236 uninfected and HIV-infected study participants from the Seattle area (Table 1). These volunteers were predominantly men having sex with men (MSM) of European-American descent. Ninety-six individuals were HIV-1 high-risk exposed seronegatives (ES), and their enrollment criteria and study procedures have been previously described (14). In addition, we included 140 HIV-1-infected participants in the University of Washington Primary Infection Clinic, whose enrollment criteria have been previously reported (40, 41). These patients were monitored in the clinic at 2-week to 4-month intervals, at which time interim illnesses, use of antiretroviral medications, and CD4⁺ T-cell counts, and HIV-1 RNA levels in plasma were ascertained. The plasma HIV-1 RNA set-point level, determined in 105 antiretroviral therapy-naive volunteers, was defined as the median log₁₀-transformed copies of HIV-1 RNA/ml of plasma between 100 days and 2 years postinfection (32). CD4⁺ T-cell slopes, calculated for 90 antiretroviral therapy-naive volunteers, were defined by using the linear regression of the square-rooted CD4⁺ T-cell counts/ml between 90 and 820 days postinfection.

View this table: [in this window] [in a new window]

TABLE 1. Demographic and genotypic comparison of study groups by HIV-1 serostatus and disease progression

HIV-1-infected and ES volunteers were well matched by ethnicity and CCR5 genotype (Table 1). There were significantly more women in the ES group (P < 0.01). ES subjects (median, 43 years; range, 26 to 69 years) were slightly older than those in the HIV-1-infected group (median, 39 years; range, 27 to 69 years). However, these differences should not be associated with TRIM5 genotype or in vitro HIV-1 susceptibility.

TRIM5 genomic DNA sequencing and analysis. Genomic DNA was isolated from Epstein-Barr virus-immortalized B-cell lymphocyte lines (47) for all study participants (QIAamp DNA kit; QIAGEN, Inc., Valencia, CA). Overlapping PCR primers were designed to cover from 4.8 kb upstream from exon 1 to 170 bp downstream from exon 8, excluding 11.8 kb containing introns 4 and 5 and the 20 bp of exon 5, according to GenBank sequences AF220027 and AK027593. Nucleotides –9688 through 2049 and nucleotides 12490 through 16789 were sequenced. Primer selection was facilitated by the use of the computer program Primer 3 (38) embedded within custom in-house scripts to enable convenient primer design. PCR primer sequences are provided in Table 2. PCR was carried out with 25 cycles of denaturing at 95°C for 20 s, annealing at 50°C for 10 s, and extension at 60°C for 4 min.

View this table: [in this window] [in a new window]

TABLE 2. Primers used to amplify the putative 5' regulatory region, exons, and introns of TRIM5 by PCR

Automated sequence analysis was performed in a 3730xl DNA analyzer (Applied Biosystems, Foster City, CA). Sequence data were acquired and managed by using a data tracking, management, and storage system described elsewhere (13). Shotgun sequence assembly, editing, and SNP detection were performed by using the Phred-Phrap-Consed package (9, 10, 15).

CCR5 genotyping. The HIV-1 coreceptor CCR5 32 genotype (23) was determined by using DNA restriction fragment length polymorphism analysis as previously described (35).

HIV-1_JR-CSF infection assay. HIV-1_JR-CSF was generated by 293T cell transfection (2) with proviral plasmid pYK-JRCSF (contributed by Irvin SY Chen and Yoshio Koyanagi [6] and provided by the NIH AIDS Research and Reference Reagent Program). Primary CD4⁺ T cells were isolated from peripheral blood mononuclear cells by negative selection using magnetic antibody bead separation (CD4⁺ T-cell Isolation Kit; Miltenyi, Auburn, CA) and stimulated by 1.5 µg of phytohemagglutinin (PHA; Remel, Lenexa, KS)/ml for 3 days at 37°C and 5% CO₂. Primary CD4⁺ T lymphoblasts were incubated with virus supernatant (multiplicity of infection = 0.015, 0.003, or 0.0006) for 4 h at 37°C, washed extensively, and cultured in four replicates of 2 x 10⁵ each in HEPES-buffered RPMI 1640 (Gibco, Carlsbad, CA) supplemented with penicillin (50 U/ml; Gibco), streptomycin (50 µg/ml; Gibco), L-glutamine (2 mM; Gibco), 10% heat-inactivated fetal bovine serum (Gemini Bio-Products, Woodland, CA), and 100 U of interleukin-2 (IL-2; Chiron, Emeryville, CA)/ml. At 3, 5, 7, 10, and 12 days postinfection, 100 µl of supernatant was harvested and stored at –70°C for batch HIV-1 quantitation. HIV-1 in supernatants was quantified by p24 enzyme-linked immunosorbent assay (Perkin-Elmer, Boston, MA) according to the manufacturer's guidelines, with results reported as total picograms of p24 antigen present in culture supernatant.

VSV-G pseudotyped HIV-1envGFP infection assay. For single-round infection using a green fluorescent protein (GFP)-expressing pseudotype, vesicular stomatitis virus (VSV) envelope was used to package vector HIV-1_LAIenvGFP (49; kindly provided by M. Emerman, Fred Hutchinson Cancer Research Center, Seattle, WA) by 293T cotransfection. PHA-stimulated CD4⁺ T cells (10⁵) were infected with serial dilutions of viral inoculum by spinoculation (31) for 1 h at 1,900 x g and 30°C in culture medium containing 20 µg of DEAE-dextran/ml. After spinoculation, IL-2 medium was added (final concentration, 50 U/ml), and infection was permitted to progress for 40 to 44 h, after which cells were fixed in 2% paraformaldehyde and assessed for GFP expression on a FACSCalibur flow cytometer (Becton Dickinson) within 24 h of fixation. FlowJo software (Tree Star, Inc., Ashland, OR) was used for analysis.

Statistical analyses. Demographic and genotypic comparison of HIV-1⁺ and ES volunteers was performed by using the Fisher exact test. Log-transformed p24 ELISA and set-point viral load results were assessed by using the Student t test. When no p24 antigen was detectable, 10 pg/ml, representing the limit of detection, was substituted. HIV-1_JR-CSF infection data from CCR532 homozygous individuals were excluded from further analysis, since these individuals served as negative controls in experiments utilizing R5-tropic viruses.

For HIV-1 serostatus and disease progression, volunteer groups were assessed for individual SNP frequency and for the distribution of the minor allele across homozygous or heterozygous genotypes. In both of these analyses, the Fisher exact test was used. Linkage disequilibrium was calculated for each pair of polymorphisms. For haplotype association analyses, we used the logistic regression model previously described (22, 52) for binary phenotypes (serostatus) and a linear regression model for continuous phenotypes (T-cell slope). For these analyses, we used Hplus software that we created for haplotype estimation (http://qge.fhcrc.org/hplus/). In these analyses, individuals' haplotypes are inferred from genotypes by using an estimating equation method, and the uncertainty of inference is incorporated in either the logistic regression or the linear regression estimation. Any association between clinical outcome and the haplotype was tested simultaneously with the haplotype inference. The odd ratios (ORs) are calculated and further tested for significance.

Top Abstract Introduction Materials and Methods Results Discussion References

 
Identification of TRIM5 variants. In order to define potential TRIM5 polymorphisms in our study population (96 ES, 110 HIV-1 infected), we sequenced 14.7 kb on chromosome 11p15 from genomic DNA, between the putative regulatory region 4.8 kb upstream from the first exon and 170 bp downstream from the 3' terminus of exon 8 (Fig. 1). An 11.8-kb sequence containing introns 4 and 5 and the 20 bp exon 5 were excluded. Because the regions encoding the RING, B-box 2 and SPRY domains as well as the N-terminal 40 amino acids of the coiled-coil domain contained the majority of coding SNPs identified, these regions, encoding amino acids 1 through 171 and 350 through 493, were sequenced in an additional 30 HIV-1-infected volunteers (Table 2).

View larger version (15K): [in this window] [in a new window]

FIG. 1. TRIM5 polymorphism. (A and B) Schematic illustrating TRIM5 genetic structure (A), with boxes above the line indicating TRIM5 exons and 48 noncoding SNPs indicated by vertical lines below the line, and TRIM5 protein (B), with functional domains indicated. Ten coding SNPs are illustrated by vertical lines. The position and amino acid change are designated for 8 nonsynonymous SNPs, and asterisks denote synonymous coding SNPs.

These sequence data revealed 48 SNPs (Table 3): 17 were located in the putative 5' regulatory region, 3 were located in the noncoding exon 1, 7 were located in the in exon 2, 2 were located in the in exon 3, 1 was located in the in exon 4, 9 were located in the in exon 8, 9 were located in the within introns, and 1 was located in the downstream to exon 8. No SNPs were detected in exons 6 and 7. Twenty SNPs were present at an allele frequency of >1% (Table 3). Ten coding SNPs were identified, including two synonymous and eight nonsynonymous polymorphisms (Fig. 1). With the exception of H419Y in the SPRY domain, all nonsynonymous SNPs were located within the RING (one nsSNP), B-box 2 (four nsSNPs), and coiled-coil (two nsSNPs) domains encoded within exon 2 (Fig. 1).

View this table: [in this window] [in a new window]

TABLE 3. Characterization and frequencies of 48 TRIM5 polymorphisms detected in exposed seronegative and HIV-1-seropositive study participants

With the exceptions of –4879C/G and 1917A/G, polymorphism distributions for all SNPs identified were in Hardy-Weinberg equilibrium. Because the frequency of the minor variant at these positions is low (Table 3), the Hardy-Weinberg violations observed result from the presence of a single volunteer in each case who was homozygous for that minor allele. This most likely represents a stochastic event rather than a real biological finding. Pairwise calculation of linkage disequilibrium revealed a high degree of linkage disequilibrium across TRIM5 (Fig. 2), increasing the certainty with which haplotypes could be inferred.

View larger version (52K): [in this window] [in a new window]

FIG. 2. TRIM5 SNPs demonstrate high linkage disequilibrium. Pairwise estimates of linkage disequilibrium are shown for all polymorphisms identified. D' = 1 indicates complete disequilibrium. For a pairwise estimate in which D' = 1, log of odds (LOD) of <2 are shown in gray and LOD of 2 are shown in black. For pairwise estimates in which D' is <1, LOD of <2 are shown in white and LOD of 2 are denoted by hatched fields.

Haplotypes of TRIM5 variants. In order to examine the role of longer sequences of DNA containing multiple SNPs, TRIM5 haplotypes were inferred for each study participant (22, 52). To accommodate partial sequence data from 30 HIV-1-infected subjects, haplotypes were inferred separately for the 25 SNPs sequenced in all 236 volunteers and for the 20 SNPs sequenced in 206 volunteers. Haplotypes encompassing 25 SNPs spanning the coding region of TRIM5 are termed "coding haplotypes," whereas haplotypes encompassing 20 SNPs within the putataive 5' regulatory region are termed "regulatory haplotypes." Three SNPs in intron 3 were excluded from these analyses because they were determined for only 206 volunteers but were not contiguous with the SNPs contained within the regulatory haplotype analysis. There were nine common coding haplotypes, accounting for 91% of the ES chromosomes and 89% of the chromosomes in the HIV-1-infected group (Table 4). Three common regulatory haplotypes accounted for 92% of chromosomes in both ES and HIV-1-infected study groups.

View this table: [in this window] [in a new window]

TABLE 4. TRIM5 coding haplotype 9 is associated with HIV-1 serostatus

Association between TRIM5 haplotypes and HIV-1 serostatus. We hypothesized that TRIM5 variants with increased activity against HIV-1 may be more frequent among ES relative to HIV-1-infected subjects. Therefore, we compared TRIM5 SNP and haplotype frequencies for ES and HIV-1⁺ individuals. None of the individual TRIM5 polymorphisms identified was associated with HIV-1 serostatus, either when analyzed by allele frequency (Table 3) or by the propensity of the minor variant to be inherited in either a heterozygous or homozygous genotype (data not shown). However, coding haplotype 9 was markedly enriched among HIV-1-infected individuals relative to ES, with an odds ratio of 5.49 (95% confidence interval [CI] = 1.83 to 16.45; P = 0.002) (Table 4). Coding haplotype 9 differed from the most common coding haplotype (haplotype 1) by the presence of the minor SNP at R136Q. The increased frequency of coding haplotype 9 in HIV-1-infected subjects also was also detected when haplotype 2, which was equally frequent in both HIV-1-infected and ES groups, was used as the referent haplotype, with odds ratio of 3.87 (95% CI = 1.31 to 11.48; P = 0.015). That the association between coding haplotype 9 and HIV-1 serostatus is observed when either haplotype 1 or haplotype 2 is used as reference demonstrates that this association is robust. None of the regulatory haplotypes was associated with HIV-1 serostatus. Of note, we observed an association of HIV-1 serostatus with the TRIM5 coding haplotype 9 but not with the individual TRIM5 SNP R136Q. This may indicate that the effect of TRIM5 on HIV-1 acquisition is dependent on combinations of genetic polymorphisms or that the factor responsible for the observed difference in HIV-1 serostatus is in linkage disequilibrium with this haplotype.

Association between TRIM5 genotype and HIV-1 set-point viral load. We hypothesized that TRIM5 polymorphisms can impact set-point viral load, perhaps by restricting postentry transcription and replication. For 105 HIV-1-infected subjects, we ascertained whether the median level of viral burden after acute infection, or set-point viral load (see Materials and Methods for a definition), was related to individual TRIM5 coding SNPs. Of the 10 coding SNPs identified, 2 synonymous and 7 nonsynonymous polymorphisms were represented among this subset of individuals. We observed no association between any TRIM5 genotype and set-point viral load, suggesting that the TRIM5 variants investigated do not play a major role in determining the steady-state level of HIV-1 production after seroconversion (Fig. 3).

View larger version (31K): [in this window] [in a new window]

FIG. 3. TRIM5 polymorphisms are not associated with altered set-point viral load in HIV-1-infected individuals. Median plasma HIV-1 RNA between 100 days and 2 years postinfection were calculated for each of 105 HIV-1-infected individuals not receiving antiretroviral therapy. Boxes demarcate interquartile range, with medians and 5th and 95th percentiles indicated. For each of nine polymorphic amino acid residues, set-point viral load data are arrayed according to dominant, heterozygous, and homozygous minor genotypes. Asterisks denote synonymous coding SNPs.

Association between TRIM5 SNPs and haplotypes and CD4⁺ T-cell loss. If the weak restriction of HIV-1 by TRIM5_hu affects disease progression, then HIV-1-infected individuals with detrimental TRIM5 polymorphisms may demonstrate more rapid CD4⁺ T-cell loss. For 90 HIV-1-infected subjects, we evaluated whether the slope of CD4⁺ T-cell loss (see Materials and Methods for a definition) was related to TRIM5 coding SNPs or haplotypes. No relationship between any individual TRIM5 coding polymorphism and CD4⁺ T-cell slope was observed (data not shown). Using a linear regression model that incorporated the uncertainty of haplotype inferences, we identified a modestly significant association between the presence of TRIM5 haplotype 4 and faster CD4⁺ T-cell loss (P = 0.021). However, because adjustment for multiple testing abolishes the significance of this finding, the association of TRIM5 haplotype 4 with more rapid disease progression in HIV-1-infected individuals cannot be decisively defined in the small sample size of our population.

Assessment of in vitro HIV-1 production versus TRIM5 genotype. Because immune-mediated control of HIV-1 infection may obscure the effect of TRIM5 variants on CD4⁺ T-cell susceptibility to HIV-1, we examined whether TRIM5 SNPs were associated with altered HIV-1 production after in vitro infection. HIV-1_JR-CSF infection of CD4⁺ T cells from 77 HIV-1 seronegative volunteers permitted investigation of five nonsynonymous and one synonymous SNP. TRIM5 genotype was compared to p24 antigen production at 5 (data not shown) or 7 days (Fig. 4) postinfection. The heterozygous TRIM5 genotype G110/G110E was associated with a marginally significant increase in p24 production at 7 days postinfection (P = 0.048), although the low allele frequency of the G110E variant (1.7%) allowed investigation of only 4 heterozygous individuals for in vitro HIV-1 production. Apart from G110E, no other nsSNPs resulted in altered CD4⁺ T-cell p24 production following HIV-1_JR-CSF infection in vitro. These findings suggest that these 8 SNPs are neither protective nor detrimental in determining the level of HIV-1 p24 production following in vitro infection.

View larger version (20K): [in this window] [in a new window]

FIG. 4. G110E polymorphism moderately associated with increased in vitro CD4+ T-cell HIV-1 production. Viral infectivity was measured by p24 antigen production 7 days after in vitro HIV-1JR-CSF infection of CD4+ T cells from 77 seronegative individuals. Boxes demarcate interquartile range, with medians and 5th and 95th percentiles indicated. For each of six polymorphic amino acid residues, p24 production data are arrayed according to dominant, heterozygous, and homozygous minor genotypes. Asterisks indicate synonymous coding SNPs.

Effect of nonsynonymous TRIM5 polymorphisms on single round infection. We further investigated the effect of individual nonsynonymous TRIM5 polymorphisms on in vitro susceptibility to HIV-1 infection. A GFP-encoding HIV-1_LAI vector pseudotyped with the VSV envelope was used for these experiments, permitting highly sensitive quantitation of HIV-1 infection independent of viral entry and allowing more direct assessment of the postentry cytoplasmic restriction exerted by TRIM5. For each codon tested, volunteers having the minor allele of interest were matched to individuals homozygous for the major variant at that TRIM5 SNP and further matched, in descending order of importance, for HIV-1 serostatus, ethnicity, other TRIM5 coding polymorphisms, treatment status, viral load, and noncoding TRIM5 polymorphisms. For H43Y, V112F, and R136Q, subjects homozygous at the polymorphism in question were investigated (Fig. 5). G110E, R119Q, R119W, and V140L were insufficiently frequent to permit examination of homozygous individuals, and heterozygous subjects were investigated. After infection, no differences in GFP positivity were observed that could be attributed to the presence of the minor variant at any of the codons tested (Fig. 5). These findings suggest that the TRIM5 variants examined do not lead to a discernible change in the efficiency of HIV-1 infection postentry, even during the first round of infection.

View larger version (31K): [in this window] [in a new window]

FIG. 5. H43Y, G110E, V112F, R119Q, R119W, R136Q, and V140L polymorphisms do not affect HIV-1 infection in primary CD4+ T cells. Primary CD4+ lymphoblasts from exposed seronegative and HIV-1-seropositive volunteers were infected with a VSV-G-pseudotyped, GFP-expressing HIV-1 vector (HIV-1LAIenvGFP). (A) Mean frequency of HIV-1 infection in three experiments comparing HIV-1 susceptibility in CD4+ T cells from five H43Y/H43Y (three HIV-1 infected and two ES) and 13 H43/H43 (five HIV-1-infected and eight ES) subjects. (B) Two experiments comparing six G110/G110E (two HIV-1-infected and four ES) and seven G110/G110 (three HIV-1-infected and four ES) individuals. (C) One experiment comparing three V112F/V112F (two HIV-1-infected and one ES) and six V112/V112 (four HIV-1-infected and two ES) individuals. (D) Three experiments comparing three R119/R119Q (one HIV-1-infected and two ES) and one R119/R119W ES subject to eight R119/R119 (two HIV-1-infected and six ES) individuals. (E) Two experiments comparing seven R136Q/R136Q (three HIV-1-infected and four ES) and seven R136/R136 (three HIV-1-infected and four ES) individuals. (F) One experiment comparing three V140/V140L and six V140/V140 HIV-1-infected individuals.

Top Abstract Introduction Materials and Methods Results Discussion References

 
TRIM5 in Old World monkeys blocks HIV-1 infection after entry (4, 8, 16, 45), but human TRIM5 is normally ineffective against HIV-1 (16, 45). Nevertheless, TRIM5_hu has been demonstrated to effect some restriction of HIV-1 infection when expressed exogenously (3, 18, 21, 43, 45, 46). We examined whether TRIM5_hu exerted any effect on HIV-1 infection of relevant target cells when normally expressed. We identified a TRIM5 coding haplotype that is significantly more common among HIV-1-infected persons than ES individuals, with and odds ratio of 5.49 (95% CI = 1.83 to 16.45; P = 0.002). Of interest, no such association between HIV-1 serostatus and TRIM5 polymorphism was detected at the level of individual SNPs. We detected a marginally significant association between TRIM5 coding haplotype 4 and more rapid CD4⁺ T-cell loss (P = 0.021), although no association between any individual coding SNP and set-point viral load could be discerned. Despite the robust relationship between TRIM5 haplotype and HIV-1 serostatus, our in vitro studies reveal no relationship between any single TRIM5 nsSNP and altered primary CD4⁺ T-cell susceptibility to HIV-1 infection, as assessed either by p24 production after multiple rounds of infection or by frequency of infection in a single-round infection assay.

Variation in human genes encoding proteins with innate antiviral function may help define individual susceptibility to acquisition of infection or to disease progression after infection. TRIM5 and other saturable nonimmune antiviral restriction factors, such as APOBEC3F (53) or APOBEC3G (25, 26, 42), may contribute to individual variability in clinical outcome through alterations in gene expression or protein function. Indeed, an association between the APOBEC3G H186R/H186R genotype and more rapid HIV-1 disease progression has been identified (1). Importantly, apart from H419Y, we did not identify any polymorphisms within the TRIM5 SPRY domain. The SPRY domain is largely responsible for defining TRIM5 activity and specificity (30, 33, 46, 51), and polymorphism in human populations within this protein region could explain variability in HIV-1 susceptibility. Nonetheless, the TRIM5 SPRY domain appears to have undergone purifying selection in humans (39a) leading to the relative absence of genetic polymorphism we observe in this region and contributing to the relative inefficiency of HIV-1 restriction by TRIM5.

To date, the majority of studies investigating the role of TRIM5_hu in HIV-1 infection have used stably transduced cell lines expressing elevated levels of TRIM5 (3, 18, 21, 43, 45, 46). Although this molecular tool has been singularly useful in the elucidation of TRIM5 antiviral function, findings observed under exogenous elevation of TRIM5 levels cannot be easily correlated with the situation in vivo. The results of our investigations in primary human CD4⁺ T cells and our study of TRIM5 genotype and haplotype versus relevant clinical outcomes are of particular significance in defining the role normally exerted by TRIM5 during HIV-1 infection. Our observation that TRIM5 variants are not associated with either altered in vitro CD4⁺ T-cell HIV-1 susceptibility or distinct clinical outcome such as disease progression or viral load set-point highlights the importance of investigating in vitro host-pathogen interactions under relevant conditions of infection and calls attention to the formidable hurdle that potential small-molecule therapeutics face in sufficiently inducing or enhancing the antiretroviral activity of TRIM5_hu.

A single TRIM5 coding haplotype was more frequently observed in HIV-1-infected individuals than in ES persons, raising the intriguing possibility that this haplotype is associated with increased risk of HIV-1 acquisition in vivo. This strong enrichment was detected when either of two different haplotypes was used as reference, demonstrating that the association is robust. Coding haplotype 9 differed from the reference haplotype 1 at R136Q. When haplotype 2 was taken as reference, coding haplotype 9 differed at SNP –2 G/C. Given that neither the R136Q nor the –2 G/C polymorphisms are individually associated with serostatus (P = 0.396 and 0.118, respectively), we cannot attribute any difference in HIV-1 susceptibility to these polymorphisms. Further, our in vitro experiments indicate that R136Q does not alter HIV-1 susceptibility in primary CD4⁺ T cells, and overexpression of the R136Q variant in feline CRFK cells similarly fails to affect HIV-1 infection (39a). Therefore, the finding that differences in serostatus are associated with TRIM5 haplotype 9 but not with the individual polymorphisms R136Q or –2 G/C may indicate that the genetic sequence responsible is in linkage disequilibrium with this haplotype, lying in TRIM5 or one of the TRIM genes adjacent to it. There is a possibility that the association of coding haplotype 9 with HIV-1 seropositivity may be spurious, but this is unlikely since the statistical significance of this result is not abolished after Bonferroni correction for multiple comparisons. The effect of coding haplotype 9 needs to be confirmed by sequencing in broader populations, including greater representation of other ethnicities.

Our investigation of the effect of nonsynonymous TRIM5 polymorphisms on in vitro HIV-1 susceptibility suggests that these coding mutations do not alter susceptibility to HIV-1 acquisition or disease directly. One hypothesis is that differences in TRIM5 expression resulting from polymorphisms either contained within or in linkage disequilibrium with TRIM5 coding haplotype 9 increases the possibility of HIV-1 infection in a given exposure. Examination of the levels of TRIM5 expression in primary resting and activated CD4⁺ T cells and monocytes will be critical in defining the role TRIM5 plays in HIV-1 infection in vivo.

 
We are indebted to the volunteers whose participation in this study made these investigations possible. Michael Emerman provided invaluable scientific discussion. We thank Jean Lee and Julie Cooper for volunteer recruitment and Deborah Lee for help with figure design.

This study was supported by National Institutes of Health grants AI47806, AI35605, and AI057005. E.C.S. is supported by National Institutes of Health T32 grants AI007140 and GM0726, funding from the Seattle Chapter of ARCS (Achievement Rewards for College Scientists), and the Poncin Scholarship Fund. M.J.M. is a recipient of the Burroughs Wellcome Clinical Scientist Award for Translational Research.

 
* Corresponding author. Mailing address: Fred Hutchinson Cancer Research Center, 1100 Fairview Ave. N, D3-100, Seattle, WA 98109-1024. Phone: (206) 667-6704. Fax: (206) 667-4411. E-mail: jmcelrat{at}fhcrc.org.

Top Abstract Introduction Materials and Methods Results Discussion References

 

1
1. An, P., G. Bleiber, P. Duggal, G. Nelson, M. May, B. Mangeat, I. Alobwede, D. Trono, D. Vlahov, S. Donfield, J. J. Goedert, J. Phair, S. Buchbinder, S. J. O'Brien, A. Telenti, and C. A. Winkler. 2004. APOBEC3G genetic variants and their influence on the progression to AIDS. J. Virol. 78:11070-11076.[Abstract/Free Full Text]
2
2. Bartz, S. R., and M. A. Vodicka. 1997. Production of high-titer human immunodeficiency virus type 1 pseudotyped with vesicular stomatitis virus glycoprotein. Methods 12:337-342.[CrossRef][Medline]
3
3. Berthoux, L., S. Sebastian, D. M. Sayah, and J. Luban. 2005. Disruption of human TRIM5 antiviral activity by Nonhuman primate orthologues. J. Virol. 79:7883-7888.[Abstract/Free Full Text]
4
4. Besnier, C., Y. Takeuchi, and G. Towers. 2002. Restriction of lentivirus in monkeys. Proc. Natl. Acad. Sci. USA 99:11920-11925.[Abstract/Free Full Text]
5
5. Beyrer, C., A. W. Artenstein, S. Rugpao, H. Stephens, T. C. VanCott, M. L. Robb, M. Rinkaew, D. L. Birx, C. Khamboonruang, P. A. Zimmerman, K. E. Nelson, C. Natpratan, et al. 1999. Epidemiologic and biologic characterization of a cohort of human immunodeficiency virus type 1 highly exposed, persistently seronegative female sex workers in northern Thailand. J. Infect. Dis. 179:59-67.[CrossRef][Medline]
6
6. Cann, A. J., J. A. Zack, A. S. Go, S. J. Arrigo, Y. Koyanagi, P. L. Green, S. Pang, and I. S. Chen. 1990. Human immunodeficiency virus type 1 T-cell tropism is determined by events prior to provirus formation. J. Virol. 64:4735-4742.[Abstract/Free Full Text]
7
7. Cao, Y., L. Qin, L. Zhang, J. Safrit, and D. D. Ho. 1995. Virologic and immunologic characterization of long-term survivors of human immunodeficiency virus type 1 infection. N. Engl. J. Med. 332:201-208.[Abstract/Free Full Text]
8
8. Cowan, S., T. Hatziioannou, T. Cunningham, M. A. Muesing, H. G. Gottlinger, and P. D. Bieniasz. 2002. Cellular inhibitors with Fv1-like activity restrict human and simian immunodeficiency virus tropism. Proc. Natl. Acad. Sci. USA 99:11914-11919.[Abstract/Free Full Text]
9
9. Ewing, B., and P. Green. 1998. Base-calling of automated sequencer traces using phred. II. Error probabilities. Genome Res. 8:186-194.[Abstract/Free Full Text]
10
10. Ewing, B., L. Hillier, M. C. Wendl, and P. Green. 1998. Base-calling of automated sequencer traces using phred. I. Accuracy assessment. Genome Res. 8:175-185.[Abstract/Free Full Text]
11
11. Follezou, J. Y., N. Y. Lan, T. X. Lien, M. E. Lafon, L. T. Tram, P. V. Hung, X. Aknine, W. Lowenstein, N. V. Ngai, I. Theodorou, J. F. Delfraissy, P. Debre, H. J. Fleury, F. Barre-Sinoussi, and N. H. Chi. 1999. Clinical and biological characteristics of human immunodeficiency virus-infected and uninfected intravascular drug users in Ho Chi Minh City, Vietnam. Am. J. Trop. Med. Hyg. 61:420-424.[Abstract]
12
12. Fowke, K. R., N. J. Nagelkerke, J. Kimani, J. N. Simonsen, A. O. Anzala, J. J. Bwayo, K. S. MacDonald, E. N. Ngugi, and F. A. Plummer. 1996. Resistance to HIV-1 infection among persistently seronegative prostitutes in Nairobi, Kenya. Lancet 348:1347-1351.[CrossRef][Medline]
13
13. Geraghty, D. E., S. Fortelny, B. Guthrie, M. Irving, H. Pham, R. Wang, R. Daza, B. Nelson, J. Stonehocker, L. Williams, and Q. Vu. 2000. The data acquisition, data storage, and data presentation in a modern genetics laboratory. Rev. Immunogenet. 2:532-540.[Medline]
14
14. Goh, W. C., J. Markee, R. E. Akridge, M. Meldorf, L. Musey, T. Karchmer, M. Krone, A. Collier, L. Corey, M. Emerman, and M. J. McElrath. 1999. Protection against human immunodeficiency virus type 1 infection in persons with repeated exposure: evidence for T-cell immunity in the absence of inherited CCR5 coreceptor defects. J. Infect. Dis. 179:548-557.[CrossRef][Medline]
15
15. Gordon, D., C. Abajian, and P. Green. 1998. Consed: a graphical tool for sequence finishing. Genome Res. 8:195-202.[Abstract/Free Full Text]
16
16. Hatziioannou, T., S. Cowan, S. P. Goff, P. D. Bieniasz, and G. J. Towers. 2003. Restriction of multiple divergent retroviruses by Lv1 and Ref1. EMBO J. 22:385-394.[CrossRef][Medline]
17
17. Hatziioannou, T., S. Cowan, U. K. Von Schwedler, W. I. Sundquist, and P. D. Bieniasz. 2004. Species-specific tropism determinants in the human immunodeficiency virus type 1 capsid. J. Virol. 78:6005-6012.[Abstract/Free Full Text]
18
18. Hatziioannou, T., D. Perez-Caballero, A. Yang, S. Cowan, and P. D. Bieniasz. 2004. Retrovirus resistance factors Ref1 and Lv1 are species-specific variants of TRIM5. Proc. Natl. Acad. Sci. USA 101:10774-10779.[Abstract/Free Full Text]
19
19. Hendel, H., S. Caillat-Zucman, H. Lebuanec, M. Carrington, S. O'Brien, J. M. Andrieu, F. Schachter, D. Zagury, J. Rappaport, C. Winkler, G. W. Nelson, and J. F. Zagury. 1999. New class I and II HLA alleles strongly associated with opposite patterns of progression to AIDS. J. Immunol. 162:6942-6946.[Abstract/Free Full Text]
20
20. Hofmann, W., D. Schubert, J. LaBonte, L. Munson, S. Gibson, J. Scammell, P. Ferrigno, and J. Sodroski. 1999. Species-specific, postentry barriers to primate immunodeficiency virus infection. J. Virol. 73:10020-10028.[Abstract/Free Full Text]
21
21. Keckesova, Z., L. M. Ylinen, and G. J. Towers. 2004. The human and African green monkey TRIM5 genes encode Ref1 and Lv1 retroviral restriction factor activities. Proc. Natl. Acad. Sci. USA 101:10780-10785.[Abstract/Free Full Text]
22
22. Li, S. S., N. Khalid, C. Carlson, and L. P. Zhao. 2003. Estimating haplotype frequencies and standard errors for multiple single nucleotide polymorphisms. Biostatistics 4:513-522.[Abstract]
23
23. Liu, R., W. A. Paxton, S. Choe, D. Ceradini, S. R. Martin, R. Horuk, M. E. MacDonald, H. Stuhlmann, R. A. Koup, and N. R. Landau. 1996. Homozygous defect in HIV-1 coreceptor accounts for resistance of some multiply-exposed individuals to HIV-1 infection. Cell 86:367-377.[CrossRef][Medline]
24
24. Liu, S. L., T. Schacker, L. Musey, D. Shriner, M. J. McElrath, L. Corey, and J. I. Mullins. 1997. Divergent patterns of progression to AIDS after infection from the same source: human immunodeficiency virus type 1 evolution and antiviral responses. J. Virol. 71:4284-4295.[Abstract]
25
25. Mangeat, B., P. Turelli, G. Caron, M. Friedli, L. Perrin, and D. Trono. 2003. Broad antiretroviral defense by human APOBEC3G through lethal editing of nascent reverse transcripts. Nature 424:99-103.[CrossRef][Medline]
26
26. Marin, M., K. M. Rose, S. L. Kozak, and D. Kabat. 2003. HIV-1 Vif protein binds the editing enzyme APOBEC3G and induces its degradation. Nat. Med. 9:1398-1403.[CrossRef][Medline]
27
27. Martin, M. P., X. Gao, J. H. Lee, G. W. Nelson, R. Detels, J. J. Goedert, S. Buchbinder, K. Hoots, D. Vlahov, J. Trowsdale, M. Wilson, S. J. O'Brien, and M. Carrington. 2002. Epistatic interaction between KIR3DS1 and HLA-B delays the progression to AIDS. Nat. Genet. 31:429-434.[Medline]
28
28. Monceaux, V., J. Estaquier, M. Fevrier, M. C. Cumont, Y. Riviere, A. M. Aubertin, J. C. Ameisen, and B. Hurtrel. 2003. Extensive apoptosis in lymphoid organs during primary SIV infection predicts rapid progression toward AIDS. AIDS 17:1585-1596.[CrossRef][Medline]
29
29. Munk, C., S. M. Brandt, G. Lucero, and N. R. Landau. 2002. A dominant block to HIV-1 replication at reverse transcription in simian cells. Proc. Natl. Acad. Sci. USA 99:13843-13848.[Abstract/Free Full Text]
30
30. Nakajima, N., R. Lu, and A. Engelman. 2001. Human immunodeficiency virus type 1 replication in the absence of integrase-mediated DNA recombination: definition of permissive and nonpermissive T-cell lines. J. Virol. 75:7944-7955.[Abstract/Free Full Text]
31
31. O'Doherty, U., W. J. Swiggard, and M. H. Malim. 2000. Human immunodeficiency virus type 1 spinoculation enhances infection through virus binding. J. Virol. 74:10074-10080.[Abstract/Free Full Text]
32
32. Patke, D. S., S. J. Langan, L. M. Carruth, S. M. Keating, B. P. Sabundayo, J. B. Margolick, T. C. Quinn, and R. C. Bollinger. 2002. Association of Gag-specific T lymphocyte responses during the early phase of human immunodeficiency virus type 1 infection and lower virus load set point. J. Infect. Dis. 186:1177-1180.[CrossRef][Medline]
33
33. Perez-Caballero, D., T. Hatziioannou, A. Yang, S. Cowan, and P. D. Bieniasz. 2005. Human tripartite motif 5 domains responsible for retrovirus restriction activity and specificity. J. Virol. 79:8969-8978.[Abstract/Free Full Text]
34
34. Perron, M. J., M. Stremlau, B. Song, W. Ulm, R. C. Mulligan, and J. Sodroski. 2004. TRIM5 mediates the postentry block to N-tropic murine leukemia viruses in human cells. Proc. Natl. Acad. Sci. USA 101:11827-11832.[Abstract/Free Full Text]
35
35. Quillent, C., E. Oberlin, J. Braun, D. Rousset, G. Gonzalez-Canali, P. Metais, L. Montagnier, J. L. Virelizier, F. Arenzana-Seisdedos, and A. Beretta. 1998. HIV-1-resistance phenotype conferred by combination of two separate inherited mutations of CCR5 gene. Lancet 351:14-18.[CrossRef][Medline]
36
36. Reymond, A., G. Meroni, A. Fantozzi, G. Merla, S. Cairo, L. Luzi, D. Riganelli, E. Zanaria, S. Messali, S. Cainarca, A. Guffanti, S. Minucci, P. G. Pelicci, and A. Ballabio. 2001. The tripartite motif family identifies cell compartments. EMBO J. 20:2140-2151.[CrossRef][Medline]
37
37. Rowland-Jones, S., J. Sutton, K. Ariyoshi, T. Dong, F. Gotch, S. McAdam, D. Whitby, S. Sabally, A. Gallimore, T. Corrah, et al. 1995. HIV-specific cytotoxic T cells in HIV-exposed but uninfected Gambian women. Nat. Med. 1:59-64.[CrossRef][Medline]
38
38. Rozen, S., and H. Skaletsky. 2000. Primer3 on the WWW for general users and for biologist programmers. Methods Mol. Biol. 132:365-386.[Medline]
39
39. Sawyer, S. L., L. I. Wu, M. Emerman, and H. S. Malik. 2005. Positive selection of primate TRIM5alpha identifies a critical species-specific retroviral restriction domain. Proc. Natl. Acad. Sci. USA 102:2832-2837.[Abstract/Free Full Text]
39
40. Sawyer, S. L., L. I. Wu, J. M. Akey, M. Emerman, and H. S. Malik. 2005. High frequency persistence of an impaired allele of the retroviral defense gene TRIM5 in humans. Curr. Biol. 16:95-100.
40
41. Schacker, T., A. C. Collier, J. Hughes, T. Shea, and L. Corey. 1996. Clinical and epidemiologic features of primary HIV infection. Ann. Intern. Med. 125:257-264.[Abstract/Free Full Text]
41
42. Schacker, T. W., J. P. Hughes, T. Shea, R. W. Coombs, and L. Corey. 1998. Biological and virologic characteristics of primary HIV infection. Ann. Intern. Med. 128:613-620.[Abstract/Free Full Text]
42
43. Sheehy, A. M., N. C. Gaddis, J. D. Choi, and M. H. Malim. 2002. Isolation of a human gene that inhibits HIV-1 infection and is suppressed by the viral Vif protein. Nature 418:646-650.[CrossRef][Medline]
43
44. Song, B., H. Javanbakht, M. Perron, H. Park do, M. Stremlau, and J. Sodroski. 2005. Retrovirus restriction by TRIM5 variants from Old World and New World primates. J. Virol. 79:3930-3937.[Abstract/Free Full Text]
44
45. Stranford, S. A., J. Skurnick, D. Louria, D. Osmond, S. Y. Chang, J. Sninsky, G. Ferrari, K. Weinhold, C. Lindquist, and J. A. Levy. 1999. Lack of infection in HIV-exposed individuals is associated with a strong CD8⁺ cell noncytotoxic anti-HIV response. Proc. Natl. Acad. Sci. USA 96:1030-1035.[Abstract/Free Full Text]
45
46. Stremlau, M., C. M. Owens, M. J. Perron, M. Kiessling, P. Autissier, and J. Sodroski. 2004. The cytoplasmic body component TRIM5 restricts HIV-1 infection in Old World monkeys. Nature 427:848-853.[CrossRef][Medline]
46
47. Stremlau, M., M. Perron, S. Welikala, and J. Sodroski. 2005. Species-specific variation in the B30.2(SPRY) domain of TRIM5 determines the potency of human immunodeficiency virus restriction. J. Virol. 79:3139-3145.[Abstract/Free Full Text]
47
48. Tosato, G. 1992. Generation of Epstein-Barr virus (EBV)-immortalized B-cell lines, p.7.22.1-7.22.3. In J. E. Coligan, A. M. Kruisbeek, D. H. Marguilles, E. M. Shevach, and W. M. Strober (ed.), Current protocols in immunology. Greene and Wiley-Interscience, Toronto, Ontario, Canada.
48
49. Towers, G., M. Bock, S. Martin, Y. Takeuchi, J. P. Stoye, and O. Danos. 2000. A conserved mechanism of retrovirus restriction in mammals. Proc. Natl. Acad. Sci. USA 97:12295-12299.[Abstract/Free Full Text]
49
50. Yamashita, M., and M. Emerman. 2004. Capsid is a dominant determinant of retrovirus infectivity in nondividing cells. J. Virol. 78:5670-5678.[Abstract/Free Full Text]
50
51. Yap, M. W., S. Nisole, C. Lynch, and J. P. Stoye. 2004. Trim5 protein restricts both HIV-1 and murine leukemia virus. Proc. Natl. Acad. Sci. USA 101:10786-10791.[Abstract/Free Full Text]
51
52. Yap, M. W., S. Nisole, and J. P. Stoye. 2005. A single amino acid change in the SPRY domain of human Trim5 leads to HIV-1 restriction. Curr. Biol. 15:73-78.[CrossRef][Medline]
52
53. Zhao, L. P., S. S. Li, and N. Khalid. 2003. A method for the assessment of disease associations with single-nucleotide polymorphism haplotypes and environmental variables in case-control studies. Am. J. Hum. Genet. 72:1231-1250.[CrossRef][Medline]
53
54. Zheng, Y. H., D. Irwin, T. Kurosu, K. Tokunaga, T. Sata, and B. M. Peterlin. 2004. Human APOBEC3F is another host factor that blocks human immunodeficiency virus type 1 replication. J. Virol. 78:6073-6076.[Abstract/Free Full Text]

---

Journal of Virology, March 2006, p. 2463-2471, Vol. 80, No. 5
0022-538X/06/$08.00+0     doi:10.1128/JVI.80.5.2463-2471.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.

This article has been cited by other articles:

• Goldschmidt, V., Ciuffi, A., Ortiz, M., Brawand, D., Munoz, M., Kaessmann, H., Telenti, A. (2008). Antiretroviral Activity of Ancestral TRIM5{alpha}. J. Virol. 82: 2089-2096 [Abstract] [Full Text]  
• Newman, R. M., Hall, L., Connole, M., Chen, G.-L., Sato, S., Yuste, E., Diehl, W., Hunter, E., Kaur, A., Miller, G. M., Johnson, W. E. (2006). Balancing selection and the evolution of functional polymorphism in Old World monkey TRIM5{alpha}. Proc. Natl. Acad. Sci. USA 103: 19134-19139 [Abstract] [Full Text]  
• Ohkura, S., Yap, M. W., Sheldon, T., Stoye, J. P. (2006). All Three Variable Regions of the TRIM5{alpha} B30.2 Domain Can Contribute to the Specificity of Retrovirus Restriction.. J. Virol. 80: 8554-8565 [Abstract] [Full Text]  
• Emerman, M. (2006). How TRIM5{alpha} defends against retroviral invasions. Proc. Natl. Acad. Sci. USA 103: 5249-5250 [Full Text]  

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Speelmon, E. C. Articles by McElrath, M. J. Search for Related Content PubMed PubMed Citation Articles by Speelmon, E. C. Articles by McElrath, M. J.