According to the World Health Organization (WHO), there were 2.1 million deaths due to HIV and AIDS last year alone and close to 35 million people still live with HIV worldwide . There remains no cure for HIV-1 and the development of an effective vaccine continues to elude researchers. Whereas current antiretroviral therapies can greatly extend the life expectancy of those living with HIV/AIDS, such treatments are expensive and remain largely inaccessible in developing countries. With the knowledge of natural immunity to HIV-1 constantly growing, it is becoming clear that intrinsic immune responses play an important role and may hold the key to an effective and accessible prophylactic .
A recently discovered gene, TRIM5 (chromosome 11), has the ability to provide innate protection against retroviruses in the form of TRIM5α, a splice variant of TRIM5 [3–7]. Studies have shown that TRIM5α of old world monkeys, such as the Rhesus Macaque, is able to effectively inhibit HIV-1 infection [4,5]. Whereas human TRIM5α does not completely inhibit HIV-1 infection, it has shown slight anti-HIV activity and is known to restrict other retroviruses, such as the N-tropic murine-leukemia virus [3,5,6]. TRIM5α has highly variable regions and restricts retroviral infection in a species-specific manner [3–8]. The B30.2 SPRY domain at the C-terminus of TRIM5α is the most variable region and has been shown to be responsible for capsid recognition and binding [9–13]. The coiled-coil is thought to be involved in TRIM5α multimerization and is essential for effective retroviral restriction [9,10,12–14]. It has been suggested that the RING and B-box domains are nonessential for capsid binding and viral restriction. The term ‘effecter’ has been applied to the RING and B-box regions as they enhance TRIM5α antiretroviral activity through possible interactions with other proteins [9,12,15]. The RING domain has been shown to affect cytoplasmic levels and distribution of TRIM5α  and act as an E3 ubiquitin ligase, targeting the virus for proteasomal degradation [12,15].
Previous studies have shown, with conflicting results, that certain polymorphisms may alter the potency of human TRIM5α against HIV-1 [3,17–19]. Two single-nucleotide polymorphisms (SNPs) in particular have been at the center of many investigations: a G to A change in rs10838525 (R136Q in exon 2) and a C to T change in rs3740996 (H43Y in exon 2). A study by Javanbakht et al.  found the amino acid changes H43Y and R136Q associated with resistance to HIV-1 infection. Another study showed that 136Q associated with increased HIV-1 infection in a European population . Neither of these studies found any association between TRIM5α SNPs and disease progression; however, Van Manen et al.  found 136Q correlated with slower disease progression and 43Y correlated with accelerated progression. These discrepancies suggest a need for further investigation into the effect human TRIM5α SNPs have on HIV-1 infection. This study set out to characterize genetic variations of TRIM5α within the Pumwani cohort at exon 2. Exon 2 contains the majority of previously identified nonsynonymous coding SNPs and makes up a large part of the functional protein (Fig. 1) [3,18]. This locus was also chosen due to the conflicting results obtained by previous studies concerning the effect of two specific SNPs on HIV-1 infection: H43Y and R136Q [3,17].
For over 20 years the HIV-1 infections of women enrolled in the Pumwani sex worker cohort in Nairobi, Kenya, have been closely monitored. Within the cohort, a group of women remain HIV-1-seronegative despite heavy exposure to HIV-1 through active sex work . To study the role of TRIM5α polymorphisms in this observed natural resistance to HIV-1 infection, 1032 women of the Pumwani cohort were genotyped at exon 2. The SNP and haplotype frequencies in the cohort were investigated and correlations with HIV-1 resistance or susceptibility and disease progression were reported. Our findings suggest that TRIM5α polymorphisms play an important role in HIV-1 infection and the SNP causing the amino acid change to 136Q might play a key role in determining natural resistance to HIV-1 in the Pumwani sex worker cohort.
The study population consisted of 1032 women enrolled in the Pumwani sex worker cohort, established in 1985 in Nairobi, Kenya. Cohort design and follow-up have been discussed elsewhere . The overall HIV-1 infection rate in the Pumwani sex worker cohort is greater than 73.7%. Women were classified as HIV-1-resistant if they remained HIV-seronegative and PCR-negative for at least 3 years while continuing active sex work and were negative at the time of this study. The 88 women who were classified as resistant in this study were all enrolled in the cohort before 1999 with an average follow-up time of 9.6 ± 4.3 years. The HIV-1-negative women who enrolled later and therefore had a shorter follow-up time were not included in the comparison between resistant women and positive women. HIV-1-negative women enrolled after 2001 were not included in the analysis for resistance or susceptibility to HIV-1 infection and all the HIV-1-infected women were included in the analysis for CD4+ T-cell decline to below 200 cells/μl. Informed consent was obtained from all women enrolled in the cohort. The ethics committees of the University of Manitoba and the University of Nairobi have approved this study.
DNA samples were isolated from whole blood and peripheral blood mononuclear cells (PBMCs) using QIAamp DNA Mini Kit (QIAGEN Inc., Mississauga, ON). A sequence-based genotyping method was used for TRIM5α exon 2 genotyping. TRIM5α exon 2 was amplified using a primer pair known to have successfully amplified exon 2 of human TRIM5α in previous studies . PCR products were sequenced and analyzed for nucleotide variations using computer software CodonExpress. These nucleotide polymorphisms were recorded and grouped into haplotypes for analysis.
TRIM5α SNP and haplotype frequencies were determined using PyPop-32-0.6.0. Associations with HIV-1-positive or resistant women in the cohort were analyzed using SPSS 13.0 for Windows (SPSS Inc., Chicago, Illinois, USA). Standard univariate methods such as the Fisher's exact test (P value) as well as Pearson chi-square analysis [odds ratio (OR) as well as Peto OR, 95% confidence interval (CI)] were utilized to determine the relationship between binary outcomes and explanatory variables. Kaplan–Meier analysis with log-rank test was used to examine the time to seroconversion for enrollees who were HIV-1-negative at enrolment. Only women enrolled in the cohort prior to 2001 were included in the survival analysis for HIV-1 infection. Since we have not retained blood samples for TRIM5α genotyping from all women enrolled during the first few years of the project, the probability of being TRIM5α-genotyped may depend on an individual's HIV status and duration of follow-up; therefore this selection mechanism had to be adjusted. According to the Hurvitz-Thompson theory, observations must be weighted inversely to the probability of being included in a sample to reach an unbiased estimate of what is being investigated . We estimated the probability of being typed using logistic regression with follow-up and HIV status as covariables. The inverse of this probability for each woman, after standardization, was then used as a sample weight. We generated a weighted parameter using logistic regression taking into account patient enrolment and samples being typed. We used this parameter to adjust for cross-tab analyses and to compare the results with/without using this parameter. Associations were confirmed using binary logistic regression (backward Wald). SNPs or haplotypes found to be significant at the P < 0.05 level were tested for independent significance using binary logistic regression and Cox regression. Kaplan–Meier survival analysis (log-rank test) was used to examine time to seroconversion and significant correlations were verified by Cox regression analysis. P values were adjusted for multiple comparisons by means of the Bonferroni method using an SPSS syntax created by David Nichols of SPSS.
TRIM5α single-nucleotide polymorphisms identified in the Pumwani sex worker cohort
Thirteen SNPs within the 1 kb genomic fragment containing exon 2 of TRIM5α were identified in the Pumwani sex worker cohort (Table 1). Among them, nine SNPs are within exon 2 (seven nonsynonymous and two synonymous), three SNPs are in intron 2, and one SNP is in the 5′ UTR. Of the coding SNPs identified, three are located in the RING, three in the B-box and two in the coiled-coil domains (Fig. 1). Four of the 13 SNPs observed in this study had not been previously identified; these polymorphisms appeared at very low frequencies (<1%). All four new SNPs were identified within exon 2 of TRIM5α at bases 404 (codon 49), 505 (codon 83), 616 (codon 120) and 713 (codon 152) of the mRNA (Fig. 1). All sequences identified with novel SNPs can be accessed through Genbank (accession numbers FJ561487 to FJ561496).
TRIM5α SNP and exon 2 haplotype frequencies
The only TRIM5α polymorphism found to be present in more than 5% of the population corresponded to the R136Q amino acid change. The three most common TRIM5α SNPs in the Pumwani sex worker cohort are those resulting in the amino acid changes R136Q (10.8%), H43Y (4.84%) and V112F (1.21%).
Single-nucleotide polymorphism distributions among different populations were compared using data provided by the Hapmap project. The three most common SNPs in the Pumwani cohort (2n = 2046) were rs10838525 (R136Q), rs3740996 (H43Y) and rs11601507 (V112F). These frequencies were compared to a Yoruban population from Nigeria (YRI; 2n = 120), Chinese (CHB; 2n = 90) and Japanese (JPN; 2n = 88) populations and a Caucasian population (CEU; 2n = 120). The SNP distributions of the Pumwani cohort and the Nigerian population are quite similar with the Nigerian population showing the following frequencies: rs10838525 at 8.30%, rs3740996 at 6.70% and rs11601507 at 0%. There are substantial differences between the TRIM5α SNP frequencies in the Pumwani cohort and those in the Caucasian and Asian populations. The Caucasian population showed higher percentages of SNPs with 34.20% rs10838525, 13.30% rs3740996 and 10.50% rs11601507. TRIM5α SNP distributions in the Chinese and Japanese populations also displayed variation with frequencies as follows: rs10838525 at 4.40% (CHB) and 2.30% (JPN), rs3740996 at 12.20% (CHB) and 19.30% (JPN) and rs11601507 at 12.50% (CHB) and 9.30% (JPN).
We analyzed TRIM5α exon 2 haplotypes observed in the Pumwani sex worker cohort. A total of 12 different haplotypes were observed based on the variations of exon 2 sequences (Table 2). Only SNPs within exon 2 were included in TRIM5α haplotypes. Intron SNPs were excluded since no reliable pattern of association was observed between these SNPs and those within the exon haplotypes. Wild-type TRIM5α was the most common haplotype, presenting in 82.5% of the population. The three most common haplotypes were new2 (containing 136Q) at 10.4%, new1 (containing 43Y) at 4.4% and new4 (containing 112F) at 1.2%.
Association of TRIM5α single-nucleotide polymorphisms and haplotypes with resistance or susceptibility to HIV-1 infection
TRIM5α SNPs were tested to identify associations with susceptibility or resistance to infection by HIV-1 using cross-tab analysis. Most SNPs were identified at a frequency well below 3% and did not show any significant correlations. The two SNPs identified at frequencies above 3% were SNPs rs3740996/T (H43Y) at 4.84% and rs10838525/A (R136Q) at 10.8%. The SNP rs3740996/T (43Y), enriched in HIV-1-negative individuals in previous studies , was evenly distributed between HIV-1-positive and resistant women in our cohort. Whereas studies have shown a SNP causing the amino acid change H43Y to be associated with protection against HIV-1 , no correlation was observed in the Pumwani cohort (Table 1). The SNP-induced amino acid change R136Q had been associated with both protection  and susceptibility  to HIV-1 in previous studies. A significant difference in the allele and phenotype frequency distributions of this SNP between HIV-1-positive and resistant women was observed in the Pumwani cohort. 136Q is strongly correlated with resistance to HIV-1 infection with an adjusted P value of 1.325E-04 (Bonferroni) and OR of 2.991 (Table 1). These results were upheld through analysis with binary logistic regression (Table 3a).
TRIM5α haplotypes were also analyzed for any associations with HIV-1 resistance or susceptibility (Table 2). Haploytype new2 was enriched in resistant women (34.1%) when compared to HIV-1-positive women (14.7%) with an adjusted P value of 1.480E-04 and OR of 2.991. Binary logistic regression confirmed these results (Table 3). Most observed haplotypes were exceedingly rare which made it difficult to determine if any significant distribution differences existed between the two groups of women. Alternatively, the effect of haplotypes as a whole was investigated by combining all individuals with polymorphic TRIM5α as compared to individuals with wild-type TRIM5α. Indeed, wild-type TRIM5α was more common in HIV-1-positive individuals (97.6%) than resistant individuals (92.0%), suggesting that any polymorphism may be advantageous. These associations remained significant after correcting for multiple comparisons (Table 2). Sensitivity analysis, using weighted analysis yielded associations that were consistent with the results obtained using nonweighted analyses.
Association of TRIM5α single-nucleotide polymorphisms and haplotypes with reduced or increased risk of HIV-1 seroconversion
The influence of TRIM5α SNPs and haplotypes on the risk of seroconversion was assessed by Kaplan–Meier survival analysis. Most SNPs and haplotypes appeared at frequencies below 3% and did not reveal any significant associations. The SNP conferring an amino acid change H43Y and exon 2 haplotype new1 were shown to have no effect on the risk of seroconversion (Fig. 2a,c). 136Q and haplotype new2 significantly decreased the risk of seroconversion (Fig. 2b,d). Women who are homozygous for 136Q, or haplotype new2, had not yet seroconverted at the time of this study (Fig. 2b,d). These significant associations were also supported by Cox regression analysis (Table 3).
To analyze the combined effect of TRIM5α polymorphisms on seroconversion, a Kaplan–Meier survival plot was generated for wild-type TRIM5α. Women who are homozygous for wild-type TRIM5α showed an increased risk of seroconversion when compared to women who are heterozygous for wild-type TRIM5α or have no wild-type TRIM5α (Fig. 2e).
Association of TRIM5α single-nucleotide polymorphisms and haplotypes with disease progression
A previous study had discovered an association between SNP H43Y and accelerated disease progression . Using Kaplan–Meier survival analysis, HIV-1-positive women with various SNPs and haplotypes were compared based on the number of days with CD4 cell counts above 200 cells/μl (data not shown). No significant correlations were found between any individual SNPs or haplotypes and the rate of disease progression.
The long-term monitoring of women in the Pumwani sex worker cohort, particularly the highly exposed persistently seronegative individuals, provides an excellent opportunity to study the natural resistance to HIV-1. Through analysis of TRIM5α exon 2 sequences of 88 HIV-1-resistant and 468 HIV-1-infected sex workers in the Pumwani sex worker cohort, we have identified a strong correlation between the SNP causing the amino acid change R136Q in the coiled-coil domain (Table 1) and HIV-1 resistance. No significant associations were found between other TRIM5α SNPs in the region examined and HIV-1 resistance/susceptibility or disease progression.
The haplotype new2, with only one nucleotide difference from the wild-type (R136Q), also significantly associated with HIV-1 resistance (Table 2). Women who were HIV-1-negative at cohort enrolment and have 136Q and haplotype new2 seroconverted significantly slower than those who do not have 136Q and haplotype new2 (Fig. 2b,d). Multivariate analysis confirmed that both 136Q and haplotype new2 were associated with resistance and a decreased risk of seroconversion independent from other SNPs and haplotypes in the region (Table 3b). These findings are consistent with the results of Javanbakht et al.  that R136Q is a protective polymorphism in HIV-1 infection. It is important to note that the association between 136Q and haplotype new2 with resistance against infection may be the result of linkage disequilibrium with unidentified SNP(s) in relatively close proximity to codon 136. Additional sequencing and analysis should be conducted to determine whether or not these associations are independent of other sequence variations.
Codon 136 is located within the coiled-coil domain of TRIM5α, which is required for effective recognition and binding of HIV-1 [9,12]. It has also been suggested that the coiled-coil is needed for multimerization of TRIM5α particles, allowing effective viral binding [12,14,16,22]. It is possible that variations in the amino acid sequence of the coiled-coil may alter multimerization and, in turn, the affinity of viral binding to the protein surface. Of note, most nonhuman primate TRIM5α (excluding chimpanzees) encodes a glutamine at codon 136 [18,11]. Many of these nonhuman forms of TRIM5α have been proven to effectively restrict HIV-1 [3–5,11]. Thus, a switch from arginine to glutamine at codon 136 in human TRIM5α could indicate a shift towards a protein that is more active against HIV-1.
Single-nucleotide polymorphism frequencies in the Pumwani cohort were similar to that of the Yoruban population from Nigeria, but quite different from those in other populations (Chinese, Japanese, Caucasian). This variation in frequency could be an important factor when considering the innate immunity of a population, particularly given the variation of the SNP encoding R136Q: 34.2% in Caucasian populations and only 4.4 and 2.3% in Chinese and Japanese populations, respectively. If 136Q exhibits a protective effect in all populations, the frequency of the SNP could have huge implications in the case of HIV-1 transmission.
Whereas Javanbakht et al.  found a connection between 43Y and protection against HIV-1 infection, we found no significant correlations between 43Y and HIV-1 resistance/susceptibility (Table 1) or risk of seroconversion (Fig. 2a,c) in the Pumwani sex worker cohort.
This study has not identified associations between TRIM5α SNPs or haplotypes and HIV-1 disease progression based on CD4 T-cell counts (data not shown). These results conflict with a recent study by Van Manen et al.  in which 43Y was found to be associated with accelerated disease progression. The same study also suggested a possible protective effect of 136Q in disease progression after the emergence of CXCR4-using HIV-1 variants, a factor we have not taken into consideration. However, on the basis of what is known of the mechanism of TRIM5α restriction, an effect on disease progression may be unexpected. TRIM5α particles able to recognize and restrict HIV-1 do so potently, inhibiting the establishment of an infection. A TRIM5α variant which is unable to prevent infection may not be efficient at recognizing or binding to HIV-1 and thus would have a limited effect on disease progression.
The ability of TRIM5α to restrict HIV-1 infection is likely dependent on the mode of viral transmission whether it is cell-free or cell-associated. There is convincing evidence that rhesus TRIM5α restriction of HIV-1 is sensitive to the mode of viral transmission; cell-free transmission is efficiently blocked by rhesus TRIM5α, whereas the extent of cell-associated restriction is much lower . Cell-associated transmission has been shown to be the principle mode of infection during HIV disease progression . Thus, restriction of the initial cell-free viral infection by the appropriate form of human TRIM5α may be possible, whereas cell-associated transmission during disease progression may remain largely uncontrolled.
Whereas altering levels of TRIM5α production during the course of an infection may have an effect on the degree of viral restriction, this has not yet been observed to link directly to viral load . However, there is evidence for interferon-induced TRIM5α transcription [25,26], raising the possibility that an HIV-1 infection could trigger more effective HIV-1 restriction through large quantities of TRIM5α production.
The specific mechanism of TRIM5α-mediated restriction remains largely unknown and it is likely that many factors are involved. Further investigation is required to enhance our understanding of TRIM5α-mediated resistance to HIV-1 infection. Analysis of the full genomic sequence of TRIM5α new2 haplotype may further clarify protective TRIM5α polymorphisms.
The work was supported by the National Institute of Health, The Canadian Institutes of Health Research, National Microbiology Laboratory of Canada and The Bill and Melinda Gates Foundation. We thank Tony Kariri for maintaining the databases of both cohorts at the University of Nairobi and Bing-hua Liang for managing the database at the University of Manitoba. We also thank the staff of the Majengo Clinic, Jane Njoki, Jane Makene, Elazabeth Bwibo, Edith Amatiwa; and the women of the Pumwani sex worker cohort for their continued participation and support. F.A.P. is a Canadian Institutes of Health Research Senior Investigator and is currently a Tier I CIHR Canada Research Chair.
H.P.: Data generation, analysis and interpretation, as well as drafting of the manuscript.
P.L.: Assisted with data generation, analysis and interpretation, as well as editing of the manuscript.
J.T.: Assisted with data generation, analysis and interpretation.
C.W.: Maintained the Pumwani sex worker cohort and was involved in the acquisition of data.
J.K.: Maintained the Pumwani sex worker cohort and was involved in the acquisition of data.
M.L.: Helped to secure funding. Conceived and designed the study. Involved with analysis and interpretation of data, as well as editing of the manuscript.
T.B.B.: Helped to secure funding, maintained the Pumwani sex worker cohort and editing of the manuscript.
F.A.P.: The overall principal investigator. Secured funding for the study. Established and maintained the Pumwani sex worker cohort and was involved in the acquisition of data.
2. Bieniasz PD. Intrinsic immunity: a front-line defense against viral attack. Nature Immunol 2004; 5:1109–1115.
3. Speelmon EC, Livingston-Rosanoff D, Li SS, Vu Q, Bui J, Geraghty DE, et al
. Genetic association of the antiviral restriction factor TRIM5α with human immunodeficiency virus type 1 infection. J Virol 2006; 80:2463–2471.
4. Stremlau M, Owens CM, Perron MJ, Kiessling M, Autissier P, Sodroski J. The cytoplasmic body component TRIM5α restricts HIV-1 infection in old world monkeys. Lett Nature 2004; 427:848–853.
5. Yap MW, Nisole S, Lynch C, Stoye JP. Trim5( protein restricts both HIV-1 and murine leukemia virus. Proc Natl Acad Sci U S A 2004; 101:10786–10791.
6. Perron MJ, Stremlau M, Song B, Ulm W, Mulligan RC, Sodroski J. TRIM5alpha mediates the postentry block to n-tropic murine leukemia viruses in human cells. Proc Natl Acad Sci U S A 2004; 101:1827–11832.
7. Lee K, KewalRamani VN. In defense of the cell: TRIM5α interception of mammalian retroviruses. Proc Natl Acad Sci U S A 2004; 101:10496–10497.
8. Sokolskaja E, Berthoux L, Luban J. Cyclophilin A and TRIM5α independently regulate human immunodeficiency virus type 1 infectivity in human cells. J Virol 2006; 80:2855–2862.
9. Stremlau M, Perron M, Lee M, Li M, Song B, Javanbakht H, et al
. Specific recognition and accelerated uncoating of retroviral capsids by the TRIM5α restriction factor. Proc Natl Acad Sci U S A 2006; 103:5514–5519.
10. Ohkura S, Yap MW, Sheldon T, Stove JP. All three variable regions of the TRIM5alpha B30.3 domain can contribute to the specificity of retrovirus restriction. J Virol 2006; 80:8554–8565.
11. Song B, Gold B, O'Huigin C, Javanbakht H, Li X, Stremlau M, et al
. The B30.2(SPRY) domain of the retroviral restriction factor TRIM5alpha exhibits lineage-specific length and sequence variation in primates. J Virol 2005; 79:6111–6121.
12. Perez-Cabellero D, Hatziioannou T, Yang A, Cowan S, Bieniasz PD. Human tripartite motif 5( domains responsible for retrovirus restriction activity and specificity. J Virol 2005; 79:8969–8978.
13. Yap MW, Nisole S, Stoye JP. A single amino acid change in the SPRY domain of human Trim5α leads to HIV-1 restriction. Curr Biol 2005; 15:73–78.
14. Mische CC, Javanbakht H, Song B, Diaz-Griffero F, Stremlau M, Strack B, et al
. Retroviral restriction factor TRIM5alpha is a trimer. J Virol 2005; 79:14446–14450.
15. Wu X, Anderson JL, Campbell EM, Joseph AM, Hope TJ. Proteasome inhibitors uncouple rhesus TRIM5alpha restriction of HIV-1 reverse transcription and infection. Proc Natl Acad Sci U S A 2006; 103:7465–7470.
16. Javanbakht H, Diaz-Griffero F, Stremlau M, Si Z, Sodroski J. The contribution of the ring and b-box 2 domains to retroviral restriction mediated by monkey TRIM5α. J Biol Chem 2005; 280:26933–26940.
17. Van Manen D, Rits MAN, Beugeling C, Van Dort K, Schuitemaker H, Kootstra NA. The effect of Trim5 polymorphisms on the clinical course of HIV-1 infection. PloS Pathog 2008; 4:1–8.
18. Javanbakht H, An P, Gold B, Peterson DC, O'Huigin C, Nelson GW, et al
. Effects of human TRIM5α polymorphisms on antiretroviral function and susceptibility to human immunodeficiency virus infection. Virology 2006; 354:15–27.
19. Nakayama EE, Carpentier W, Costagliola D, Shioda T, Iwamoto A, Debre P, et al
. Wild type and H43Y variant of human TRIM5alpha show similar anti-human immunodeficiency virus type 1 activity both in vivo and in vitro. Immunogenetics 2007; 59:511–515.
20. Fowke KR, Nagelkerke NJ, Kimani J, Simonsen JN, Anzala AO, Bwayo JJ, et al
. Resistance to HIV-1 infection among persistently seronegative prostitutes in Nairobi, Kenya. Lancet 1996; 348:1347–1351.
21. Xue H, Yang G. Weights in Horvitz-Thompson statistic for complex samples. Wei Sheng Yan Jiu 2000; 29:61–63.
22. Li X, Sodroksi J. The TRIM5( b-box 2 domain promotes cooperative binding to the retroviral capsid by mediating higher-order self-association. J Virol 2008; 82:11495–11502.
23. Richardson MW, Carroll RG, Stremlau M, Korokhov N, Humeau LM, Silvestri G, et al
. Mode of transmission affects the sensitivity of human immunodeficiency virus type 1 to restriction by Rhesus TRIM5α. J Virol 2008; 82:11117–11128.
24. Sourisseau M, Sol-Foulon N, Porrot F, Blanchet F, Schwartz O. Inefficient human immunodeficiency virus replication in mobile lymphocytes. J Virol 2007; 81:1000–1012.
25. Rajsbaum R, Stove JP, O'Garra A. Type 1 interferon-dependent and independent expression of tripartite motif proteins in immune cells. Eur J Immunol 2008; 38:619–630.
26. Carthegena L, Parise MC, Ringeard M, Chelbi-Alix MK, Hazan U, Nisole S. Implication of TRIM5alpha and TRIMCyp in interferon-induced anti-retroviral restriction activities. Retrovirology 2008; 5:59.