Trim5a Structure

Protruding loops forming binding surface

Hydrophobic core of B30.2/SPRY

Protruding loops forming binding surface

Hydrophobic core of B30.2/SPRY

Figure 14.9 Structure and polymorphism of TRIM5a. TRIM5a is a member of the tripartite motif (TRIM) family of proteins and comprises RING finger, B-box, and coiled coil domains; an additional SPRY domain is specific to this a isoform of the TRIM5 proteins (a domain found also in the immunoglobulin family). TRIM5a is the longest of four splice variants of the TRIM5 gene found on chromosome 11. 'Lollipops' shown above the domains are at the locations of reported common human nonsynonymous polymorphisms (Newman and Johnson 2007). The SPRY patch is shown with black triangles highlighting the five codons in the whole protein that have evidence of evolving under positive selection based on analysis of hominids and Old World monkeys. Redrawn with permission from Newman and Johnson (2007) and from Sawyer et al. (2005) .

skew for the SPRY domain consistent with a recent selective sweep in the population removing variation in this region of the protein (Sawyer et al. 2006). Only a limited analysis of within species variation has been carried out among other primate species. Among Asian macaques and African monkeys, sequence diversity was clustered in the coiled coil domain and variable portions of B20.2/ SPRY - the latter in marked contrast to human subjects.

HIV-1 disease association studies have found no consistent effect of the TRIM5 genotype on disease progression and variable, but modest, effects on disease susceptibility (Javanbakht et al. 2006; Speelmon et al. 2006). No

Box 14.5 Analysis of primate sequence diversity reveals ancient positive selection in TRIM5a and defines a key functional element of the protein

Sawyer and colleagues sequenced the coding region of the TRIM5a gene for 17 primate genomes and analysed rates of synonymous and nonsynonymous changes between species (Sawyer et al. 2005). Such analysis allows estimation of selective pressures that have acted on a gene: in the majority of protein coding genes, rates of synonymous changes are greater than nonsynonymous as amino acid changes are usually deleterious to protein function and subject to purifying selection, resulting in their loss from the population (Section 10.3). For TRIM5a the reverse was true with ratios of nonsynonymous to synonymous changes greater than one for many branches of the primate phylogeny analysed: the range of hominids and Old and New World monkeys sequenced revealed that positive selection had been acting on this protein for more than 33 million years (Sawyer et al. 2005). Different branches of the phylogeny showed different ratios with marked evidence of positive selection for the hominid clade when sequences from the whole gene were analysed.

Strikingly, the region of TRIM5a that was found to have undergone the most intense positive selection was the SPRY domain. When hominids and Old World monkeys were analysed, only five residues in the whole protein were identified with a high level of confidence as having undergone positive selection, and all were found to lie in a very short region of the SPRY domain, dubbed the SPRY patch. Such a tight clustering of positive selection was felt by Sawyer et al. to be likely to predict a 'point of physical contact between two proteins locked in genetic conflict' and to be reminiscent of the situation in the major histocompatibility complex (MHC) where colocalization of positive selection to regions of the protein involved in antigen recognition has been found (Section 12.4) (Sawyer et al. 2005). Functional analysis demonstrated this was indeed the case: substituting the patch region between human and rhesus TRIM5a using chimeric proteins showed it contained the key amino acids conferring species-specific restriction for HIV-1.

individual effect of TRIM5 SNPs was seen for in vitro HIV-1 susceptibility of CD4+ T cells (Speelmon et al. 2006). In contrast, a SNP (rs3740996, c.127C>T, p.H43Y) in the loop region of the RING domain of TRIM5a (see Fig. 14.9) was found to be associated with impaired replication of murine N-MLV retrovirus (Sawyer et al. 2006). Based on frequencies in different populations, the estimated age of the H43Y haplotype was 500 000 years, similar to the estimate of 600 000 years for an average neutral human polymorphism (Kreitman and Di Rienzo 2004). While rare in Africans, the allele was found to be very common in Central and South Americans suggesting that there has not been a detrimental selective pressure to reduce the allele frequency in these populations from the time of the small founder populations on human migration to the Americas some 15 000-30 000 years ago (Sawyer et al. 2006). While not having any significant effect on HIV-1

restriction in human cells, the high frequency of this variant in particular populations may predispose them to greater or reduced risk in future retroviral infections.

The analysis of TRIM5a in humans and primates has provided dramatic insights into our evolutionary history with exposure to ancient and more recent retroviruses likely to determine our current and future susceptibilities.

0 0

Post a comment