Copy Number Variation and the Co-Evolution of Primate and Viral Genomes

Summary

article has not abstract

“Let it … be borne in mind how infinitely complex and close-fitting are the mutual relations of all organic beings to each other and to their physical conditions of life; and consequently what infinitely varied diversities of structure might be of use to each being under changing conditions of life.” — Charles Darwin, On the Origin of Species

November 24, 2008, marked the 149th anniversary of the first publication of Charles Darwin's seminal work entitled “On the Origin of Species.” The above quote comes from Darwin's answer to his question about how the struggle for existence might shape patterns of variation. Of course, in the 19th century, Darwin was making inferences simply based on observations of morphological variation. Yet, if he were alive today, he would be struck by how prescient his statement was, even as applied to questions about the long co-evolution of primates and viral pathogens, including lentiviruses [1].

The human genome was originally thought to be structurally stable, but it turns out to be quite dynamic, with many genomic regions duplicated or deleted among individuals to the extent that they exist in variable copy numbers. Within these copy number variations (CNVs), genes that encode proteins involved in immune responses are over-represented [2], including chemokines that play key roles in host defense against infectious diseases [3]–[5]. This observation implies that our genomes have DNA sequences that may memorialize immune strategies used to combat ancient pathogens. The past 5 years have witnessed an intense interest in understanding the extent of CNV in primate genomes [6],[7] and their contributions to disease susceptibility in humans [8]. In this issue of PLoS Genetics, Degenhardt and colleagues provide a link between CNV and disease susceptibility in non-human primates [9].

Asian macaques—including rhesus, pigtail, and cynomolgus—are commonly used as animal models to study the determinants of AIDS pathogenesis and evaluate HIV-1 vaccine candidates. After being challenged with Simian Immunodeficiency Virus (SIV)—the simian counterpart of HIV—some macaques rapidly develop features similar to AIDS, whereas others do so more slowly. A similar clinical conundrum exists in humans, as many people who are HIV-1–positive progress rapidly to AIDS, whereas others resist disease progression, despite not receiving antiretroviral therapy. Both viral and host factors contribute to the variability in AIDS progression rates in humans [10].

Among host factors that may contribute to variable HIV-AIDS susceptibility, significant attention has focused on the role of variations in genes that influence HIV transmission, such as the genes that encode CC chemokine receptor 5 (CCR5), the major HIV co-receptor required for cell entry of virus, and CCR5 chemokine ligands such as CC ligand 3 (CCL3) and its paralog CCL3L1 [10]. For example, homozygosity for a 32-bp deletion in the coding sequence of CCR5 abolishes CCR5 expression and confers near-absolute protection against acquiring HIV [10]. CCR5 ligands can block entry of HIV into cells by “gumming” up the site on CCR5 to which HIV-1 binds and by reducing cell surface expression of CCR5 [5]. Among the chemokines that bind to CCR5, CCL3L1 has the most potent HIV-suppressive properties [5]. Additionally, CCL3L genes were shown to be subject to CNV in humans and chimpanzee [11],[12]. A low copy number of the CCL3L1-containing segmental duplication was found to be associated with reduced CCL3/CCL3L1 chemokine levels, reduced chemotaxis of CCR5-expressing cells, and reduced proportions of HIV target cells that express CCR5 [5],[11],[12]. This discovery prompted investigators to inquire whether intersubject differences in CCL3L1 copy number might be a basis for variable HIV-AIDS susceptibility. A low copy number of the CCL3L1-containing segmental duplication was shown to be associated with or correlate with an increased risk of acquiring HIV infection [12]–[17], a faster rate of progression to AIDS or CD4+ T cell depletion [12],[16],[18],[19], higher HIV viral loads [12],[13],[20], lower HIV-specific immune responses [20], and lower cell-mediated immune responses [18].

In this issue of PLoS Genetics, Degenhardt and colleagues tested whether a low CCL3L copy number was associated with a faster rate of progression to AIDS in macaques challenged experimentally with SIV [9]. They found that macaques with a low copy number of CCL3L genes experience a significantly more rapid rate of progression to experimental AIDS, with the CCL3L CNV accounting for ∼18% of the variability in experimental AIDS progression rates.

Indian rhesus macaques progress more quickly to experimental AIDS than do Chinese macaques [21]. Degenhardt et al. suggest that the lower CCL3L copy number in Indian rhesus macaques may underlie the more rapid progression to AIDS in Indian versus Chinese macaques. Thus, in addition to serving as a determinant of interindividual differences in the outcome of experimental AIDS, CCL3L gene dose may account for some of the observed interpopulation differences in simian AIDS progression rates.

Previous studies have shown that there is a clear genetic distinction between rhesus macaques that originate from India versus China [22]. Thus, population structure is a possible confounding variable whenever phenotypic differences between these populations are investigated. Degenhardt et al. controlled for population structure using a battery of microsatellites and demonstrated that the CCL3L CNV was a better predictor of outcome than population affiliation. While other genes may also influence progression to simian AIDS in experimentally infected rhesus macaques [23], Degenhardt et al.'s results show that the CCL3L CNV has strong effects on progression to simian AIDS. These results have practical implications for efforts to develop an effective HIV vaccine. To distinguish more clearly between vaccine efficacy and intrinsic variation in host response, it may be important to stratify rhesus macaques by CCL3L CNV.

Understanding the role of chemokine CNVs in primate disease is made more complicated by several observations. At least in humans, there are multiple CCL3L (CCL3L1, CCL3L2, and CCL3L3) and CCL4L (CCL4L1 and CCL4L2, paralogs of CCL4) genes, which are found on chromosome 17q12; a similar diversity might exist in nonhuman primates (Figure 1A). However, the human CCL3L-CCL4L–containing locus has been subjected to complex homologous recombination events [24],[25], such that individuals may vary not only in the total copy number of CCL3L and CCL4L genes but also their individual components [11],[16]. Furthermore, the mRNA structure of the different CCL3L and CCL4L genes appears to vary (Figure 1B and 1C). For example, while human CCL4L1 and CCL4L2 share 100% sequence identity in the coding regions, a fixed mutation at the intron–exon boundary of CCL4L1 results in the production of aberrantly spliced transcripts (Figure 1B and 1C), and a higher CCL4L1 copy number has been associated with an increased risk of acquiring HIV infection [26] and faster rate of progression to AIDS [16]. With these features in mind, future studies will need to consider such questions as: Are the different copies of CCL3L and CCL4L in rhesus macaques identical or do they encode transcripts/proteins with different functions? How many of these copies are actually pseudogenes? Similar to what is observed in humans [16],[26], could CCL4L genes also contribute to simian AIDS independently of or in combination with distinct CCL3L genes? Could such complexity also confound genotype–phenotype studies in humans that investigate the relationship between CCL3L or CCL4L CNV with disease susceptibility? In addition to these CCL3L-CCL4L–related genetic factors that may complicate the analyses of association studies, there might be other confounders to consider. For example, co-infection with other viruses (e.g., hepatitis C virus [HCV]) may modify the association between CCL3L1 copy number and risk of acquiring HIV infection [17].

Using real-time PCR-based approaches, Degenhardt et al. confirmed an earlier report that chimpanzees have variable copy numbers of CCL3L genes [9],[12], as do other nonhuman primates including orangutan, African green monkey, and Sooty Mangabey; on average the CCL3L copy numbers in nonhuman primates are much higher than those found in human populations [9],[12]. Furthermore, analyses of the chimpanzee genome (from the Clint reference sequence) revealed at least four distinct CCL3L genes (Figure 1D). These results differ with those of Perry et al., who, using an array-based method, found that chimpanzees have two CCL3L copies per diploid genome [7]. These contrasting results underscore the challenges of accurately quantifying CNVs, a particularly important issue given the intense interest in understanding the role of CNVs in disease susceptibility [8].

One possible reason for the extensive variability in CCL3L copy number in primates may reflect that the variability represents an ancient host defense mechanism. While this hypothesis needs to be tested with additional empirical data, it is consistent with the observation that there is a parallel to primate chemokine CNV in viruses: many viral pathogens have hijacked DNA sequences found in primates and adapted them to encode chemokine receptors and chemokines that specifically target and, in some cases, neutralize the primate chemokine system [27]. These viral-encoded antichemokine strategies highlight the importance of the chemokine system in host defense against infections.

Darwin, an astute observer of nature, might ask, “Why does there appear to be so much structural variation for genes encoding chemokines?” The ancient and dynamic battle between mammalian hosts and pathogens has exerted unrelenting selection pressure on the host genome, promoting the development of a complex and adaptable immune system. Conversely, the successful replication and persistence of latent viruses within the mammalian host implies that they have evolved the means to evade or manipulate host immune defenses. In the case of viruses, it is clear that they have targeted the immune responses mediated by chemokines [27]. Is the expansion and diversification of the chemokine gene family, as a consequence of gene duplication [3],[4], evidence of the co-evolution of host defenses and viral pathogens? The elegant study by Degenhardt et al. gets us closer to answering this question, but much work remains. Nevertheless, Darwin would be pleased that the paradigm he established more than a century ago continues to be robust for explaining the “varied diversities of structure.”