Skip to main content
NCBI home page
American Journal of Human Genetics logoLink to American Journal of Human Genetics
. 2012 Jul 13;91(1):27–37. doi: 10.1016/j.ajhg.2012.05.008

The Evolutionary Landscape of Cytosolic Microbial Sensors in Humans

Estelle Vasseur 1,2,3,4,5, Michele Boniotto 1,2,5, Etienne Patin 1,2, Guillaume Laval 1,2, Hélène Quach 1,2, Jeremy Manry 1,2, Brigitte Crouau-Roy 3,4, Lluis Quintana-Murci 1,2,∗
PMCID: PMC3397270  PMID: 22748209

Abstract

Host-pathogen interactions are generally initiated by host recognition of microbial components or danger signals triggered by microbial invasion. This recognition involves germline-encoded microbial sensors or pattern-recognition receptors (PRRs). By studying the way in which natural selection has driven the evolution of these microbial sensors in humans, we can identify genes playing an essential role and distinguish them from other, more redundant genes. We characterized the sequence diversity of the NOD-like receptor family, including the NALP and NOD/IPAF subfamilies, in various populations worldwide and compared this diversity with that of other PRR families, such as Toll-like receptors (TLRs) and RIG-I-like receptors (RLRs). We found that most NALPs had evolved under strong selective constraints, suggesting that their functions are essential and possibly much broader than previously thought. Conversely, most NOD/IPAF subfamily members were subject to more relaxed selective constraints, suggesting greater redundancy. Furthermore, some NALP genes, including NLRP1, NLRP14, and CIITA, were found to have evolved adaptively. We identified those variants conferring a selective advantage on some human populations as the most likely targets of positive selection. More generally, the strength of selection differed considerably between the major families of microbial sensors. Endosomal TLRs and most NALPs were found to evolve under stronger purifying selection than most NOD/IPAF subfamily members and cell-surface TLRs and RLRs, suggesting some degree of redundancy in the signaling pathways triggered by these molecules. This study provides novel perspectives and experimentally testable hypotheses concerning the relative biological relevance of the various families of microbial sensors in humans.

Introduction

The innate immune system is responsible for the immediate response of the host to infectious or noxious assaults.1 It is based on germline-encoded microbial sensors, known as pattern-recognition receptors (PRRs),2 which recognize conserved pathogen-associated molecular patterns (PAMPs) from microorganisms or damage-associated molecular patterns (DAMPs) resulting from tissue damage or cellular stress.3 The effective sensing of PAMPs and DAMPs induces the activation of signaling pathways that culminate in the induction of inflammatory responses, which facilitate the eradication of pathogens and infected, injured, or dying cells.2,3 Several PRR families, including the Toll-like receptors (TLRs), the C-type lectin receptors (CLRs), the RIG-I-like receptors (RLRs), and the NOD-like receptors (NLRs), have been identified. The members of these families can be distinguished on the basis of ligand specificity, shared protein domains, cellular distribution, and downstream signaling pathways.2,4–7 The use of different families of PRRs provides the host with a certain degree of functional redundancy and multiple mechanisms for responding to a diverse range of pathogens.2

PRRs can be membrane bound, cytoplasmic, or secreted. The membrane-bound TLRs—the most studied group of PRRs—and CLRs survey the extracellular milieu and endosomal compartments, whereas the more recently discovered RLRs and NLRs scan the cytosol for signs of infection or danger.2–7 The importance of the role of NLRs, in particular, has been increasingly recognized in the last few years.6,7 In humans, the NLR proteins are encoded by a family of 22 genes and are characterized by three distinct domains: the ligand-sensing leucine-rich repeats (LRRs); the NACHT domain, which mediates oligomerization, and an effector domain, which can be a pyrin domain (PYD), a CARD (caspase recruitment domain), or a BIR (baculovirus IAP repeat) domain. Different NLRs have different N-terminal domains, and this domain defines different subfamilies, including the large NALP subfamily, the NODs, and other proteins, such as CIITA, NLRC4 (also known as IPAF), and NAIP.8 Substantial advances in the functional characterization of some NLRs, such as NOD1, NOD2, and the two NALP members NLRP1 and NLRP3 in particular, have highlighted the roles of these proteins in the sensing of microbial and nonmicrobial danger signals. Upon ligand recognition, these sensors either activate NF-κB or MAP kinases to induce inflammatory responses or activate large cytoplasmic complexes called inflammasomes, which link the sensing of microbial or danger signals to both the proteolytic activation of proinflammatory cytokines and the initiation of cell death.6,7,9,10 Increasing evidence also suggests that NLRs have a diverse range of biological functions, extending well beyond pathogen detection.6 These include roles in autoimmunity, autophagy, development, reproduction, and tissue homeostasis (see Kufer and Sansonetti6 and references therein). However, the precise functions, ligands, and sensing mechanisms of most NLRs remain to be elucidated.

Although considerable progress has been made in our understanding of the biology of the human NLRs, as well as of the other families of human PRRs, their relative contributions to host survival remain largely unknown. To address this, we used a population-genetics approach to define the way in which natural selection has driven the evolution of these microbial sensors. We thus characterized the levels of genetic diversity of the NLR members in various human populations and compared our findings with those for other PRR families, including cell-surface TLRs, endosomal TLRs, and cytosolic RLRs. This approach has proved indispensable and complementary to immunological, clinical, and epidemiological genetics studies because it has made it possible to distinguish between essential and more redundant functions of host genes.11–13 In this context, we have recently provided a proof of concept of the power of evolutionary genetics in the context of infection by using the paradigm of human interferons (IFNs).14 Here, we provide the first comprehensive view of the evolutionary landscape characterizing the major families of microbial sensors in humans. We demonstrate that the strength of selection differs considerably among the various families of human PRRs and highlight major differences in the biological relevance of the mechanisms triggered by these molecules in the natural setting.

Material and Methods

DNA Samples

Genetic variation at the NLRs was assessed in a total of 370 chromosomes from the HGDP–CEPH panel.15 This subpanel includes 62 sub-Saharan Africans, 62 Europeans, and 61 East Asians. Sub-Saharan African populations were composed of 19 Bantu from Kenya, 21 Mandenka from Senegal, and 22 Yoruba from Nigeria; European populations included 20 French, 14 Italians, 6 Orcadians, and 22 Russians; and East-Asian populations were composed of 10 Japanese, 4 Cambodians, 15 Han Chinese, and 32 individuals from Chinese minorities (for further details on the ethnic groups studied, see Table S1 in the Supplemental Data available online). This study was approved by the Institut Pasteur Ethics Committee (n° RBM 2008.06).

DNA Resequencing

We resequenced all the exonic regions of the 21 NLRs (except from the last exon of NLRP8, i.e. 3% of the coding region, which could not be amplified) and at least an equivalent amount of nonexonic portions, including intronic, 5′, and 3′ regions. Note that the exonic portions we resequenced generally correspond to the longest isoform (considered for all analyses). We inspected sequence files with the GENALYS software.16 To avoid SNP discovery errors, we analyzed sequences by using two different operators, and we reamplified and resequenced ambiguous polymorphisms. To determine ancestral states at each SNP by parsimony, we used the UCSC database to retrieve the orthologous sequences of chimpanzee, gorilla, orangutan, rhesus, baboon, marmoset, tarsier, mouse-lemur, and bushbaby.

Data Analysis

Interspecies Neutrality Tests

To estimate the direction and strength of selection within the human species as a whole, we measured dS and dN—the proportion of silent and nonsynonymous fixed differences between humans and chimpanzees—together with pS and pN,—the proportion of silent and nonsynonymous polymorphic sites observed within humans—by using DnaSP package v. 5.1.17 Then, we used the McDonald-Kreitman Poisson random field (MKPRF) method18,19 to estimate ω (ω = θN/θS α ((dN + pN)/(dS + pS)), where θN and θS are estimates of the rate of nonsynonymous and silent mutations) and γ (with γ α log (dN/pN). Under neutrality, ω is not significantly different from 1. Values of ω below 1 indicate a deficit of nonsynonymous variants (both in terms of polymorphism within humans and divergence with chimpanzees), which is consistent with selection against amino-acid-altering variation (purifying selection). Values of ω greater than 1 reflect an excess of amino acid changes, which is consistent with selection favoring amino acid mutations (positive selection). Concerning γ, the parameter is negative if a gene displays an excess of amino acid polymorphism within humans with respect to amino acid divergence between species (weak negative and/or balancing selection). In contrast with purifying selection, weak negative selection does tolerate the occurrence of nonsynonymous mutations provided that they do not increase in frequency within the population (i.e., they are nonlethal but slightly deleterious mutations).20,21 Conversely, positive γ values reflect an excess of amino acid divergence with respect to amino acid polymorphism (positive selection). The MKPRF test was performed on the sequenced NLRs, and their results were compared with those from the RLRs and TLRs.22,23 To assess the contributions of divergence and polymorphism to the patterns of purifying selection observed at some genes, we compared the pN/pS and dN/dS ratios of the PRRs with significant ω < 1 to a genome-wide distribution of 1,596 genes exhibiting ω < 1.18 Using simulations conditioned on the sample sizes and gene lengths of PRRs, we showed that the MKPRF test has the power to detect purifying selection at a gene even when this evolved neutrally at the divergence level; i.e., ω can be significantly lower than 1 when dN/dS = 1 and pN/pS < 0.6 (Figure S1).

Functional Diversity within Protein Domains

We assessed the πN/πS ratios by using the DnaSP package v. 5.117 and compared them within CARD, PYD, NBD, and LRR domains in all NLRs. Because NALPs (except NLRP10) and NOD/IPAF share an LRR domain in the C-terminal and a central NACHT domain, we compared the πN/πS ratio within these domains in all NLRs. Because NLRs differ in their N termini, we compared the πN/πS ratio within the PYD domain between NALPs only and within the CARD domain between NLRP1, NOD1, NOD2, and NLRC4. The domain division was retrieved from the Uniprot database.24 Note that the LRR domains include the whole region carrying the known LRR motifs (including inter-LRR short regions) because of the constantly evolving data concerning the bounds of these motifs.

Diversity Indices and Intraspecies Tests

We performed haplotype reconstruction by using the Bayesian method implemented in Phase (v.2.1.1).25 The algorithm was run five times, and consistent results were obtained across runs. The most likely run was retained for subsequent analyses. We used Haploview software26 to obtain and visualize levels of linkage disequilibrium (LD). For each population, the different summary statistics and the sequence-based neutrality tests were performed with DnaSP package v. 5.1.17 To detect more recent signatures of positive selection, we used various haplotype-based tests and levels of population differentiation. Specifically, we used the derived intra-allelic nucleotide diversity (DIND) test based on the ratio iπA/iπD, where iπA and iπD are the levels of nucleotide diversity associated with the haplotypes carrying the ancestral and derived allele for a given SNP, respectively.23 The rationale of this test is that a derived allele that is under positive selection and that is at high population frequencies should present lower levels of nucleotide diversity at linked sites than would be expected. We also used tests based on the levels of haplotype homozygosity; such tests included the cross-population extended haplotype homozygosity (XP-EHH) test27 obtained from the HGDP-selection browser and, when available, the integrated haplotype scores (iHS)28 obtained from the HapMap phase II dataset.29 To determine the levels of population differentiation, we assessed the FST statistics derived from the analysis of variance (ANOVA)30 for each SNP and for each pair of populations.

Correction for the Mimicking Effects of Demography

To consider the impact of demography on the patterns of diversity, we used simulation-based or empirical procedures. For the allele-frequency-spectrum and DIND tests, we incorporated into our neutral expectations two demographic models based on multiple, noncoding genomic regions sequenced in a set of populations similar to those used here.31,32 p values for the various neutrality tests were estimated from 104 coalescent simulations, performed with SIMCOAL 2.0,33 under a finite-site neutral model and considering the recombination rate of the concerned region as reported in UCSC.34 Each of the 104 coalescent simulations was conditional on the sample size and the number of segregating sites observed in each gene. For the population differentiation tests, we compared the observed FST values at each SNP at NLRs against a background FST distribution of 640,000 SNPs genotyped in the same panel of individuals we sequenced in this study.35 Because the genome-wide FST distribution of the HGDP-CEPH dataset includes loci targeted by positive selection,36 the comparison of NLR FST against the HGDP-CEPH distribution represents a conservative approach to detect selection. Because FST values depend on allele frequencies, FST comparisons were conditioned to SNPs presenting similar expected heterozygosity.

Results

Genetic Diversity of the NLR Family Members in Human Populations

We resequenced the 21 NLR genes in a panel of 185 healthy individuals of African, European, and East Asian descent (Table S1) by using Sanger resequencing. This sequencing method was the most appropriate choice because, in contrast with whole-genome sequence datasets where most NLR genes have been sequenced at low coverage (e.g., 1000 Genomes), it allows the reliable detection of low-frequency variants, which are essential for accurately detecting and estimating the intensity of selection. The NLRs can be broadly divided into two large subfamilies.8 The first, the NALP subfamily, is comprised of the 14 PYD-containing NALPs, which are encoded by NLRP1–14. The second, which is here collectively referred to as to the NOD/IPAF subfamily, includes the five NODs—NOD1 (CARD4), NOD2 (CARD15), NLRC3 (NOD3), NLRC5 (NOD4), and NLRX1 (NOD9)—together with CIITA (NLRA) and NLRC4 (also known as IPAF), which have the N-terminal CARD domain common to most NODs. The NLR NAIP was not resequenced here because of its highly repeated genomic organization.37 For each individual, we generated 171.2 kb of resequenced data, 67.9 kb of which corresponded to coding regions and the rest of which corresponded to the 5′and 3′ UTRs and introns.

We identified 2,084 SNPs, including 396 nonsynonymous SNPs, four nonsense variants, and 12 coding-region indels (Figure 1 and Table S2). We first investigated the genetic diversity of the NLRs by estimating levels of nucleotide diversity (π) over the entire sequenced region for each gene. We then compared the π values obtained for each gene with those expected under neutrality, which correspond to the mean diversity levels observed for 20 autosomal noncoding regions resequenced in a panel of populations of African, European, and Asian ancestry.31 We observed remarkable differences in the levels of nucleotide diversity between genes and between populations (Figure S2). NALPs generally displayed higher or similar levels of π than noncoding regions (0.1-0.4 SE in Africa and Asia), whereas most NOD/IPAF members had levels of nucleotide diversity lower than neutral expectations (0.4–0.6 SE in all populations). We next determined the significance of these empirical observations and formalized the extent to which selection has acted upon these gene families. To do this, we used tests that distinguish the putative functional impact of different types of mutations (e.g., nonsynonymous versus silent). More generally, we compared the effects of natural selection on the NLRs and other major families of microbial sensors between human and chimpanzee lineages and within different human populations. This should be helpful in attempts to predict their relative biological relevance.

Figure 1.

Figure 1

Open in a new tab

Distribution of the Nonsynonymous and Nonsense Variants Identified across the NLRs in This Study

The location of each nonsynonymous and nonsense variant within the different protein domains is shown. AD stands for activation domain. Concerning NLRX1, the N-terminal domain is neither a PYRIN nor a CARD.

The NALP and NOD/IPAF Subfamilies Have Evolved under Different Functional Constraints

To investigate whether and how natural selection has driven the observed heterogeneous patterns of diversity, we first estimated the direction and strength of selection within the human species as a whole. We used the McDonald-Kreitman Poisson random-field (MKPRF) test,18,19 which compares polymorphism within humans with the divergence between humans and chimpanzees at nonsynonymous and silent sites. Specifically, this method provides two measures of the selective pressure: ω, which compares nonsynonymous and silent mutation rates calculated from both divergent and polymorphic sites, and γ, which relies on the ratio of divergence and polymorphism at nonsynonymous sites. Under neutrality, ω is not significantly different from 1. Values < 1 indicate a deficit of nonsynonymous variants (consistent with the action of strong negative selection, i.e., purifying selection), whereas values >1 reflect selection favoring nonsynonymous variants (positive selection). With respect to γ, negative values indicate an excess of nonsynonymous polymorphism within humans with respect to divergence (weak negative selection and/or balancing selection), whereas positive values reflect an excess of nonysnonymous divergence with respect to that observed for silent sites (positive selection).

Ten of the 14 NALPs (NLRP3, 4, 5, 7, 8, 9, 10, 11, 12, and 13) and NLRC5 had a ω value significantly lower than 1, indicating a deficit of nonsynonymous variants (Figure 2) and reflecting the action of purifying selection. Conversely, most NOD/IPAF subfamily members (NOD1, NLRC3, NLRC4, NLRX1, and CIITA), and NLRP2 had significantly negative γ values, attesting to an excess of nonsynonymous SNPs segregating in the human population with respect to divergence (Figure S3). Inspection of the allele frequency spectra showed that the excess of nonsynonymous polymorphism observed at these genes displaying significantly negative γ values could be accounted for mostly by low-frequency variants (Figure S4). This suggests that the observed excess of nonsynonymous variants is most likely explained by a regime of weak negative selection, which has tolerated the accumulation of amino acid changes but has prevented them from increasing to high frequency in the population. Calculation of the ratio of nonsynonymous to silent nucleotide diversities (πN/πS) per domain for each gene showed that the differences in selective constraints between the NALP and NOD/IPAF subfamilies could not be attributed to a particular protein domain. These results indicate that NLRs have evolved into two distinct evolutionary groups; most NALPs have been targeted by strong purifying selection, whereas a weaker regime of negative selection has driven the evolution of most NOD/IPAF members.

Figure 2.

Figure 2

Open in a new tab

Estimation of Purifying Selection Acting on Individual NLR Genes and Genes from Other Major Families of Microbial Sensors

We assessed the strength of purifying selection by calculating ω via the MKPRF test. Scale bars indicate 95% confidence intervals, and red diamonds indicate genes with ω estimates significantly lower than 1. The results for the population selection parameter γ are presented in Figure S3. Note that nonsense mutations and coding indels were either absent or present at very low frequency (<1%) in NLRs (Table S2) and, more generally, in all considered families of PRRs other than the cell-surface molecules TLR10 and TLR5, for which 5% and up to 23%, respectively, of individuals from the general population carried a nonsense mutation.22,23

Some NLRs Have Been Subject to Positive Selection in Specific Populations

With a view to identifying functional variation that might have conferred a selective advantage on the host, we then investigated whether some NLRs had evolved adaptively. We thus performed various intra-species neutrality tests on various aspects of the data, including the allele frequency spectrum (i.e., Tajima's D, Fu and Li's D∗ and F∗, and Fay and Wu's H tests), levels of population differentiation (i.e., FST), and haplotype-based tests (i.e., DIND and iHS tests) (for extensive reviews on these statistical tests, see Nielsen20 and Nielsen et al.21). At a first glance, tests of the allele frequency spectrum showed that most NOD/IPAF members were characterized by an excess of singletons, as the significantly negative values obtained for the sequence-based neutrality tests attest (Table 1). These patterns could be explained by local positive selection; when a mutation (or a haplotype) is selected and increases in frequency, the nonselected mutations become rare and increase the number of singletons overall. However, because no significant signals of positive selection were detected on the basis of independent tests, the most parsimonious explanation is that weak negative selection has maintained polymorphisms at low frequency, leading to the excess of observed singletons. In light of this, NOD/IPAF members seem to have mainly evolved under the action of weak negative selection, as also attested by the γ estimates (see MKPRF results in Figure S3). More generally, with respect to positive selection, we defined genes under selection conservatively as those for which significant deviations from neutrality were obtained at least in two independent tests for selection based on different aspects of the data (e.g., allele-frequency-spectrum tests and FST) in a given population. Statistical significance of the tests was assessed with either coalescent simulations adjusted on human demography or empirical genome-wide distributions (see Material and Methods for details). With these conservative criteria, most NLRs showed no significant deviation from neutral expectations, although NLRP1, NLRP14, and CIITA represented notable exceptions (Table 1, Figure 3, and Figure S5).

Table 1.

Neutrality Tests across NLR Genomic Regions

Gene Africa
 Europe
 Asia
TD D∗ F∗ Hn DH TD D∗ F∗ Hn DH TD D∗ F∗ Hn DH
NLRP1 −1.24 0.08 −0.6 −2.193∗∗ ∗ −0.608 −0.907 −0.933 −2.98∗/+ NS −0.324 −1.361 −1.119 −0.74 NS
NLRP2 −0.52 −0.59 −0.67 −0.75 NS 0.001 −0.836 −0.577 −0.36 NS −0.013 −1.334 −0.941 −0.57 NS
NLRP3 −0.53 −1.81 −1.55 0.885 NS −0.462 −1.191 −1.08 0.056 NS 0.123 −0.478 −0.296 0.259 NS
NLRP4 −0.6 −2.36 −1.94 0.538 NS 0.261 −1.282 −0.78 0.233 NS 0.884 −0.688 −0.061 −0.03 NS
NLRP5 −0.71 −0.92 −0.99 −0.3 NS 0.603 0.602 0.723 −0.29 NS 0.151 −0.359 −0.172 −0.74 NS
NLRP6 −0.75 −1.66 −1.53 −0.12 NS −0.525 −2.769∗/+ −2.236∗/+ 0.07 NS 0.258 −1.689 −1.124 −1.33 NS
NLRP7 0.216 −0.55 −0.27 −0.987∗ NS 1.388 −0.457 0.347 −1.56 NS 0.476 −1.516 −0.848 −1.06 NS
NLRP8 −0.5 −0.43 −0.55 0.078 NS 0.757 −1.168 −0.469 0.224 NS 0.662 −0.917 −0.339 0.196 NS
NLRP9 −1.45 −1.45 −1.74 0.61 NS −0.942 −2.131+ −1.997+ 0.513 NS −0.957 −1.211 −1.333 0.621 NS
NLRP10 −1.772∗ −1.29 −1.78 0.532 NS −0.42 −0.823 −0.81 0.661 NS −0.03 −0.617 −0.494 0.872 NS
NLRP11 −0.5 −0.93 −0.9 0.498 NS −0.716 −0.783 −0.907 −0.28 NS 0.205 −1.137 −0.709 0.275 NS
NLRP12 −0.87 −1.73 −1.63 −1.739∗∗ NS 0.489 −0.432 −0.065 −2.072∗ NS 0.671 −1.604 −0.82 −1.994∗ NS
NLRP13 −0.69 −1.28 −1.23 −0.15 NS 0.914 −0.148 0.344 −0.59 NS 1.261 0.88 1.242 −0.87 NS
NLRP14 −0.87 −1.12 −1.22 0.47 NS 0.047 0.018 0.035 −0.19 NS −1.134 −2.256+ −2.18∗/+ −0.59 NS
NOD1 0.081 −0.98 −0.64 −0.2 NS 0.127 −0.715 −0.444 −1.39 NS 0.066 −3.024∗/+ −2.107∗/+ −1.44 NS
NOD2 −1.41 −3.6 −3.22 0.209 NS −0.344 −3.278∗/+ −2.509∗/+ 0.032 NS −1.614∗/+ −4.516∗∗/++ −4.063∗∗/++ −1.28 ∗
NLRC3 −1.47 −1.1 −1.52 −1.097∗ ∗∗ −1.692∗/+ −2.151 −2.35∗/+ −2.344∗ ∗∗ −1.704∗/+ −1.737 −2.069+ −1.58 ∗∗
NLRC4 −1.64 −2.13 −2.32 0.383 NS −0.73 −3.746∗∗/+ −3.166∗∗/+ 0.199 NS −1.112 −3.239∗/+ −2.924∗/+ 0.032 NS
NLRC5 −0.76 −1.98 −1.72 0.05 NS 0.155 0.357 0.322 −0.52 NS −0.189 −2.374∗/+ −1.697∗/+ −0.81 NS
NLRX1 −1.63 −1.69 −2 −1.225∗ ∗∗ −0.925 −4.097∗∗/+ −3.447∗∗/+ 0.273 NS −1.295 −3.1∗/+ −2.889∗/+ 0.021 NS
CIITA −0.84 −1.76 −1.63 0.072 NS 0.194 −0.412 −0.187 −0.22 NS 0.385 −2.136+ −1.284 0.129 NS
Open in a new tab

Abbreviations are as follows: TD, Tajima's D; D∗, Fu & Li's D∗; F∗, Fu & Li's F∗; Hn, normalized Fay & Wu's H; DH, Tajima's D and Fay & Wu's H p values combined. A double asterisk indicates a p value ≤ 0.01, and a single asterisk indicates a p value ≤ 0.05, both according to the model of Voight et al.32 A double plus sign indicates a p value ≤ 0.01, and a single plus sign indicates a p value ≤ 0.05, both according to the model of Laval et al.31 Most of the members of the NOD/IPAF subfamily gave strongly negative values of Fu & Li's D∗ and F∗, indicative of an excess of singletons; this excess was significant in Europe and Asia in most cases. However, most gave no other significant signature of positive selection, suggesting that these patterns are the consequence of weak negative selection, as the results obtained for the population selection parameter γ in the MKPRF test attest (Figure S3).

Figure 3.

Figure 3

Open in a new tab

Levels of Population Differentiation at the NLRs

The FST statistic is presented as a function of heterozygosity for each SNP in (A) Africans versus Europeans, (B) Africans versus East-Asians, and (C) Europeans versus East-Asians. The 95th and 99th percentiles of the Human Genome Diversity Panel-Centre d'Etude du Polymorphisme Humain (HGDP-CEPH) genotyping dataset for the same individuals as those studied here are shown as dashed lines, whereas the blue area corresponds to the 99.9th percentile. Black and red points represent silent and nonsynonymous SNPs, respectively. For each outlier SNP, the gene name, followed by its position respective to the ATG, is indicated. Outlier SNPs separated by a comma correspond to SNPs in complete LD, and nonsynonymous SNPs are underlined. Note that, in Asian populations, a group of SNPs in NLRP6, including the nonsynonymous SNP 1652C>A (Leu163Met), displayed some of the highest levels of differentiation of any of the SNPs studied.

The strongest signatures of positive selection were detected for NLRP1 (NALP1). Tests based on the allele frequency spectrum (particularly Fay & Wu's H and DH tests) showed an excess of high-frequency derived alleles in Africa and Europe (Table 1). We identified a group of 15 SNPs for which derived alleles with a frequency >90% were found in Africa and Europe. In Europe, these high-frequency derived alleles were in strong linkage disequilibrium (LD) with 29 high-frequency ancestral alleles in a haplotype block of ∼45 kb (Figure S6 and Table S3). The LD block was shorter in Africa, but most of the SNPs characterizing the European haplotype were present at similar frequencies. In Asia, most of the corresponding derived alleles were fixed (Table S3), accounting for the lack of significance of Fay and Wu's H values (i.e., this test is known to lose power when the targeted allele has reached fixation). These observations suggest the occurrence of an event of positive selection worldwide (i.e., a selective sweep) that is still underway in Africa and Europe but has been completed in Asia. In haplotype-based tests, 11 of the mutations characterizing the European haplotype gave iHS values >2 in Europeans from HapMap, and one mutation gave an iHS value >2 in Africans from HapMap (Table S3).29 These significant iHS values were observed at highly frequent ancestral alleles only because derived alleles characterizing the selected haplotype were not available from HapMap.29 Overall, our results are thus consistent with a worldwide event of positive selection targeting a long haplotype that includes seven nonsynonymous SNPs: SNP 23999C>G Thr246Ser, SNP 25607C>G Thr782Ser, SNP 42035C>T Thr878Met, SNP 49993T>C Ile995Thr, SNP 53312G>A Val1119Met, SNP 62372C>G Leu1241Val, and SNP 68479C>T Arg1366Cys (see Table S3).

At NLRP1, we detected another positive-selection event restricted to Europe. The DIND test identified the nonsynonymous SNP 51015G>A (Val1059Met) and four linked intronic variants (Figure 4A), which were not in LD (r2 = 0.04) with the long haplotype described above (Figure S6). Two of these SNPs (including the Val1059Met variant) had iHS values <−2 (Table S2), providing evidence of the action of positive selection on the derived alleles. Furthermore, the frequencies of the derived alleles were higher in Europe than elsewhere (33% in Europe versus 4%–8% elsewhere; Figure S7). FST analyses identified another nonsynonymous variant, SNP 1911T>A (Leu155His), as being strongly differentiated in Europe (FST = 0.40 and 0.35 for Africa/Europe and Europe/Asia, respectively) (Figure 3). Leu155His was found to be in intermediate LD (r2 = 0.52) with the Val1059Met variant and the four intronic variants (Figure S6). In light of these results, and given that DIND and iHS tests are known to be especially adapted to detect recent positive selection, our data are consistent with an independent, more recent event of positive selection targeting NLRP1 in Europe.

Figure 4.

Figure 4

Open in a new tab

Detection of Recent Positive Selection Acting on NLRP1 in Europe and on NLRP14 in Asia

We plotted iπA/iπD values against derived allele frequencies (DAFs) for (A) NLRP1 in Europe and (B) NLRP14 in Asia. We obtained p values by comparing the iπA/iπD values for NLRP1 and NLRP14 with the expected values obtained from 104 simulations by using a best-fitted demographic model of human populations; this model is the most conservative in the context of the detection of positive selection.31 The upper dashed line on the graph corresponds to the 99th percentile, and the lower line corresponds to the 95th percentile. Black and red points represent silent and nonsynonymous SNPs, respectively. Outlier SNPs separated by a comma correspond to SNPs in complete LD, and nonsynonymous SNPs are underlined. As for NLRP1, in addition to the selected SNP 51015G>A (Val1059Met), our analyses also identified another nonsynonymous SNP (SNP 62201G>A, Val1184Met) linked to an intronic variant (SNP 63236G>A). This signal might be a complex repercussion on the worldwide selective sweep. For the DIND analyses of all genes in all populations, see Figure S5.

Another positive-selection signal was identified for NLRP14 (also known as NALP14). This gene displayed an excess of low-frequency alleles in Asia, and the DIND test identified the nonsynonymous SNP 19221G>A (Glu808Lys) variant as an outlier in the three populations, particularly in Asia (Table 1, Figure 4B, and Figure S5). Haplotype-based tests confirmed this signature, and XP-EHH values greater than 2 were recorded for Asia.27 Interestingly, in most HGDP-CEPH Asian populations, the highest XP-EHH value in this genomic region was observed for the Glu808Lys variant and two intronic SNPs.35 Together, these observations suggest that Glu808Lys is the most likely target of either stronger or earlier selection in Asian populations, consistent with the clear star-like shape of the NLRP14 network, particularly in Asia (Figure S8).

Finally, we detected a signature of positive selection for the NOD/IPAF member CIITA on the basis of the FST analysis, which identified SNP −286 (rs3087456) as highly differentiated in Europe and SNPs −730 (rs4781010) and −561 (rs2071170) as highly differentiated in Asia (Figure 3). The signal in Europe was confirmed by the >2 iHS value obtained for SNP –712 (rs12596540) and other HapMap SNPs (rs8052975, rs6498114, rs12922863, rs6498116, rs12928665, and rs11074934) in high LD (r2 > 0.6) with SNP −286. These results support the occurrence in Europe of a positive-selection event targeting the CIITA promoter region.

Lastly, the cases of NLRP6, NLRP12, and NOD2 are worthy of note, despite the fact that the signatures of selection observed at these genes were supported by a single neutrality test and should thus be interpreted cautiously. For NLRP6, neutrality tests revealed an excess of low-frequency alleles in Europe (Table 1), but no other signal of selection was detected. In turn, in Asia we identified the nonsynonymous SNP 1652C>A (Leu163Met) as being strongly differentiated (FST = 0.58 and 0.51 for Africa/Asia and Europe/Asia, respectively) (Figure 3). With respect to NLRP12, we detected an excess of high-frequency derived alleles in all populations, which is consistent with a worldwide selective sweep (Table 1). For NOD2, given the strongly negative values of the tests, particularly in Asia, the highly significant excess of rare alleles in Europe and Asia suggests the action of positive selection (Table 1).

Global Patterns of Selection Differ between the Major Families of Microbial Sensors

Finally, we evaluated the patterns of selection for the NALP and NOD/IPAF subfamilies in the context of the results we obtained for other major families of microbial sensors.22,23 These families included the TLRs—both those located in the endosome (TLR3, TLR7, TLR8, and TLR9) and those expressed on the cell surface (TLR1, TLR2, TLR4, TLR5, TLR6 and TLR10) —and the cytosolic RLRs—DDX58 (RIG-I), IFIH1 (MDA5), and DHX58 (LGP2). The MKPRF test was performed on the 34 genes encoding these receptors. On the basis of these test results, we identified a group of genes subject to strong selective constraints—this group included most NALPs and the endosomal TLRs—and a group of genes subject to weaker evolutionary constraints—this group included most NOD/IPAF subfamily members, the cell-surface TLRs, and the cytosolic RLRs (Figure 2). We next evaluated the respective contributions of divergence and polymorphism to the patterns of purifying selection observed in the group of NALPs and endosomal TLRs, which present ω values significantly lower than 1. To do so, we compared their pN/pS and dN/dS values with the genome-wide distribution of pN/pS and dN/dS for genes presenting the same features (i.e., for 1,596 genes with significant ω < 1; see Figure S1).18 None of our PRR genes presented pN/pS or dN/dS values that were significantly higher than expected, suggesting that the evolutionary constraints observed for this group of PRRs result from the continuous action of purifying selection since the human-chimpanzee divergence.

Discussion

In this study, we have shown that the subfamilies of NLRs have followed very different evolutionary pathways. NALPs, the least-studied group of NLRs, have mostly evolved under strong purifying selection and are characterized by an overall deficit of functional diversity. Indeed, this gene subfamily shows a significant enrichment in genes under the action of strong purifying selection (p = 1 × 10−4, 71% observed versus 20% expected at the genome-wide level on the basis of ∼11,600 genes, see Bustamante et al.18). By contrast, most members of the NOD/IPAF subfamily have been subject to more relaxed selection constraints, reflecting a higher degree of redundancy. These observations are consistent with the idea that most NALPs acquired a function that is essential and nonredundant in host survival, as well as with the rapid elimination of mutations of these genes from the population as a result of their highly deleterious effects. Our observations are supported by medical genetic studies involving NLRP3 (NALP3; which is subject to the highest degree of evolutionary constraint) in which missense mutations have been associated with rare severe inflammatory diseases (MIM 606416).38,39 NLRP3 expression is essentially limited to immune and nonkeratinizing epithelial cells,40 and the protein encoded by this gene is known to activate caspase-1 in the sensing of bacteria or DAMPs.6 The purifying-selection regime under which this gene has evolved suggests an important role of this sensor in caspase-1-mediated immunity signaling and in processes that are independent of caspase-1 or that are unrelated to pathogen recognition.6,41

NALPs are encoded by a multigene family, most of the members of which are under strong selective constraints. This situation contrasts with theoretical predictions about multigene families, many of the members of which can become pseudogenes or are subject to relaxed selective constraints.42 This observation highlights the important role that most NALPs might have, probably in a much more diverse range of functions than the mere sensing of microbial and danger signals.6 For example, increasing evidence suggests that NALPs are involved in the maintenance of intestinal homeostasis, as shown for the murine Nalp3, the absence of which is associated with greater tissue damage and colitis.43,44 It is also becoming evident that many NALP genes are expressed specifically in gametes and embryos,45 consistent with an important role in early development and reproduction. Mutations of the human NLRP7, for example, are associated with abnormal human pregnancies, spontaneous abortions, and intrauterine growth retardation (MIM 609661).46 These reproductive and developmental functions might be more closely related to immune functions than previously anticipated.8 Indeed, activation of the proinflammatory cytokine IL-1β after inflammasome formation is essential for ovulation and oocyte maturation.47 However, the genuine functions of most NALPs remain poorly documented and little studied. In this context, our evolutionary data are particularly informative because they are consistent with key roles for most NALPs in host survival, highlighting the need for functional data on this major family of ATPases in humans in order for its physiological relevance to health and disease to be assessed.

Our analyses have also revealed that some NALP and NOD/IPAF members have evolved adaptively, attesting to the presence of functional variation that might confer an advantage to specific human populations. The patterns depicted at NLRP1 (NALP1), which displays the strongest signals of positive selection, are consistent with an event of positive selection at a haplotype comprising seven nonsynonymous changes. NLRP1 is one of the few NALPs for which a PRR function has been well documented.6,48 These seven identified amino acid changes might therefore have conferred a selective advantage related to microbial sensing and might underlie differences in susceptibility to infections and immunity-related disorders. Furthermore, at NLRP1 we identified an independent, more recent positive-selection signature restricted to Europe; in this signature, two nonsynonymous changes (Val1059Met and Leu155His) displayed positive-selection signals. Interestingly, the Leu155His variant has been shown to be associated with various autoimmune diseases, including Addison's disease, type I diabetes, and vitiligo (MIM 606636).49,50 However, most of our analyses localized the signature of selection to Val1059Met, rather than to Leu155His, suggesting that Val1059Met is the actual target of selection. The action of selection on Val1059Met might have thus increased the frequency, among Europeans, of the genetically linked Leu155His mutation, which is nowadays held responsible for several autoimmune diseases. These findings provide support for the hypothesis that the current high incidence of autoimmune or inflammatory disorders results from past adaptation to infectious agents.11,51 Finally, the positively selected SNP −286 (rs3087456) at CIITA has been associated with different susceptibilities to inflammatory diseases, such as rheumatoid arthritis, multiple sclerosis, and myocardial infarction (MIM 600005).52 The overlap between the positive selection signatures observed for some NLRP1 and CIITA variants and previous associations with disease states provides proof of concept for the use of this evolutionary approach to predict the functional impact of other, as-yet-uncharacterized positively selected variants of NLRP1 and NLRP14 and to evaluate their potential implications in human disease. The signatures of positive selection observed at other NLRs, such as NLRP6, NLRP12, and NOD2, were not confirmed by various independent tests of selection and therefore failed to pass our stringent criteria. However, we cannot rule out the possibility that positive selection, probably in a more modest way, might have targeted these genes.

Pathogens harbor multiple ligands that are sensed by multiple families of PRRs through crosstalk between the corresponding signaling pathways,2 which may display various degrees of redundancy. Our results for NLRs therefore cannot be interpreted in isolation. Our study revealed important differences in the intensity of selection driving the evolution of the major families of microbial sensors and provided information about the biological relevance of the mechanisms triggered by these molecules (Figure 5). For PRRs specialized in the sensing of nucleic acids, particularly those from viruses, we found that endosomal TLRs were under stronger evolutionary constraints than cytosolic RLRs, suggesting a nonredundant, essential role for endosomal TLRs. Indeed, TLRs and RLRs use specific adaptors to initiate their respective signaling cascades, but these pathways ultimately converge in the production of type-I IFNs and proinflammatory cytokines.4,5 The more-relaxed evolutionary constraints on RLRs than on endosomal TLRs might therefore reflect some redundancy of this system in antiviral immunity. For PRRs involved in the sensing of nonnucleic acid products, mostly from bacteria, and stress signals, we found that the members of the NALP family were generally subject to strong purifying selection, whereas NOD/IPAF subfamily members and cell-surface TLRs have evolved under weaker constraints. This supports a higher degree of redundancy of these latter two groups of microbial sensors.

Figure 5.

Figure 5

Open in a new tab

Hierarchical Model Outlining the Evolutionary Dynamics and Biological Relevance of the Various Families of PRRs

This representation is based on the intensity of the selective constraints (based on the MKPRF results) detected for the 34 PRRs. These analyses allowed us to distinguish three groups of genes: genes under purifying selection (ω < 1, in red), genes under weaker selective constraints (γ < 0, in yellow), and genes for which no deviation from neutrality was detected (in gray). Color intensity is proportional to the –log(p value) of ω or γ tests. Cellular sublocalization, protein domains, and ligands are given as an indication but are not exhaustive.

In conclusion, our analyses allowed us to distinguish three groups of innate-immunity genes that differ in their evolutionary patterns: genes under strong selective constraints, genes under weaker constraints, and genes for which no deviation from neutrality was detected. Mutations in genes evolving under the effects of strong purifying selection are likely to be associated with severe clinical phenotypes and, therefore, strongly constrained genes are candidates for involvement in individual, rare Mendelian deficiencies. Conversely, mutations in genes evolving under more relaxed constraints (i.e., weak negative selection or neutrality) will generally have a more modest impact on host survival, although they might subtly modulate complex susceptibility to disease at the population level, as illustrated by the case of NOD2 variation and susceptibility to Crohn disease.53–55 These data open new research perspectives and facilitate the formulation of experimentally testable hypotheses by providing a general hierarchical model for the biological relevance of the various microbial sensors, some of which are essential and some of which are more expendable. These findings should stimulate future functional studies aiming to determine whether the strong constraints on the genes for some of these sensors, including the little-studied NALPs, provide evidence of the importance of the PRR functions putatively mediated by these sensors or, more generally, for broader processes extending to basic, early developmental mechanisms and the maintenance of body homeostasis.

Acknowledgments

We would like to thank Luis B. Barreiro, Jean-Laurent Casanova, and Philippe Sansonetti for critical reading of the manuscript. This work was supported by Institut Pasteur, the Agence Nationale de la Recherche (ANR-08-MIEN-009-01), the Fondation pour la Recherche Médicale, the Centre National de la Recherche Scientifique, Merck-Serono, and an École Polytechnique Fédérale de Lausanne-Debiopharm Life Sciences Award to L.Q.-M.

Supplemental Data

Document S1. Figures S1–S8 and Tables S1 and S3
mmc1.pdf (1.6MB, pdf)
Document S2. Table S2
mmc2.xls (740.5KB, xls)

Web Resources

The URLs for data presented herein are as follows:

References

  • 1.Janeway C.A., Jr., Medzhitov R. Innate immune recognition. Annu. Rev. Immunol. 2002;20:197–216. doi: 10.1146/annurev.immunol.20.083001.084359. [DOI] [PubMed] [Google Scholar]
  • 2.Takeuchi O., Akira S. Pattern recognition receptors and inflammation. Cell. 2010;140:805–820. doi: 10.1016/j.cell.2010.01.022. [DOI] [PubMed] [Google Scholar]
  • 3.Matzinger P. The danger model: A renewed sense of self. Science. 2002;296:301–305. doi: 10.1126/science.1071059. [DOI] [PubMed] [Google Scholar]
  • 4.Kawai T., Akira S. TLR signaling. Semin. Immunol. 2007;19:24–32. doi: 10.1016/j.smim.2006.12.004. [DOI] [PubMed] [Google Scholar]
  • 5.Loo Y.M., Gale M., Jr. Immune signaling by RIG-I-like receptors. Immunity. 2011;34:680–692. doi: 10.1016/j.immuni.2011.05.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Kufer T.A., Sansonetti P.J. NLR functions beyond pathogen recognition. Nat. Immunol. 2011;12:121–128. doi: 10.1038/ni.1985. [DOI] [PubMed] [Google Scholar]
  • 7.Schroder K., Tschopp J. The inflammasomes. Cell. 2010;140:821–832. doi: 10.1016/j.cell.2010.01.040. [DOI] [PubMed] [Google Scholar]
  • 8.Martinon F., Mayor A., Tschopp J. The inflammasomes: Guardians of the body. Annu. Rev. Immunol. 2009;27:229–265. doi: 10.1146/annurev.immunol.021908.132715. [DOI] [PubMed] [Google Scholar]
  • 9.Tschopp J., Martinon F., Burns K. NALPs: A novel protein family involved in inflammation. Nat. Rev. Mol. Cell Biol. 2003;4:95–104. doi: 10.1038/nrm1019. [DOI] [PubMed] [Google Scholar]
  • 10.Martinon F., Burns K., Tschopp J. The inflammasome: A molecular platform triggering activation of inflammatory caspases and processing of proIL-beta. Mol. Cell. 2002;10:417–426. doi: 10.1016/s1097-2765(02)00599-3. [DOI] [PubMed] [Google Scholar]
  • 11.Barreiro L.B., Quintana-Murci L. From evolutionary genetics to human immunology: How selection shapes host defence genes. Nat. Rev. Genet. 2010;11:17–30. doi: 10.1038/nrg2698. [DOI] [PubMed] [Google Scholar]
  • 12.Quintana-Murci L., Alcaïs A., Abel L., Casanova J.L. Immunology in natura: Clinical, epidemiological and evolutionary genetics of infectious diseases. Nat. Immunol. 2007;8:1165–1171. doi: 10.1038/ni1535. [DOI] [PubMed] [Google Scholar]
  • 13.Casanova J.L., Abel L., Quintana-Murci L. Human TLRs and IL-1Rs in host defense: Natural insights from evolutionary, epidemiological, and clinical genetics. Annu. Rev. Immunol. 2011;29:447–491. doi: 10.1146/annurev-immunol-030409-101335. [DOI] [PubMed] [Google Scholar]
  • 14.Manry J., Laval G., Patin E., Fornarino S., Itan Y., Fumagalli M., Sironi M., Tichit M., Bouchier C., Casanova J.L. Evolutionary genetic dissection of human interferons. J. Exp. Med. 2011;208:2747–2759. doi: 10.1084/jem.20111680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Cann H.M., de Toma C., Cazes L., Legrand M.F., Morel V., Piouffre L., Bodmer J., Bodmer W.F., Bonne-Tamir B., Cambon-Thomsen A. A human genome diversity cell line panel. Science. 2002;296:261–262. doi: 10.1126/science.296.5566.261b. [DOI] [PubMed] [Google Scholar]
  • 16.Takahashi M., Matsuda F., Margetic N., Lathrop M. Automated identification of single nucleotide polymorphisms from sequencing data. J. Bioinform. Comput. Biol. 2003;1:253–265. doi: 10.1142/s021972000300006x. [DOI] [PubMed] [Google Scholar]
  • 17.Rozas J., Sánchez-DelBarrio J.C., Messeguer X., Rozas R. DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics. 2003;19:2496–2497. doi: 10.1093/bioinformatics/btg359. [DOI] [PubMed] [Google Scholar]
  • 18.Bustamante C.D., Fledel-Alon A., Williamson S., Nielsen R., Hubisz M.T., Glanowski S., Tanenbaum D.M., White T.J., Sninsky J.J., Hernandez R.D. Natural selection on protein-coding genes in the human genome. Nature. 2005;437:1153–1157. doi: 10.1038/nature04240. [DOI] [PubMed] [Google Scholar]
  • 19.Sawyer S.A., Hartl D.L. Population genetics of polymorphism and divergence. Genetics. 1992;132:1161–1176. doi: 10.1093/genetics/132.4.1161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Nielsen R. Molecular signatures of natural selection. Annu. Rev. Genet. 2005;39:197–218. doi: 10.1146/annurev.genet.39.073003.112420. [DOI] [PubMed] [Google Scholar]
  • 21.Nielsen R., Hellmann I., Hubisz M., Bustamante C., Clark A.G. Recent and ongoing selection in the human genome. Nat. Rev. Genet. 2007;8:857–868. doi: 10.1038/nrg2187. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Vasseur E., Patin E., Laval G., Pajon S., Fornarino S., Crouau-Roy B., Quintana-Murci L. The selective footprints of viral pressures at the human RIG-I-like receptor family. Hum. Mol. Genet. 2011;20:4462–4474. doi: 10.1093/hmg/ddr377. [DOI] [PubMed] [Google Scholar]
  • 23.Barreiro L.B., Ben-Ali M., Quach H., Laval G., Patin E., Pickrell J.K., Bouchier C., Tichit M., Neyrolles O., Gicquel B. Evolutionary dynamics of human Toll-like receptors and their different contributions to host defense. PLoS Genet. 2009;5:e1000562. doi: 10.1371/journal.pgen.1000562. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Consortium T.U., UniProt Consortium The Universal Protein Resource (UniProt) 2009. Nucleic Acids Res. 2009;37(Database issue):D169–D174. doi: 10.1093/nar/gkn664. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Stephens M., Donnelly P. A comparison of bayesian methods for haplotype reconstruction from population genotype data. Am. J. Hum. Genet. 2003;73:1162–1169. doi: 10.1086/379378. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Barrett J.C., Fry B., Maller J., Daly M.J. Haploview: Analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21:263–265. doi: 10.1093/bioinformatics/bth457. [DOI] [PubMed] [Google Scholar]
  • 27.Sabeti P.C., Varilly P., Fry B., Lohmueller J., Hostetter E., Cotsapas C., Xie X., Byrne E.H., McCarroll S.A., Gaudet R., International HapMap Consortium Genome-wide detection and characterization of positive selection in human populations. Nature. 2007;449:913–918. doi: 10.1038/nature06250. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Voight B.F., Kudaravalli S., Wen X., Pritchard J.K. A map of recent positive selection in the human genome. PLoS Biol. 2006;4:e72. doi: 10.1371/journal.pbio.0040072. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Frazer K.A., Ballinger D.G., Cox D.R., Hinds D.A., Stuve L.L., Gibbs R.A., Belmont J.W., Boudreau A., Hardenbol P., Leal S.M., International HapMap Consortium A second generation human haplotype map of over 3.1 million SNPs. Nature. 2007;449:851–861. doi: 10.1038/nature06258. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Excoffier L., Smouse P.E., Quattro J.M. Analysis of molecular variance inferred from metric distances among DNA haplotypes: Application to human mitochondrial DNA restriction data. Genetics. 1992;131:479–491. doi: 10.1093/genetics/131.2.479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Laval G., Patin E., Barreiro L.B., Quintana-Murci L. Formulating a historical and demographic model of recent human evolution based on resequencing data from noncoding regions. PLoS ONE. 2010;5:e10284. doi: 10.1371/journal.pone.0010284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Voight B.F., Adams A.M., Frisse L.A., Qian Y., Hudson R.R., Di Rienzo A. Interrogating multiple aspects of variation in a full resequencing data set to infer human population size changes. Proc. Natl. Acad. Sci. USA. 2005;102:18508–18513. doi: 10.1073/pnas.0507325102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Laval G., Excoffier L. SIMCOAL 2.0: A program to simulate genomic diversity over large recombining regions in a subdivided population with a complex history. Bioinformatics. 2004;20:2485–2487. doi: 10.1093/bioinformatics/bth264. [DOI] [PubMed] [Google Scholar]
  • 34.Fujita P.A., Rhead B., Zweig A.S., Hinrichs A.S., Karolchik D., Cline M.S., Goldman M., Barber G.P., Clawson H., Coelho A. The UCSC Genome Browser database: Update 2011. Nucleic Acids Res. 2011;39(Database issue):D876–D882. doi: 10.1093/nar/gkq963. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Li J.Z., Absher D.M., Tang H., Southwick A.M., Casto A.M., Ramachandran S., Cann H.M., Barsh G.S., Feldman M., Cavalli-Sforza L.L., Myers R.M. Worldwide human relationships inferred from genome-wide patterns of variation. Science. 2008;319:1100–1104. doi: 10.1126/science.1153717. [DOI] [PubMed] [Google Scholar]
  • 36.Pickrell J.K., Coop G., Novembre J., Kudaravalli S., Li J.Z., Absher D., Srinivasan B.S., Barsh G.S., Myers R.M., Feldman M.W., Pritchard J.K. Signals of recent positive selection in a worldwide sample of human populations. Genome Res. 2009;19:826–837. doi: 10.1101/gr.087577.108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Boniotto M., Tailleux L., Lomma M., Gicquel B., Buchrieser C., Garcia S., Quintana-Murci L. Population variation in NAIP functional copy number confers increased cell death upon Legionella pneumophila infection. Hum. Immunol. 2012;73:196–200. doi: 10.1016/j.humimm.2011.10.014. [DOI] [PubMed] [Google Scholar]
  • 38.Aganna E., Martinon F., Hawkins P.N., Ross J.B., Swan D.C., Booth D.R., Lachmann H.J., Bybee A., Gaudet R., Woo P. Association of mutations in the NALP3/CIAS1/PYPAF1 gene with a broad phenotype including recurrent fever, cold sensitivity, sensorineural deafness, and AA amyloidosis. Arthritis Rheum. 2002;46:2445–2452. doi: 10.1002/art.10509. [DOI] [PubMed] [Google Scholar]
  • 39.Hoffman H.M., Mueller J.L., Broide D.H., Wanderer A.A., Kolodner R.D. Mutation of a new gene encoding a putative pyrin-like protein causes familial cold autoinflammatory syndrome and Muckle-Wells syndrome. Nat. Genet. 2001;29:301–305. doi: 10.1038/ng756. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Kummer J.A., Broekhuizen R., Everett H., Agostini L., Kuijk L., Martinon F., van Bruggen R., Tschopp J. Inflammasome components NALP 1 and 3 show distinct but separate expression profiles in human tissues suggesting a site-specific role in the inflammatory response. J. Histochem. Cytochem. 2007;55:443–452. doi: 10.1369/jhc.6A7101.2006. [DOI] [PubMed] [Google Scholar]
  • 41.Allen I.C., Scull M.A., Moore C.B., Holl E.K., McElvania-TeKippe E., Taxman D.J., Guthrie E.H., Pickles R.J., Ting J.P. The NLRP3 inflammasome mediates in vivo innate immunity to influenza A virus through recognition of viral RNA. Immunity. 2009;30:556–565. doi: 10.1016/j.immuni.2009.02.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Ohno S. Springer; New York: 1970. Evolution by gene duplication. [Google Scholar]
  • 43.Allen I.C., TeKippe E.M., Woodford R.M., Uronis J.M., Holl E.K., Rogers A.B., Herfarth H.H., Jobin C., Ting J.P. The NLRP3 inflammasome functions as a negative regulator of tumorigenesis during colitis-associated cancer. J. Exp. Med. 2010;207:1045–1056. doi: 10.1084/jem.20100050. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Zaki M.H., Boyd K.L., Vogel P., Kastan M.B., Lamkanfi M., Kanneganti T.D. The NLRP3 inflammasome protects against loss of epithelial integrity and mortality during experimental colitis. Immunity. 2010;32:379–391. doi: 10.1016/j.immuni.2010.03.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Zhang P., Dixon M., Zucchelli M., Hambiliki F., Levkov L., Hovatta O., Kere J. Expression analysis of the NLRP gene family suggests a role in human preimplantation development. PLoS ONE. 2008;3:e2755. doi: 10.1371/journal.pone.0002755. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Murdoch S., Djuric U., Mazhar B., Seoud M., Khan R., Kuick R., Bagga R., Kircheisen R., Ao A., Ratti B. Mutations in NALP7 cause recurrent hydatidiform moles and reproductive wastage in humans. Nat. Genet. 2006;38:300–302. doi: 10.1038/ng1740. [DOI] [PubMed] [Google Scholar]
  • 47.Gérard N., Caillaud M., Martoriati A., Goudet G., Lalmanach A.C. The interleukin-1 system and female reproduction. J. Endocrinol. 2004;180:203–212. doi: 10.1677/joe.0.1800203. [DOI] [PubMed] [Google Scholar]
  • 48.Hsu L.C., Ali S.R., McGillivray S., Tseng P.H., Mariathasan S., Humke E.W., Eckmann L., Powell J.J., Nizet V., Dixit V.M., Karin M. A NOD2-NALP1 complex mediates caspase-1-dependent IL-1beta secretion in response to Bacillus anthracis infection and muramyl dipeptide. Proc. Natl. Acad. Sci. USA. 2008;105:7803–7808. doi: 10.1073/pnas.0802726105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Jin Y., Mailloux C.M., Gowan K., Riccardi S.L., LaBerge G., Bennett D.C., Fain P.R., Spritz R.A. NALP1 in vitiligo-associated multiple autoimmune disease. N. Engl. J. Med. 2007;356:1216–1225. doi: 10.1056/NEJMoa061592. [DOI] [PubMed] [Google Scholar]
  • 50.Magitta N.F., Bøe Wolff A.S., Johansson S., Skinningsrud B., Lie B.A., Myhr K.M., Undlien D.E., Joner G., Njølstad P.R., Kvien T.K. A coding polymorphism in NALP1 confers risk for autoimmune Addison's disease and type 1 diabetes. Genes Immun. 2009;10:120–124. doi: 10.1038/gene.2008.85. [DOI] [PubMed] [Google Scholar]
  • 51.Sironi M., Clerici M. The hygiene hypothesis: An evolutionary perspective. Microbes Infect. 2010;12:421–427. doi: 10.1016/j.micinf.2010.02.002. [DOI] [PubMed] [Google Scholar]
  • 52.Swanberg M., Lidman O., Padyukov L., Eriksson P., Akesson E., Jagodic M., Lobell A., Khademi M., Börjesson O., Lindgren C.M. MHC2TA is associated with differential MHC molecule expression and susceptibility to rheumatoid arthritis, multiple sclerosis and myocardial infarction. Nat. Genet. 2005;37:486–494. doi: 10.1038/ng1544. [DOI] [PubMed] [Google Scholar]
  • 53.Hampe J., Cuthbert A., Croucher P.J., Mirza M.M., Mascheretti S., Fisher S., Frenzel H., King K., Hasselmeyer A., MacPherson A.J. Association between insertion mutation in NOD2 gene and Crohn's disease in German and British populations. Lancet. 2001;357:1925–1928. doi: 10.1016/S0140-6736(00)05063-7. [DOI] [PubMed] [Google Scholar]
  • 54.Hugot J.P., Chamaillard M., Zouali H., Lesage S., Cézard J.P., Belaiche J., Almer S., Tysk C., O'Morain C.A., Gassull M. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn's disease. Nature. 2001;411:599–603. doi: 10.1038/35079107. [DOI] [PubMed] [Google Scholar]
  • 55.Ogura Y., Bonen D.K., Inohara N., Nicolae D.L., Chen F.F., Ramos R., Britton H., Moran T., Karaliuskas R., Duerr R.H. A frameshift mutation in NOD2 associated with susceptibility to Crohn's disease. Nature. 2001;411:603–606. doi: 10.1038/35079114. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Document S1. Figures S1–S8 and Tables S1 and S3
mmc1.pdf (1.6MB, pdf)
Document S2. Table S2
mmc2.xls (740.5KB, xls)

Articles from American Journal of Human Genetics are provided here courtesy of American Society of Human Genetics

RESOURCES