MBE Advance Access originally published online on February 22, 2008
Molecular Biology and Evolution 2008 25(5):997-1001; doi:10.1093/molbev/msn049
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Articles |
Overdominance in the Human Genome and Olfactory Receptor Activity
Department of Genetics, Physical Anthropology and Animal Physiology, Faculty of Science and Technology, University of the Basque Country, Leioa, Bizkaia, Spain
E-mail: santos.alonso{at}ehu.es.
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResults and DiscussionAcknowledgementsReferences |
|---|
We investigate the contribution of overdominance to the maintenance of polymorphism in the human genome during the recent evolution of our species. Using the HapMap genotypic information, we have detected that the Gene Ontology term "olfactory receptor activity" is a molecular function overrepresented in genes that have SNPs (Single Nucleotide Polymorphisms) showing higher than expected number of heterozygotes in the HapMap populations. Our results suggest that the diversity of a subset of human olfactory receptors (ORs) may have been maintained by balancing selection, in the form of overdominance. This observation may suggest that the loss of OR genes during the evolution of the human lineage may have been accompanied by an increased capability to discriminate odorants with closely similar structures.
Key Words: overdominance • heterozygote advantage • balancing selection • olfactory receptors • HapMap
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResults and DiscussionAcknowledgementsReferences |
|---|
Understanding the mechanisms by which genetic variability is maintained is a central issue in evolution (Kimura 1983
; Hey 1999; Nei 2005). Nevertheless, it has been claimed that we remain essentially ignorant of the forces that maintain variation, despite its ease of measurement (Gillespie 1991). Among the forces that can maintain a polymorphism, balancing selection was a major candidate in the 1950’s and 1960’s (Lerner 1954; Dobzhansky 1955; Lewontin and Hubby 1966), but its appeal withered with the advent of the neutral theory of molecular evolution (Kimura 1983). With a renewed interest in selection, recent research has focused on identifying those regions of the human genome under positive directional selection (for a review, see Sabeti et al. 2006 and references therein), although, however, others claim that the driving force of evolution is mutation and that natural selection is only of secondary importance (Nei 2007). As regards balancing selection, Asthana et al. (2005) and Bubb et al. (2006) investigated if it could be responsible for maintaining a significant number of long-term polymorphisms in humans and found that it was unlikely at the human–chimpanzee timescale. However, at a more recent timescale, balancing selection in the form of overdominance has been documented in humans at least in a few cases. Thus, for instance, overdominance has been reported for HBB in relation to sickle-cell anemia (Allison 1956), as well as for other malaria-resistance associated genes (G6PD; Verrelli et al. 2002), the cystic fibrosis gene (Schroeder et al. 1995), or the gene associated with the prion disease "kuru" (PRPN) (Mead et al. 2003). Prompted by these observations, we decided to investigate the contribution of overdominance during the recent evolution of our species to the maintenance of polymorphism in the human genome. Given that in overdominance heterozygotes have a higher fitness than both homozygotes, a simple approach to detect this kind of selection is to assess deviations from Hardy–Weinberg proportions in the direction of an excess of heterozygotes (Hedrick 2005). Therefore, we calculated the ratio of observed heterozygosity to expected gene diversity for every autosomal coding nonsynonymous SNP in each of the 3 HapMap populations (CEU, YRI, and CHB + JPT). |
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResults and DiscussionAcknowledgementsReferences |
|---|
Populations
Genotypes for over 3 million SNPs are available from The HapMap project Web page for the following populations: 60 unrelated individuals from the Yoruba population in Ibadan, Nigeria (YRI), 45 unrelated individuals from Tokyo, Japan (JPT), 45 unrelated individuals from Han Chinese in Beijing, China (CHB), and 60 unrelated Utah residents with ancestry from northern and western Europe, from the Centre d'Etude de Polymorphisms Human (CEU). We grouped JPT and CHB data into one new group termed CHB + JPT.
Databases
To estimate the ratio of observed heterozygosity to expected gene diversity for every autosomal coding nonsynonymous SNP, we used the HapMap SNP information (release 22, based on National Center for Biotechnology Information build 36 and dbSNP, the Single Nucleotide Polymorphism database, build 126). Additional related information was downloaded from the UCSC, the University of California Santa Cruz, public MySQL server (ftp://hgdownload.cse.ucsc.edu/mysql/hg18/).
Genome coordinates for copy number variations (CNVs; build 36) were taken from http://projects.tcag.ca/variation/. Genome coordinates for segmental duplications (build 36) were obtained from http://projects.tcag.ca/humandup/, and pseudogenes genome coordinates (build 36) from http://www.pseudogene.org.
Genomic coordinates for olfactory receptor (OR) pseudogenes were obtained from HORDE (Human Olfactory Receptor Data Exploratorium), (Available at http://bioportal.weizmann.ac.il/HORDE/). We also used classifier for olfactory receptor pseudogenes (CORP), a probabilistic method for annotation of OR pseudogenes, in order to increase our chances to detect pseudogenized ORs (http://bioportal.weizmann.ac.il/HORDE/CORP/).
Coalescent Simulations
For the neutral simulations, we used the coalescent-based program ms by Hudson (2002). To correct for demography, we used an out-of-Africa model of human evolution. Demographic parameters were initially set to those reported by Schaffner et al. (2006). Iteratively, these values were modified until the pairwise FST distributions simulated under these values showed a good agreement with pairwise FST distributions obtained from real genomic data from the 3 major geographical human groups from the HapMap project (see Izagirre et al. 2006) (P values for the Kolmogorov–Smirnov D statistic: 0.891 for Caucasians vs. Asians, 0.104 for Caucasians vs. Africans, and 0.683 for Asians vs. Africans, which indicate that both simulated and real FST distributions are not statistically different). Finally, demographic parameters assumed an ancestral population size of 24 000 for the African population (population 1) and 7 700 for both Asians and Caucasians (populations 2 and 3, respectively), with a migration rates matrix Mij = {0, 0.05, 0.4, 0.1, 0 3, 0.8, 2.5, 0} for i and j values going from 1 to 3. Going backward in time, we assumed 2 bottlenecks with instant population reduction followed each by a population fusion: one at approximately 40 000 years ago, in which the Chinese population reduced its size to approximately one-sixth. About 2 000 years after this episode, the Chinese population fusses with the European population. Assuming a generation time of 20 years, this represents an F value of 0.04 for this bottleneck. A second population bottleneck takes place about 90 000 years ago. In this episode, the Eurasian population suffers a reduction in size to one-sixth of its previous size. About 10 000 years after this bottleneck (F = 0.21), the Eurasian population fusses with the African population.
Subsequent simulations to estimate the probability of a given ratio value were run under these demographic parameters using again ms program of Hudson (2002). We fixed the number of segregating sites to 1 and sample sizes were equal to the corresponding HapMap sample sizes of each population group. For each of the 10 000 simulation and for each sample set of 2n alleles, we formed n genotypes from pairs of consecutive alleles (i.e., first allele plus second allele formed genotype number 1 and so on). For each set of so formed genotypes, we scored the number of observed heterozygotes and calculated the expected heterozygosity from the sample allele frequencies. Finally, we obtained a distribution of 10 000 ratio values from which we estimated the cumulative probability of a given ratio value.
Gene Ontology Analysis
For the Gene Ontology (GO) analysis, we used Onto-Express (http://vortex.cs.wayne.edu/Projects.html). The reference list of genes consisted of all those genes for which nonsynonymous SNPs had been genotyped in HapMap, excluding both possible pseudogenes and those genes contained within segmental duplications and CNV regions. Lists of genes containing nonsynonymous SNPs with ratios higher than the established cutoff point were the query for the overrepresentation analysis. These lists were further refined as indicated in the main text. Molecular function terms were explored by means of the hypergeometric function. Multiple test correction used the false discovery rate approach implemented in Onto-Express.
|
Results and Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResults and DiscussionAcknowledgementsReferences |
|---|
Using the databases described in the Materials and Methods, we calculated the ratio of observed heterozygosity to expected gene diversity (hereafter, the "ratio" values for short) for every autosomal coding nonsynonymous SNP in each of the 3 HapMap populations. To determine the cutoff value against which to check the significance of observed ratio values, we took as references SNP rs334 at locus HBB and SNP rs1799990 at PRNP. The former is responsible for the overdominance described for hemoglobin-associated malaria resistance. The latter is responsible for a coding polymorphism at PRPN codon 129 that has been associated to overdominance in relation to the prion disease "kuru" (Mead et al. 2003
). In the HapMap YRI population, the ratios for rs334 and rs1799990 were 1.119 and 1.139, respectively. Simulations (Materials and Methods) show that a ratio value of 1.11 lies at the 90th percentile of a distribution of ratios, given the assumption of neutrality and a simplified demographic history, which takes into account an out-of-Africa model of human evolution (see Materials and Methods). An alternative approach using the distribution of ratios of intronic SNPs (presumably neutral) from HapMap YRI, shows that a ratio of 1.11 corresponds to the 89th percentile of the distribution. Thus, we decided to use a cutoff ratio of 1.11 for YRI (a malaria-endemic population) as this would have allowed us to detect rs334, at locus HBB, and SNP rs1799990, at PRNP, in a blind test. The cost of this choice is a false positive rate of approximately 10%. For CEU and CHB + JPT, we used a cutoff ratio of 1.12 and 1.125, respectively, which provide similar false positive rates. In addition, regions of the human genome containing CNVs, segmental duplications, and pseudogenes were removed from the analysis to avoid additional possible false positives (see Materials and Methods).
Rather than focusing on individual SNPs or genes, we investigated if there were particular GO terms that were overrepresented for genes with SNPs showing higher than expected ratios. The most salient feature of the GO analysis was the molecular function term "olfactory receptor activity," which was highly significant for the 3 population groups analyzed (multiple test corrected P values 2.6 x 10–5, 1.7 x 10–3, and 7.2 x 10–6 for YRI, CEU, and CHB + JPT, respectively) (data available as supplementary materials 1, 2, and 3 [Supplementary Material online], respectively). The claims that 1) some ORs, although structurally intact, could be functionally inactive due to mutations in key domains (Menashe et al. 2006); 2) some ORs might be segregating a nonfunctional allele (Menashe et al. 2006); and 3) that some ORs may not be specifically expressed in olfactory tissue (i.e., might not function as ORs) (Zhang et al. 2007), prompted us to repeat the GO analysis in order to account for these claims. Thus, we removed from our query and reference lists those ORs that 1) showed a CORP probability of being a pseudogene greater than or equal to 0.5, 2) are "segregating pseudogenes" (ORs with disruptions still segregating with the intact allele), and 3) are functional ORs but are not significantly expressed in olfactory epithelium (Zhang et al. 2007). Removing these ORs led to multiple test corrected P values of 0.014, 0.094, and 0.033 for YRI, CEU, and CHB + JPT, respectively. However, we noticed that 1 OR (OR2AG2) was classified within the molecular function class "unknown function" and therefore is excluded in the computation of the P value of the olfactory receptor activity class. OR2AG2 is, however, a functional gene and is expressed in the olfactory epithelium. To account for this, we substituted OR10C1 for OR2AG2. We chose OR10C1 for no particular reason except that this locus is considered as belonging to the olfactory receptor activity molecular function class by the GO classification system used by Onto-Express and is thus not excluded from the computation of the P value. Then, we repeated the GO analysis for the Caucasian set. We observed that this correction reduces the P value for olfactory receptor activity class in Caucasians to 0.050. Overall, this suggests that olfactory receptor activity is a molecular function enriched in genes with higher heterozygosity than expected.
Using OR pseudogenes as a control set of presumably neutral genes, we observe that at least in YRI, the proportion of functional ORs that show 1 or more SNPs with a ratio higher than 1.11 is significantly higher than the proportion of OR pseudogenes (not including segregating pseudogenes) satisfying the same condition (66 out of 228, a 28.95% vs. 88 out of 395, a 22.28%; Fisher exact test, 1-sided P value for the same or a stronger association 0.0397). This adds support to the idea that the enrichment of in genes with higher heterozygosity than expected within the OR class is not simply the result of generalized neutral evolution for ORs.
One may be concerned by the possibility that high sequence similarity among OR paralogs may result in a heterozygosity excess caused by the possible simultaneous amplification and genotyping of more than 1 OR paralog. The observation that few ORs are shared among the 3 populations plays down this possibility. Thus, out of 47 ORs that 1) were fully HORDE functional, 2) showed a significant expression in the olfactory epithelium, and 3) showed a significant high ratio (at 10% cutoff points), only 3 (OR2D3, OR5T2, and OR10G4) were shared among the 3 population groups. Three (OR5B3, OR5K3, and OR10G7) were shared between Africans and Asians, 1 between Africans and Caucasians (OR7G2), and 2 (OR5M11 and OR2AG2) between Caucasians and Asians. However, we decided to account for the possible simultaneous amplification and genotyping of more than 1 OR paralog by removing from the reference and query lists of OR genes those loci that have paralogs with a similarity in sequence greater than 70%, a figure close to the average sequence similarity among OR paralogs (72.3%). After applying these conditions, we were left with 20 ORs; of these, only OR2D3 remained shared among the 3 populations and only OR5M11 remained shared between Caucasians and Asians (see table 1)
|
Under these conditions, we obtained multiple test corrected P values for the molecular function term olfactory receptor activity of 0.023, 0.133, and 0.170 for YRI, CEU, and CHB + JPT, respectively. Thus, under these more stringent conditions only YRI show significant evidence for olfactory receptor activity as a molecular function class enriched in genes showing overdominance.
If instead of a 10%, we used a reduced false positive rate of 5% (cutoff ratios 1.16 for YRI, 1.165 for CEU, and 1.167 for CHB + JPT; cutoff points obtained by simulations), multiple test corrected P values for the molecular function term olfactory receptor activity remain significant for YRI (P = 0.008) and CHB + JPT (P = 0.006) but not for CEU (P = 0.102; P = 0.062 after correction for OR2AG2). Further removal from our query and reference lists of 1) those ORs that showed a CORP probability of being a pseudogene greater than or equal to 0.5, 2) that are segregating pseudogenes, or 3) that are not significantly expressed in olfactory epithelium, resulted in multiple test corrected P values of 0.034 for YRI, 0.5 for CEU (P = 0.382 after correction for OR2AG2), and 0.318 for CHB + JPT. Finally, exclusion of ORs having paralogs with greater than 70% sequence similarity resulted in a multiple test corrected P value of 0.073 for YRI (no positives were left for CEU or CHB + JPT). A list of OR loci positive under these conditions is shown in table 1. Although strictly nonsignificant, this multiple test corrected P value borders significance and indicates that although at this level of significance we cannot strictly say that the olfactory receptor activity molecular function class as a whole is under overdominance, at least a subset of genes belonging to this class seem to have clearly evolved under this evolutionary force. Besides, these 5% cutoff point values may be too stringent as, for instance, those loci that are textbook examples of overdominance in the human genome, like HBB, would not have been detected using a 5% significance level.
As regards the evolution of the OR gene family, Nimura and Nei (2003, 2005a, 2005b, 2007) proposed that this family is subject to a birth-and-death model of evolution, in which new genes are formed by gene duplication and some of the duplicate genes differentiate in function, whereas others are inactivated or deleted from the genome. At the population level, Gilad, Bustamante, et al. (2003) proposed that a subset of human ORs might have been the target of positive selection. Our results may suggest that, at least in the African YRI population, the diversity of an additional subset of ORs may have been maintained by balancing selection, in the form of overdominance. In the Euro-Asiatic populations, this signature may have been erased due to the population bottleneck associated to the out-of-African exodus.
Our observation seems to make sense in the light of the proposal by Lancet (1994), according to which, heterozygosity in ORs could potentially double the number of different odorant-binding sites encoded in the genome. This proposal is of particular relevance in the context of locus and allele exclusion, that is, the observation that in mammals each olfactory neuron expresses just 1 allele from a single OR gene (Shykind 2005). Thus, heterozygosity in ORs would be important for the olfactory acuity of the individual. Further, if structurally related ORs recognize overlapping sets of odorants (Malnic et al. 1999; Kajiya et al. 2001), the increased intraindividual genetic diversity in, at least, a subset of ORs may lead to an increased capability to discriminate among closely structurally related odorants. Actually, Kelly et al. (2007) have shown that polymorphisms in OR7D4 account for a significant proportion in the perception of steroidal odors. Thus, this increased capability may be a specialization response to the substantial loss of functional ORs in humans in comparison to other primates (Gilad, Man, et al. 2003) or to the mouse (Niimura and Nei 2005a, 2007). It can be argued, however, that the presence of this kind of selection is at odds with so many OR gene losses during human evolution. In this regard, the conservation of duplicated loci, or the promotion of gene duplication itself, would seem an easier mechanism to increase the sensitivity to odorants than the maintenance of different homologous alleles. However, if we have to implement a system of locus exclusion so that only a single OR specificity appears in a single olfactory neuron, if this has to be achieved by silencing all other ORs in that neuron, and if this silencing is accomplished by the own OR protein being expressed (Serizawa et al. 2003), it seems that the cost of silencing all other ORs would impose a transcriptional burden on the chosen OR. This burden would be higher as the number of OR paralog clusters increases. In turn, maintenance of different homologous alleles doubles the potential number of ORs and comes at no additional transcriptional cost. Contrarily, it can also be argued that, given that olfaction seems to be more important for mice than for humans, then why so many OR genes have been maintained as different loci in mice in the existence of this cost of gene silencing? In this regard, Nozawa et al. (2007) showed that variation in the number of OR genes can be only partly due to physiological requirements and that genomic drift (random duplication and deletion of genes) can also be responsible for a substantial part of this variation. Using CNV variation in ORs as a proxy for genomic drift, Nozawa et al. (2007) showed that, in humans, the proportion of polymorphic functional ORs is 30% and 35% for OR pseudogenes. This nonsignificant difference suggests that, if OR pseudogenes evolve in a neutral fashion, then copy number changes, and by extension, variation in the number of OR genes, have also occurred in a neutral fashion, and therefore, the number of OR genes may not be directly related to fitness (Nozawa et al. 2007). Thus, in the context of the cost of gene silencing, maintaining 
350 new genes by genomic drift in could amount to maintaining moderately deleterious alleles at moderate frequencies in the presence of genetic drift.
In any case, maintenance of different homologous alleles and conservation of duplicated genes need not be mutually exclusive mechanisms. Rather, they may be part of the same tool kit of evolution.
Finally, it still remains to be assessed how the differences in odorant perception and/or discrimination can affect the fitness of human individuals. It has always been assumed that the ability to detect odorant molecules is an essential survival tool in many behavioral processes, including reproduction and predation (Dryer 2000), but it has been proposed that additional factors, like experience and memory, can also affect olfactory perception (Wilson and Stevenson 2003). Perhaps, the contribution to fitness may be evaluated in the future in model animals, along with the assessment of their intraspecific sequence variation in OR genes.
|
Acknowledgements |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResults and DiscussionAcknowledgementsReferences |
|---|
The authors wish to thank 2 anonymous reviewers, whose comments substantially improved the manuscript. Work funded by grants UPV05/150 from the University of the Basque Country and IP-453-07 from the Basque Government.
|
Footnotes |
|---|
Jianzhi Zhang, Associate Editor
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResults and DiscussionAcknowledgementsReferences |
|---|
Allison AC. The sickle-cell and haemoglobin C genes in some African populations. Ann Hum Genet (1956) 21:67–89.[Web of Science][Medline]
Asthana S, Schmidt S, Sunyaev S. A limited role for balancing selection. Trends Genet (2005) 21:30–32.[CrossRef][Web of Science][Medline]
Bubb KL, Bovee D, Buckley D. (12 co-authors). Scan of human genome reveals no new loci under ancient balancing selection. Genetics (2006) 173:2165–2177.
Dobzhansky T. A review of some fundamental concepts and problems of population genetics. Cold Spring Harb Symp Quant Biol (1955) 20:1–15.
Dryer L. Evolution of odorant receptors. Bioessays (2000) 22:803–810.[CrossRef][Web of Science][Medline]
Gilad Y, Bustamante CD, Lancet D, Pääbo S. Natural selection on the olfactory receptor gene family in humans and chimpanzees. Am J Hum Genet (2003) 73:489–501.[CrossRef][Web of Science][Medline]
Gilad Y, Man O, Pääbo S, Lancet D. Human specific loss of olfactory receptor genes. Proc Natl Acad Sci USA (2003) 100:3324–3327.
Gillespie JH. The causes of molecular evolution. Oxford series in ecology and evolution (1991) New York: Oxford University Press.
Hedrick PW. Genetics of populations (2005) Sudbury (MA): Jones and Bartlett publishers.
Hey J. The neutralists, the fly and the selectionist. Trend Ecol Evol (1999) 14:35–38.[CrossRef][Medline]
Hudson RR. Generating samples under a Wright-Fisher neutral model of genetic variation. Bioinformatics (2002) 18:337–338.
Izagirre N, García I, Junquera C, de la Rúa C, Alonso S. A scan for signatures of positive selection in candidate loci for skin pigmentation in humans. Mol Biol Evol (2006) 23:1697–1706.
Kajiya K, Inaki K, Tanaka M, Haga T, Kataoka H, Touara K. Molecular bases of odor discrimination: reconstitution of olfactory receptors that recognize overlapping sets of odorants. J Neurosci (2001) 21:6018–6025.
Kelly A, Zhuang H, Chi Q, Vosshall LB, Matsunami H. Genetic variation in a human odorant receptor alters odour perception. Nature (2007) 449:468–472.[CrossRef][Medline]
Kimura M. The neutral theory of molecular evolution (1983) Cambridge (UK): Cambridge University Press.
Lancet D. Olfaction. Exclusive receptors. Nature (1994) 372:321–322.[CrossRef][Medline]
Lerner IM. Genetic homeostasis (1954) Edinburgh (UK): Oliver & Boyd.
Lewontin RC, Hubby JL. A molecular approach to the study of genic heterozygosity in natural populations. II. Amount of variation and degree of heterozygosity in natural populations of Drosophila pseudoobscura. Genetics (1966) 54:595–609.
Malnic B, Hirono J, Sato T, Buck LB. Combinatorial receptor codes for odours. Cell (1999) 96:713–723.[CrossRef][Web of Science][Medline]
Mead S, Stumpf MPH, Whitfield J. (11 co-authors). Balancing selection at the prion protein gene constituent with prehistoric kuru-like epidemics. Science (2003) 300:640–643.
Menashe I, Aloni R, Lancet D. A probabilistic classifier for olfactory receptor pseudogenes. BMC Bioinformatics (2006) 7:393.[CrossRef][Medline]
Nei M. Selectionism and neutralism in molecular evolution. Mol Biol Evol (2005) 22:2318–2342.
Nei M. The new mutation theory of phenotypic evolution. Proc Natl Acad Sci USA (2007) 104:12235–12242.
Niimura Y, Nei M. Evolution of olfactory receptor genes in the human genome. Proc Natl Acad Sci USA (2003) 100:12235–12240.
Niimura Y, Nei M. Evolutionary changes of the number of olfactory receptor genes in the human and mouse lineages. Gene (2005a) 346:23–28.[CrossRef][Web of Science][Medline]
Niimura Y, Nei M. Evolutionary dynamics of olfactory receptor genes in fishes and tetrapods. Proc Natl Acad Sci USA (2005b) 102:6039–6044.
Niimura Y, Nei M. Extensive gain and losses of olfactory receptor genes in mammalian evolution. PLoS ONE (2007) 2:e708.[CrossRef]
Nozawa M, Kawahara Y, Nei M. Genomic drift and copy number variation of sensory receptor genes in humans. Proc Natl Acad Sci USA (2007) 104:20421–20426.
Sabeti PC, Schaffner SF, Fry B, Lohmueller J, Varily P, Shamovsky O, Palma A, Mikkelsen TS, Altshuler D, Lander ES. Positive natural selection in the human lineage. Science (2006) 312:1614–1620.
Schaffner SF, Foo C, Gabriel S, Reich D, Daly D, Altshuler D. Calibrating a coalescent simulation of human genome sequence variation. Genome Res (2006) 15:1576–1583.[CrossRef][Web of Science]
Schroeder SA, Gaughan DM, Swift M. Protection against bronchial asthma by CFTR delta F508 mutation: a heterozygosity advantage in cystic fibrosis. Nat Med (1995) 1:703–705.[CrossRef][Web of Science][Medline]
Serizawa S, Miyamichi K, Nakatani H, Suzuki M, Saito M, Yoshihara Y, Sakano H. Negative feedback regulation ensures the one receptor-one olfactory neuron rule in mouse. Science (2003) 302:2088–2094.
Shykind BM. Regulation of odorant receptors: one allele at a time. Hum Mol Genet (2005) 14:R33–R39.
Verrelli BC, McDonald JH, Argyropoulus G, Destro-Bisol G, Froment A, Drousioutou A, Lefranc G, Helal AN, Loiselet J, Tishkoff SA. Evidence for balancing selection from nucleotide sequence analyses of human G6PD. Am J Hum Genet (2002) 71:1112–1128.[CrossRef][Web of Science][Medline]
Wilson DA, Stevenson RJ. The fundamental role of memory in olfaction perception. Trends Neurosci (2003) 26:243–247.[CrossRef][Web of Science][Medline]
Zhang X, de la Cruz O, Pinto JM, Nicolae D, Firestein S, Gilad Y. Characterizing the expression of the human olfactory receptor gene family using a novel DNA microarray. Genome Biol (2007) 8:R86.[CrossRef][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. M. Andres, M. J. Hubisz, A. Indap, D. G. Torgerson, J. D. Degenhardt, A. R. Boyko, R. N. Gutenkunst, T. J. White, E. D. Green, C. D. Bustamante, et al. Targets of Balancing Selection in the Human Genome Mol. Biol. Evol., December 1, 2009; 26(12): 2755 - 2764. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Fumagalli, U. Pozzoli, R. Cagliani, G. P. Comi, S. Riva, M. Clerici, N. Bresolin, and M. Sironi Parasites represent a major selective force for interleukin genes and shape the genetic predisposition to autoimmune conditions J. Exp. Med., June 8, 2009; 206(6): 1395 - 1408. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||