Genes and Immunity (2003) 4, 251–257 & 2003 Nature Publishing Group All rights reserved 1466-4879/03 $25.00 www.nature.com/gene Evolution of the beta 2 gene in primates

M Boniotto1, A Tossi2, M DelPero3, S Sgubin1, N Antcheva2, D Santon1, J Masters4 and S Crovella1 1Dipartimento di Scienze della Riproduzione e dello Sviluppo, Universita` di Trieste, Trieste, Italy; 2Dipartimento di Biochimica, Biofisica e Chimica delle Macromolecole, Universita` di Trieste, Trieste, Italy; 3Dipartimento di Biologia Animale e dell’Uomo, Universita` di Torino, Torino, Italy; 4Natal Museum, Pietermaritzburg, South Africa

With the aim of further investigating the molecular evolution of beta defensin genes, after having analysed beta defensin 1 (DEFB1) in humans and several nonhuman primate species, we have studied the evolution of the beta defensin 2 gene (DEFB2), which codifies for a peptide with antimicrobial and chemoattractant activity, in humans and 16 primate species. We have found evidence of positive selection during the evolution of orthologous DEFB2 genes at two points on a phylogenetic tree relating these primates: during the divergence of the platyrrhines from the catarrhines and during the divergence of the Cercopithecidae from the Hylobatidae, Great Apes and humans. Furthermore, amino acid variations in Old World Monkeys seem to centre either on residues that are involved in oligomerisation in the human molecule, or that are conserved (40–80%) in beta- in general. It is thus likely that these variations affect the biological function of the molecules and suggest that their synthesis and functional analysis might reveal interesting new information as to their role in innate immunity. Genes and Immunity (2003) 4, 251–257. doi:10.1038/sj.gene.6363958

Keywords: ; beta defensins; innate immunity; primates; nucleotide sequences; amino-acid sequences

Introduction Results The human beta defensin 2 (hBD2) was originally The alignment of nucleotide sequences of the primate isolated from the skin of psoriatic patients, and was DEFB2 coding region is shown in Table 1. We calculated found to be constitutively expressed in foreskin, lung the number of synonymous nucleotide substitutions per and trachea using the RT-PCR method.1 Moreover, synonymous site (dS) and the number of nonsynon- different authors have demonstrated that the hBD2 gene ymous nucleotide substitutions per nonsynonymous is upregulated upon inflammatory stimuli and its site (dN) by pairwise comparison among the 17 primate expression can subsequently be detected in many other DEFB2 sequences (Table 2). We restricted the analysis tissues.2–5 The peptide has a broad antimicrobial activity by studying only the sequence corresponding to the ranging from Gram negative bacteria to yeasts.1 It antimicrobially active, mature DEFB2. We hypothesised appears that hBD2 kills the pathogens via an electrostatic that only this region has been subject to a positive interaction between the oligomeric form with and a selection caused by direct interaction with pathogens negatively charged microbial surface.6 It has also been and cellular receptors. We found that the dN value was demonstrated that hBD2 shows a chemoattractant higher than dS in several of the nucleotide sequence activity for immature dendritic cells and memory T cells.7 comparisons, indicating that a positive selection pressure Different genes coding for proteins involved, directly acts at the level of this region (see Table 2). or indirectly, in the innate immune response have In order to identify at which node the active portion evolved rapidly by positive selection in response to of the molecule was altered under selection, we have changes in the pathogens to which the host is exposed.8,9 computed the number of synonymous (bS) and non- In this study, we investigate the molecular evolution of synonymous (bN) substitutions for each branch of the beta defensin 2 (DEFB2) gene in humans and 16 primate phylogenetic tree depicting the evolutionary relation species. Comparisons among orthologous DEFB2 genes among the primate species under investigation. For this from several primate species give new insights into the analysis, we used a phylogenetic tree that is widely evolution of the DEFB2 gene and the relation between accepted by primatologists (Figure 1).10 Several branches the structure of the protein and its function. showed greater values of bN than bS, but bN was significantly greater than bS in branch x (Z ¼ 2.56; Po0.05) and in branch y (Z ¼ 2.75; Po 0.05). These results indicate that positive selection has acted in the lineage leading to the clade including the homind and hylobatid families. The deduced amino-acid sequences for the primate Correspondence: Dr S Crovella, Dipartimento di Scienze della Riprodu- zione e dello Sviluppo, Universita` di Trieste, Via dell’Istria, 65/1- 34100 species are reported in Figure 2. Humans and Great Apes Trieste, Italy. E-mail: [email protected] (GA) showed identical amino-acid sequences, with the DEFB2 evolution in primate M Boniotto et al 252 Table 1 Alignment of nucleotide sequences from coding region of orthologous primate DEFB2 genes

HSS ATGAGGGTCTTGTATCTCCTCTTCTCGTTCCTCTTCATATTCCTGATGCCTCTTCCAGGTGTTTTTGGTGGTATAGGCGA PTR ...... A.... PPY ...... GGO ...... HMO ...... GC...... T..A....C...... A....A.GA. HCO ...... A....A.GA. HLA ...... C...... C....C...... A....A.GA. PME ...... C...... A....A.GA. PCR ....A....C...... A....A.GA. POB ....A....C...... A....A.GA. MMU ....A....C...... G...... A....A.GA. MFA ....A....C...... G...... A....A.GA. PPA ....A....C...... A....A.GA. CAE ....A....C...... C.C...... A....A.GA. CER ....A....C...... C...... A....A.GA. CPR ....A....C...... C...... A....C.GA. CJA ....A....C...... G...... A..C.A.GA.

HSS TCCTGTTACCTGCCTTAAGAGTGGAGCCATATGTCATCCAGTCTTTTGCCCTAGAAGGTATAAACAAATTGGCACCTGTG PTR ...... A...... PPY ...... GGO ...... HMO ...... T...... C...... HCO ...... T...... C...... HLA ...... T...... C...... PME ....A...... T...... G...... C...... T..C...GT..... PCR ....A...... T...... G...... G...... T..C...GT..... POB ....A...... T...... G...... C...... T..C...GT..... MMU ...... G...G...... T...... T.....G...... T...... GT..... MFA ...... G...G...... T...... T.....G...... T...... GT..... PPA ...... G...G...... T...... T.....G...... T...... GT..... CAE ...... G...... T...... G...... T..C...GT..... CER ...... G...... T...... G...... T..C...GT..... CPR ...... G...... T...... G...... T..C..TGT..... CJA ..T.T...... A..TG...... T...... G...... G...C...... T..C....T.....

HSS GTCTCCCTGGAACAAAATGCTGCAAAAAGCCATGA PTR ...... PPY ...... GGO ...... HMO ...... HCO ...... CG...... HLA ...... CG...... PME ..G..T...C..T...... PCR ..G..T...C..T...... POB ..G..T...C..T...... MMU ..G..T...C..T...... MFA ..G..T...C..T...... PPA ..G..T...C..T...... CAE .CG..T...C..T...... CER ..G..T...C..T...... CPR ..G..T...C..T...... CJA ..G.....CT..T......

exception of a Gly3-Ser substitution in the N-terminal based on the hBD2 sequence, relating these residues region seen only in P. troglodytes. The hylobatid sequences to the secondary structure elements is also provided show very little variation with respect to the hBD2. in Figure 2. Amino-acid changes found in the gibbon Three amino acid substitutions were observed in the (Hylobatidae) sequences, with respect to the human and N-terminal region of the active peptide, while only one GA sequences, are located exclusively at the N-terminus, conservative substitution was present in the C-terminal preceding the short a-helical stretch. Cercopithecid region. The Old World Monkeys (OWM, Cercopitheci- sequence variations are more extensive and mainly dae) are characterised by very similar amino acid involve the loops connecting the b-sheet, in particular sequences and showed a 71% average similarity with between b-strands 2 and 3. However, variations seem to hBD2. In the Cercopithecidae we found variations such be clustered on one side of each monomeric structure as Val6-Ile substitution in the Colobinae and Leu9– (see Figure 3). They do not greatly affect the hydro- Lys10-Val–Arg and His16-Leu substitutions in maca- phobicity of the molecules (the average per residue ques and baboons. Finally, the New World Monkey hydrophobicity according to the Eisenberg scale11 varies (NWM) Callithrix jacchus showed 60% similarity with from À0.2 in the PCR sequence to À0.22 in the human hBD2. and À0.23 in the PPA sequences, respectively). The To assess the functional implications of the observed charge varies from +5/+7 (depending on the charge state amino-acid changes in primate defensins, the positions of histidine) in the NWM sequence, to +6/+7 in the of affected residues were traced on the dimeric structure human or GA sequences, and to +7/+8 or +7/+9 in most of hBD26 as shown in Figure 3. A topological diagram, OWM sequences.

Genes and Immunity DEFB2 evolution in primate M Boniotto et al 253 Table 2 Synonymous (dS, below diagonal) and nonsynonymous (dN, above diagonal) distances between DEFB2 sequences of primates

HSS PTR PPY GOR HMO HCO HLAR PME PCR POB MMU MFA PPA CAE CER CPR CJA

HSS 0.011 0.000 0.000 0.038 0.056 0.056 0.147 0.160 0.147 0.174 0.174 0.174 0.147 0.147 0.141 0.236 PTR 0.035 0.011 0.011 0.033 0.050 0.050 0.140 0.153 0.140 0.167 0.167 0.167 0.140 0.140 0.141 0.236 PPY 0.000 0.035 0.000 0.038 0.056 0.056 0.147 0.160 0.147 0.174 0.174 0.174 0.147 0.147 0.141 0.236 GOR 0.000 0.035 0.000 0.038 0.056 0.056 0.147 0.160 0.147 0.174 0.174 0.174 0.147 0.147 0.141 0.236 HMO 0.090 0.111 0.090 0.090 0.016 0.016 0.102 0.114 0.102 0.127 0.127 0.127 0.102 0.102 0.103 0.206 HCO 0.108 0.128 0.108 0.108 0.017 0.000 0.122 0.134 0.122 0.148 0.148 0.148 0.122 0.122 0.122 0.229 HLAR 0.108 0.128 0.108 0.108 0.017 0.000 0.122 0.134 0.122 0.148 0.148 0.148 0.122 0.122 0.122 0.229 PME 0.130 0.152 0.130 0.130 0.036 0.053 0.053 0.011 0.000 0.044 0.044 0.044 0.022 0.022 0.022 0.159 PCR 0.091 0.112 0.091 0.091 0.000 0.017 0.017 0.000 0.011 0.055 0.055 0.055 0.032 0.032 0.033 0.146 POB 0.130 0.152 0.130 0.130 0.036 0.053 0.053 0.000 0.000 0.044 0.044 0.044 0.022 0.022 0.022 0.159 MMU 0.052 0.071 0.052 0.052 0.035 0.052 0.052 0.072 0.035 0.072 0.000 0.000 0.022 0.022 0.022 0.166 MFA 0.052 0.071 0.052 0.052 0.035 0.052 0.052 0.072 0.035 0.072 0.000 0.000 0.022 0.022 0.022 0.166 PPA 0.052 0.071 0.052 0.052 0.035 0.052 0.052 0.072 0.035 0.072 0.000 0.000 0.022 0.022 0.022 0.166 CAE 0.130 0.152 0.130 0.130 0.036 0.053 0.053 0.073 0.036 0.073 0.072 0.072 0.072 0.000 0.000 0.152 CER 0.090 0.111 0.090 0.090 0.000 0.017 0.017 0.036 0.000 0.036 0.035 0.035 0.035 0.036 0.000 0.152 CPR 0.149 0.193 0.149 0.149 0.072 0.089 0.089 0.111 0.073 0.111 0.109 0.109 0.109 0.111 0.072 0.140 CJA 2.365 2.525 2.365 2.365 n/c n/c n/c n/c n/c n/c 2.468 2.468 2.468 2.735 n/c n/c dN is underlined when it is greater than dS; n/c=not calculated.

Figure 1 Phylogenetic tree of the 17 primate species analysed. bn/bs ratios are shown above the branches. When both bn and bs are estimated to be 0 the ratio is not shown. Branches for which DEFB2 values of bn are significantly greater than those for bs are indicated as x and y.

The variable residues were also analysed in the context points on the phylogenetic tree: during the divergence of of positional conservation in beta defensin molecules in the platyrrhines (NWM) from the catarrhines (OWM) general, by aligning the 17 primate sequences reported and during the divergence of the Cercopithecidae from here with the 17 primate beta defensins 1 sequences,12 the Hylobatidae, Great Apes and humans (Figure 1). To hBD3, and 14 bovid, one pig, nine rodent and eight assess whether the variations at the amino-acid level avian beta defensin sequences obtained from the AMSDb could have functional implications, we analysed them in antimicrobial sequences database, for a total of 67 relation to the extensive data available for the structure sequences. The alignment of human and macaque and activity of the human peptide hBD2, whose structure DEFB2 sequences with selected sequences from other is shown in Figure 3.6 The monomeric structure consists mammals is shown in Figure 4. of a short N-terminal a-helix packed on top of a three- stranded b-sheet. This typical b-defensin scaffold depends mainly on its disulfide linkages to provide for Discussion the correct folding; only two small residues are sig- nificantly buried (Ala13 and Gly29 in hBD-2),13 and these In this study, we have found evidence of positive are highly conserved among beta defensins in general selection during the evolution of orthologous DEFB2 (see Figure 4). It is thus quite likely that all the primate genes. Evidence of positive selection was found in two DEFB2 molecules maintain this monomeric structure,

Genes and Immunity DEFB2 evolution in primate M Boniotto et al 254

Figure 2 Alignment of primate DEFB2 deduced amino-acid sequences. Dots indicate amino-acid identity to the human sequence. The asterisk indicates the beginning of the active mature peptide. The topological diagram shown above the sequences is based on the crystal structure of the human peptide.

Figure 3 Structure of human DEFB2. Fold and dimerisation of human BD2 (PDB entry 1FD3). Each monomer forms a three-stranded b-sheet with one helix flanking it at the N-terminus. The dimer is formed by interaction between strands b in each monomer, and involves residues Pro5 and Ala13 to Pro17 [6]. The approximate positions of a-carbons for residues involved in variations within primate sequences are indicated with spheres.

despite the reported amino-acid variations. At high in Figure 3. Dimerisation is stabilised by H-bonding concentrations, such as those required for crystallisation, interactions involving residues 5 and 13–17.6 In the and likely also those reached locally at the microbial crystal structure, these dimers further oligomerise to membrane, hBD2 forms six-stranded b-sheets consti- form compact octamers via H-bonding interactions tuted by the side-by-side association of two monomers involving residues Gly1, Gly3 and Asp4 of the N-terminal at the level of the first strand (b1) in each, as shown loop of one monomer and residues Thr7 and Lys10 in the

Genes and Immunity DEFB2 evolution in primate M Boniotto et al 255

Figure 4 Sequence alignment of mammalian beta defensins. The strictly conserved cysteines are highlighted in black, significantly conserved residues or residue types are highlighted in grey. The hBD2 numbering scheme is used. helical stretch from the other monomer, in each dimer. mechanism of membrane permeabilisation that involves It is conceivable that the variations observed in other accumulation and oligomerisation at the negatively primate DEFB2 sequences could affect dimerisation and/ charged bacterial surface. The observed N-terminal or oligomerisation and thus also antimicrobial activity. changes may thus imply that a particular form of Furthermore, human beta defensins have been shown oligomerisation has evolved specifically in apes and to exert chemoattractant activity on T-lymphocytes and humans. These changes might also have some effect dendritic cells because of a specific interaction with the on interaction with the chemokine receptor: the hBD2 CCR6 receptor for the b-chemokine MIP-3a/CCL20.7 For N-terminal stretch appears to be more similar to the hBD2, it has been suggested that the N-terminal stretch functionally significant DCCL stretch of the CCL20 from Asp4 to Leu9 resembles an equivalent Asp–Cys– molecule (Asp4 and Leu9 are within the 8.5 A˚ distance Cys–Leu (DCCL) stretch in the CCL20 important for observed for the DCCL motif15) than that of the other receptor activation, while residues 14–24 and 36–40 primate sequences, where the Asp residue is shifted to provide a region that could bind the receptor in a similar position 1. manner to a groove present in CCL20.14 Moreover, the The other changes observed in CJA and the Cerco- overall distribution of basic residues, which is also likely pithecidae occur mainly at the level of the C-terminus, important for chemokine binding, is similar in the two between residues 26 and 35 (see Figure 2). The stretch molecules. Thus, variations in the sequence might also 26–29 (QIGT) is quite conserved among different affect this aspect of biological activity. mammalian BD2 peptides (50–100% conservation de- When the principal residue variations in the hylobatid, pending on the residue, see Figure 4), and appears to OWM or NWM DEFB2 sequences are displayed on the have been maintained in the human sequence but to dimeric structure of hBD2, it is apparent that most have shifted away from the more common form in these variable residues cluster on one side of the monomeric primates. Thus, Gln26 and Thr29, respectively, observed structures (see Figure 3). Furthermore, while residues in 54 and 81% of mammalian BD2 peptides are present in involved in dimerisation (except His16-Leu in the the human, ape and gibbon sequences, but are replaced macaques) are unaffected, the Gly1-Asp, Gly3-Ser, by a His and a Val, respectively, in the other primates. Asp4-Asn substitutions seen in all primates except the Furthermore, Pro at position 33 (also present in the CJA great apes as well as the Lys10-Arg substitution sequence) is also moderately conserved (40% of mam- observed for some Cercopithecidae, involve those re- malian peptides), while Ser33 in the Cercopithecidae sidues concerned with oligomerisation in the human seems unique among mammals. peptide. The antimicrobial activity of the hBD2 has been The significance of these variations at a functional level proposed to occur via an electrostatic charge-based remains to be determined. Amino-acid substitutions

Genes and Immunity DEFB2 evolution in primate M Boniotto et al 256 in this region, among the different primate species, are City, CA, USA) with 1 U of Taq Gold (Applera Genomics) not localised in DEFB2 regions that determine dimerisa- and an annealing temperature of 551C. The pGEM-T tion in the human peptide, with the exception of the Easy Vector System (Promega Corporation, Madison, WI, macaque sequences, where the His16-Val substitution USA) was used for cloning of PCR products. Recombi- occurs in the b2 sheet (see MMU, MFA and PPA in Figure nant plasmids were propagated in DH5-a bacteria. A 2). They also do not appear to affect the region that has total of 10 recombinants were chosen and plasmids been proposed to mimic the receptor recognition groove extracted using standard protocols16 in order to exclude of CCL20,14,15 which is indicated as possibly important the possibility of obtaining sequences from paralogous for DEFB2 chemoattractant activity. Conversely, varia- genes or pseudogenes eventually present in the genome tions at the N-terminus clearly affect the potential of the species studied. DNA sequencing of the p-GEM receptor activation motif (see above), and also the plasmids was performed using the BigDye Terminator molecules’ cationicity. Cycle Sequencing Ready Reaction Kit v. 2.0 (Applera We have already investigated beta defensin 1 (DEFB1) Genomics). DNA sequences were detected and analysed gene evolution in humans and nonhuman primates.12 on an automated ABI Prism 3100 Genetic Analyser While evidence of positive selection has been found for (Applera Genomics). DEFB2 in nonhuman primates, we have not detected a Multiple alignments of the nucleotide and amino-acid positive selection during the evolution of DEFB1: in fact sequences were performed using the program CLUSTAL DEFB1 resulted more conserved than DEFB2 in all the X.18 Pairwise comparisons between nucleotide sequences nonhuman primate species analysed. The causes and were performed to investigate the extent of evolutionary consequences of these different conservation patterns constraint at the amino-acid level in the molecule. remain to be explained. The number of synonymous nucleotide substitutions In conclusion, we report that significant mutations per synonymous site (dS) and the number of nonsynon- have occurred during the evolution of primate DEFB2 ymous nucleotide substitutions per nonsynonymous site gene, in response to positive selection for different (dN) between each pair of orthologous sequences were products in humans, great apes and gibbons. Further- computed following the Nei and Gojobori’s method,19 more, amino-acid variations in OWMs seem to occur using the Mega 2.0 software.20 Finally, to investigate the at residues that are involved in oligomerisation, as possibility of the occurance of positive selection and observed in the human sequence, and at residues that to identify the evolutionary lineages in which it would are moderately conserved in beta defensins in general. It have occurred, the number of synonymous (bS) and is thus likely that these variations affect the biological nonsynonymous (bN) substitutions per site was esti- function of the molecules and suggest that their synthesis mated from the pairwise distances for each tree branch.21 and functional analysis might reveal interesting new The structure of hBD2 was obtained from the PDB information as to their role in innate immunity. database (entries 1FD3 and 1FD4).6 A positional residue frequency analysis for beta defensin sequences was performed on 67 sequences obtained from the AMSDb Materials and methods antimicrobial sequences database (www.bbcm.units. it/Btossi/), using the conserved cysteines for correct Both exons of DEFB2 and short portions of their alignment, to allow an estimate of residue conservation untranslated flanking and intronic regions have been in key positions. investigated in 16 primate species for interspecific variability. We analysed three species of great apes (Pan troglodytes (PTR), Gorilla gorilla (GGO) and Pongo pygmaeus (PPY)), three species of the family Hylobatidae Acknowledgements (Hylobates concolor (HCO), H. lar (HLA) and H. moloch This research was in part supported by grants from the (HMO)), nine species of Cercopithecidae (Macaca fasci- Italian Ministry of the Universities and Scientific Re- cularis (MFA), Macaca mulatta (MMU), Papio anubis (PPA), search (PRIN 2001). N Antcheva is supported by a grant Cercopithecus aethiops (CAE), C. erythrogaster (CER), C. from the EU PANAD project QLRT-2000-00411. The preussi (CPR), Presbytis cristata (PCR), P. obscurus (POB) collaboration of J Masters is made possible by the NRF and P. melalophos (PME)), and one platyrrhine species Science Liaison Centre, Pretoria. (Callithrix jacchus (CJA)). Genomic DNAs were extracted from liver, bone and muscle tissues using the phenol/chloroform protocol.16 DNA was also extracted from hairs using the Chelex 100 References method.17 Genomic amplification of orthologous DEFB2 genes was performed using primers designed on the 1 Harder J, Bartels J, Christophers E, Schroder JM. A peptide basis of the published human sequence (GenBank antibiotic from human skin. Nature 1997; 387: 861. accession number AF071216). For the first exon, the 2 Harder J, Meyer-Hoffert U, Teran LM et al. Mucoid Pseudo- forward primer was 50-CTGAATTCTAACCTCTGTAAT- monas aeruginosa, TNF-alpha, and IL-1beta, but not IL-6, GAGCAT-30 and the reverse primer was 50-AGA- induce human beta-defensin-2 in respiratory epithelia. Am GAACTTCTACGCCATTCTTCC-30. For the second J Respir Cell Mol Biol 2000; 22: 714– 721. 3 Hao HN, Zhao J, Lotoczky G, Grever WE, Lyman WD. exon, the forward primer was 50-CTGGAGAGCA- 0 0 Induction of human beta-defensin-2 expression in human GAAAAAGGTGTTT-3 and the reverse primer was 5 - astrocytes by and . J Neurochem 0 TGTTTTTCCAACATGAGTTTTGAGTT-3 . For both the 2001; 77: 1027–1035. exons 45 cycles of PCR were performed using the 4 Liu AY, Destoumieux D, Wong AV et al. Human beta-defensin- GeneAmp PCR System 9700 (Applera Genomics Foster 2 production in keratinocytes is regulated by interleukin-1,

Genes and Immunity DEFB2 evolution in primate M Boniotto et al 257 bacteria, and the state of differentiation. J Invest Dermatol 2002; 14 Perez-Canadillas JM, Zaballos A, Gutierrez J et al. NMR 118: 275–281. solution structure of murine CCL20/MIP-3alpha, a chemokine 5 Singh PK, Jia HP, Wiles K et al. Production of beta-defensins that specifically chemoattracts immature dendritic cells and by human airway epithelia. Proc Natl Acad Sci USA 1998; 95: lymphocytes through its highly specific interaction with 14961–14966. the beta-chemokine receptor CCR6. J Biol Chem 2001; 276: 6 Hoover DM, Rajashankar KR, Blumenthal R et al. The 28372–28379. structure of human beta-defensin-2 shows evidence of higher 15 Hoover DM, Boulegue C, Yang D et al. The structure of human order oligomerization. J Biol Chem 2000; 275: 32911–32918. inflammatory protein-3alpha /CCL20. Linking 7 Yang D, Chertov O, Bykovskaia SN et al. Beta-defensins: antimicrobial and CC chemokine receptor-6-binding activities linking innate and adaptive immunity through dendritic and with human beta-defensins. J Biol Chem 2002; 277: CCR6. Science 1999; 286: 525–528. 37647–37654. 8 Tanaka T, Nei M. Positive Darwinian selection observed at the 16 Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: A variable-region genes of immunoglobulins. Mol Biol Evol 1989; Laboratory Manual. Cold Spring Harbor Laboratory Press: 6: 447–459. New York, 1989. 9 Hughes AL, Ota T, Nei M. Positive Darwinian selection 17 Walsh PS, Metzger DA, Higuchi R. Chelex 100 as a medium promotes charge profile diversity in the antigen-binding cleft for simple extraction of DNA for PCR-based typing from of class I major-histocompatibility-complex molecules. Mol forensic material. Biotechniques 1991; 10: 506–513. Biol Evol 1990; 7: 515–524. 18 Aiyar A. The use of CLUSTAL W and CLUSTAL X for multiple 10 Purvis A. A composite estimate of primate phylogeny. Philos sequence alignment. Methods Mol Biol 2000; 132: 221–241. Trans R Soc London B Biol Sci 1995; 348: 405–421. 19 Nei M, Gojobori T. Simple methods for estimating the 11 Eisenberg D. Three-dimensional structure of membrane and numbers of synonymous and nonsynonymous nucleotide surface proteins. Annu Rev Biochem 1984; 53: 595–623. substitutions. Mol Biol Evol 1986; 3: 418–426. 12 Del Pero M, Boniotto M, Zuccon D et al. Beta-defensin 1 gene 20 Kumar S, Tamura K, Jakobsen IB, Nei M. MEGA2: molecular variability among non-human primates. Immunogenetics 2002; evolutionary genetics analysis software. Bioinformatics 2001; 53: 907–913. 17: 1244–1245. 13 Bauer F, Schweimer K, Kluver E et al. Structure determination 21 Zhang J, Rosenberg HF, Nei M. Positive Darwinian selection of human and murine beta-defensins reveals structural after gene duplication in primate ribonuclease genes. Proc Natl conservation in the absence of significant sequence similarity. Acad Sci USA 1998; 95: 3708–3713. Protein Sci 2001; 10: 2470–2479.

Genes and Immunity