Chemokine ligand–receptor binding relationships. Five subfamilies of chemokines, CXC, CC, XC, CX3C, and CX, have been recognized on the basis of the arrangement of the two N-terminal residues of four conserved cysteines. One and three amino acids separate the first and second cysteines in the CXC and CX3C chemokines, respectively, whereas the two cysteines are adjacent to each other in the CC subfamily. The XC (or C) subfamily lacks the first and paired third cysteine residues. The fifth subfamily, CX, which has so far been identified only in zebrafish, lacks one of the two N-terminal cysteine residues but retains the third and fourth (Nomiyama et al. 2008). Chemokines can also be functionally classified into several groups, based on their mode of expression and function (Zlotnik & Yoshie 2000; Moser et al. 2004; Mantovani et al. 2006). These groups are shown in different colors. Both the common names and systematic nomenclature are shown in the figure. Recently, Islam et al. (2011) showed that mouse Ccl8 serves as an agonist for Ccr8 but not for Ccr2, whereas human CCL8 binds CCR2 but not CCR8. We previously proposed that the mouse ortholog of human CCL8 is Ccl12 rather than Ccl8 (Nomiyama et al. 2003). Thus, mouse Ccl8 is now likely to be a mouse-specific gene without a human counterpart. Consistent with this, mouse Ccl12 has been shown to bind Ccr2, as does human CCL8 (Sarafi et al. 1997). Therefore, the mouse genes require renaming. The other discrepancies between human and mouse chemokine gene names are described in our previous review (Nomiyama et al. 2010). All known chemokine receptors are seven-transmembrane G protein-coupled receptors. Chemokine receptors are classified according to their ability to bind a specific subclass of chemokines (CXCR, CCR, XCR, and CX3CR) (Murphy et al. 2000). However, mouse Cxcr3 and human CX3CR1 have been shown to bind ligands of a different subclass, mouse Ccl21 (Soto et al. 1998) and human CCL26 (Nakayama et al. 2010), respectively, in addition to their cognate ligands. The receptor(s) for the CX chemokines has not yet been identified. Thus far, 18 signaling chemokine receptors have been identified in the human genome. Besides these classic chemokine receptors, five atypical (nonsignaling) chemokine receptors have been identified (DARC, CCBP2, CCRL1, CCRL2, and CXCR7) (Graham 2009; Leick et al. 2010; Naumann et al. 2010). These atypical receptors bind chemokines but do not elicit standard chemotactic responses after ligand binding. Both DARC and CCBP2 primarily bind inflammatory chemokines of the CXC and CC subfamilies. The ligand specificity of the receptors shown here may change by post-translational modification of the ligands (Mortier et al. 2008).

Number of chemokine and chemokine receptor genes identified in vertebrate genomes. We had previously identified chemokine genes from 10 mammalian genomes (Nomiyama et al. 2010). Here, we have omitted five genomes from the survey list but added two genomes to cover a wide range of mammals. In total, we searched seven mammalian genomes (human, mouse, cow, elephant, opossum, wallaby, and platypus) for the analyses. In addition, three birds (chicken, zebra finch, and duck), a reptile (anole lizard), an amphibian (Xenopus), four teleost fish (medaka, stickleback, zebrafish, and Tetraodon), a cartilaginous fish (elephant shark) and a jawless fish (sea lamprey) were included in the survey. Phylogenetic relationships of these organisms among chordates are shown. More detailed taxonomic classifications are shown in Fig. S1 in Supporting Information. The first split in vertebrates occurred between jawed and jawless vertebrates (gnathostomes and agnathans), followed by the divergence of jawed vertebrates into cartilaginous and bony fish (chondrichthyes and osteichthyes). Divergence times (Mya, million years ago) (Hedges & Kumar 2003) are not to scale. A hypothetical origin time for the adaptive immune system is indicated. The timings of the two successive rounds of WGD (1R and 2R) and the teleost-specific WGD (3R) are also shown. Although the timing of the 2R has long been in dispute, Kuraku et al. (2009) recently showed that both 1R and 2R occurred before the split between jawed and jawless vertebrates. Recent studies indicate that tunicates (previously known as urochordates) are the invertebrates most closely related to vertebrates (Delsuc et al. 2006). The amino acid sequences of the chemokines and their database accession numbers are shown in Fig. S1A in Supporting Information. The chemokine receptor sequences and their accession numbers (Nomiyama et al. 2011) have been updated and are shown in Fig. S1B in Supporting Information. Phylogenetic trees of vertebrate chemokines and chemokine receptors are shown in Fig. S4 in Supporting Information.

Conserved synteny analysis of vertebrate chemokine CCL25. (A) Comparative maps of CCL25 gene regions. Fish-specific CCL44 genes are also shown. Arrows indicate transcriptional orientation. Comparative maps of other chemokine and chemokine receptor genes are shown in Fig. S3 in Supporting Information. (B) Conserved synteny dot plots. The plots were drawn using the Synteny Database (Catchen et al. 2009, 2011). In the plots, fish orthologs of genes on Hsa19 (0–20 Mb) are indicated as red crosses on fish chromosomes in the order found on the human chromosome (gene orders on fish chromosomes are different from those of humans). Correspondence of the chromosomes among teleosts containing the CCL25 genes was examined as follows. First, the CVL number was obtained using the human CCL25 gene locus (chromosome 19, 8.1 Mb) (Nakatani et al. 2007). Supplemental Fig. S2 in reference 20 shows that the human genes in this CVL block “88” are orthologous to the genes on medaka chromosomes 4 and 17, where the two medaka CCL25 genes are located. Next, protochromosome numbers of the teleost, gnathostome, and vertebrate ancestors (m, A1 and A, respectively) were identified from Supplemental Table S2 in reference 20. Using the teleost protochromosome number, the orthologous chromosomes of the three teleosts were then identified in the Supplemental information of reference 25. In case of teleost-specific genes, teleost protochromosome numbers and correspondence of the chromosomes in each teleost can be obtained by consulting the Supplemental information in reference 25. Dre2, Ola17, and Tni15, all containing CCL25a, were derived from one of the duplicated chromosomes of teleost protochromosome m. Dre11, Ola4, and Tni1, all containing CCL25b, were derived from another duplicated copy of the same protochromosome. Dre11 and Dre22 are the products of chromosome fission. Human and teleost chromosomes containing the CCL25 regions were all derived from gnathostome protochromosome A1 and also from vertebrate ancestral chromosome A (see also Fig. 4). Synteny dot plots of other chemokine and chemokine receptor genes and the CVL numbers are shown in Fig. S5 and Table S4 in Supporting Information, respectively.

Proposed ancestry of vertebrate chemokine CCL25. Vertebrate protochromosome A, on which an ancestral CCL25 gene is assumed to reside, was duplicated by the 1R and 2R WGDs and also by a fission event between the 1R and 2R, resulting in six gnathostome protochromosomes (Nakatani et al. 2007). The CCL25 gene on gnathostome protochromosome A1 was transferred to teleost and amniote protochromosomes, whereas the genes on the other gnathostome protochromosomes were lost. Fish-specific CCL44 must have been generated by tandem duplication of the ancestral CCL25 gene on teleost protochromosome m. Two copies of the teleost CCL25 gene were maintained on duplicated chromosomes, and one of the CCL44 copies on one of the duplicated chromosomes may have been lost. Tetraodon CCL44 and bird CCL25 genes have not yet been identified. The evolutionary history of other chemokine and chemokine receptor genes are shown in Fig. S6 in Supporting Information. 3R indicates the teleost-specific WGD.

Vertebrate ancestral genes for chemokines and chemokine receptors. The vertebrate and gnathostome protochromosomes on which chemokine and chemokine receptor genes were localized are shown. Among the genes contained by sea lamprey and elephant shark, only those that are shared by vertebrate or gnathostome ancestors are shown. Because the genome sequences of sea lamprey and elephant shark are still fragmented, it is not known whether the genes are linked on the same chromosomes. The lines that link the vertebrate chemokine ancestors with the chemokine receptor ancestors indicate the ligand–receptor relationships based on the human chemokine system. The receptor for CXCL14 and the ligand(s) for CXCR3L have not yet been identified. The ligand(s) for the ancestral XCR1 is unknown. The predicted locations of HOX gene clusters on the protochromosomes are also indicated. Colocalization of HOX clusters with chemokine receptor genes on vertebrate protochromosome E shows that some of the chemokine receptor genes accompanied HOX cluster duplication (DeVries et al. 2006). HOXD, HOXA, and HOXB on gnathostome protochromosomes E0, E1, and E2, respectively, were omitted for simplicity (see also Fig. S6 in Supporting Information). 1R and 2R indicate the two successive rounds of WGD.