Fish & Shellfish Immunology 35 (2013) 2040e2045
Contents lists available at ScienceDirect
Fish & Shellfish Immunology journal homepage: www.elsevier.com/locate/fsi
Short communication
The evolutionary analysis on complement genes reveals that fishes C3 and C9 experience different evolutionary patterns Shanchen Wang, Rixin Wang, Tianjun Xu* Laboratory of Fish Biogenetics & Immune Evolution, College of Marine Science, Zhejiang Ocean University, Zhoushan, PR China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 August 2013 Received in revised form 14 October 2013 Accepted 21 October 2013 Available online 30 October 2013
Complement is a humoral factor of innate immunity and plays an essential role in altering the host of the presence of potential pathogens and clearing of invading microorganisms. The third complement component (C3) not only is regarded as the crossing of the three pathways of complement activation, but also serves one of the bridges linking innate and acquired immunity. The nine complement component (C9) can combine with C5b, C6, C7 and C8 to form MAC which bounds to the surface of microorganisms to kill them. The evidence of evolution on C3 genes which have multiple functions and plays central role in innate immunity was documented in our previous study. Now we were interested in the evolution of C9 genes which were the terminal complement components. For these reasons, we want to explore the evolutionary patterns of C9 and whether C3 and C9 experience different evolutionary patterns. In our study, we used the sliding window method to separately calculate the values of u among fishes and mammals of C3 and C9 codons. In order to detect the positive selection sites, we used the maximum likelihood (ML) method to study the evolutionary pattern on C3 and C9 genes. Positive selection sites were detected in mammalian C9 genes and no positive selection sites were detected in fishes C9 genes. However, no positive selection sites were detected in mammalian C3 genes and positive selection sites were detected in fishes C3 genes. The result indicated that C3 and C9 had different evolutionary patterns on mammals and fishes. In conclusion, different living environments lead to different evolutionary patterns on C3 and C9 in mammals and fishes. Besides, different complement components may have different evolutionary patterns on mammals and fishes. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Innate immunity Adaptive immunity C3 C9 Molecular evolution
1. Introduction The complement system was first described as a heat-sensitive factor in fresh serum that ‘complemented’ the effect of specific antibody in the lysis of bacteria and red blood cells [1]. Now, the complement system, which consists of more than thirty-five plasma and cell-surface proteins, is known to be a highly sophisticated host-defense system that plays key roles in innate and adaptive immunity [2,3]. Activation of the complement is initiated through any of three distinct pathways: the classical, alternative and lectin pathways [4,5]. The classical pathway is activated by immune complexes formed between antigens and antibodies [6,7]. The alternative pathway is directly activated by a large number of microorganisms [8,9]. The lectin pathway requires the interaction of lectins such as mannose-binding lectin (MBL) and ficolins, with sugar moieties based on the surface of microorganism [10,11].
* Corresponding author. Tel./fax: þ86 580 2550826. E-mail address: [email protected] (T. Xu). 1050-4648/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fsi.2013.10.018
Activation of the complement system through any the three pathways results in the activation of C3. The first complement component, C1, is Ca2þ dependent protein complex that consists of C1r and C1s bound to one molecule of C1q. When surface-bound IgM or IgG binds to C1q, the classical pathway is activated. Activated C1s cleaves C4 and C2 into C4a, C4b, C2a and C2b, respectively. The C4bC2a enzymatic complex is known as C3 convertase which cleaves C3 into C3a and C3b [12e14]. C3 belongs to A2M alpha-2-macroglobulin superfamily and plays a central role in complement activation which emerged over 700 million years ago [15]. C3b can become covalently bound to the cell surface. Once bound, C3b can form a complex with C4bC2a to give rise to C5 convertase. Then C5 is cleaved into C5a and C5b. C5a plays an important role in the inflammation process and C5b initiates the self-assembly of the MAC by subsequent binding of C6, C7, C8 and C9. MAC can create a pore in the cell membrane leading to cell lysis and death. With the deep study, complement molecules have been found and studied in a variety of organisms, principally in vertebrates. The homologs of C3 have been found in invertebrates [16e18]. These findings proved that the origin and evolution of the complement
S. Wang et al. / Fish & Shellfish Immunology 35 (2013) 2040e2045
system are traced to the earliest radiations of the animal kingdom [19,20]. Now the study on complement system is on evolutionary process rather than molecular structure and cloning. In our previous study, we revolve around the complement component to research. The evidence of evolution on C3 genes was reported. We found that three periods of positive selection events had happened on C3 genes during evolutionary history and C3 genes experienced different evolutionary patterns on fishes and mammals for their living environments [21]. Now, the evolution of C9 genes which are the terminal complement components assembling with C5b, C6, C7 and C8 to form MAC is attracted us. Besides, many fishes C9 genes were cloned and studied, for example, Paralichthys olivaceus [22], Fugu rubripes [23], Cyprinus carpio [24], Oncorhynchus mykiss [25] and so on. So we want to explore the evolutionary patterns and evolutionary mechanism of C9 genes. Besides, it is known that C3 not only is the crossing of the three pathways of complement activation, but also serves as one of the bridges linking innate and acquired immunity [26]. The activated C3 is able to covalently bind to foreign microorganisms, which are presented to the T cells to result in cellular immune response, or further lysed by the MAC. However, C9 is a single-chain glycoprotein that is involved in the formation of the MAC on the surface of target cells [27,28]. The MAC is assembled by C5b, C6, C7, C8 and C9, once assembled on the target surface it forms transmembrane channels resulting in cell lysis [29,30]. Thus in the history of the evolution, the question on the relationship of C3 and C9 was attracted us. So we are interested in the evolutionary patterns on C3 and C9. In our study, we used the maximum likelihood (ML) method to study the evolution pattern on C3 and C9, respectively. We want to explore the evolution pattern on C9 genes and test whether C3 and C9 have difference on evolution. 2. Materials and methods 2.1. Taxonomic coverage We derived nine C9 sequences from nine fish species, twenty-five C9 sequences from twenty-five mammals, one C9 sequence from amphibian; nineteen C3 sequences from ten fishes, twelve C3 sequences from mammals, together with two C3 sequences from invertebrates and one from bird (Table S1). All the C9 and C3 sequences in our study were derived from GenBank (http://www.ncbi.nlm.nih. gov/Genbank/) and Ensemble (http://www.ensemble.org/) database for evolutionary analyses.
2041
indicative of neutral evolution, purifying selection and positive selection. The phylogenetic tree is used to detect whether different environment affects the evolution pattern on aquatic and terrestrial organisms in C9. In order to investigate the evolutionary process of C3 and C9 genes in mammals and fishes, PAML 4 [37] is used to estimate the ratio of u. The PAML package includes a free-ratio model, site-specific model, a branch-specific model (two-ratio) and a branch-site model (model A) [38]. The free-ratio model is used to allow the u ratio to vary on each branch which is used to test whether this model fitted the data significantly better than the one ratio model which is used to identify the selective pressures in all genes by the likelihood ratio test (LRT). Branch-site model was used to detect positive selection on the ancestral foreground lineages. The site model which includes six models of codons substitution is applied to subset of fishes and mammals C3 and C9 sequences to detect the possible positive selection sites among these lineages. The following codons substitution models were utilized: M0 (one ratio) with assuming a unique value of u across all sites, M3 (discrete) which has three discrete classes of site with different u values for sites in the phylogeny, M1a (nearly neutral model) with two site classes (0 < u < 1, u ¼ 1) estimating the frequency p0 and p1, M2a (positive selection) which added a class of positively selected sites comparing with M1a, M7(b) which do not allow the positive selection with assuming the frequencies of different u values (0 < u < 1) to follow a beta distribution whose shape is controlled by two parameters (p and q), M8 (b & u) which is similar to M7 by adding a proportion p1 with u > 1 [34]. In all cases, significant differences between models can be evaluated by calculating twice the difference of the log-likelihood values (2DlnL) following a chi-squared distribution with the number of degrees of freedom equal to the difference in the number of free parameters between models [39]. The Bayes empirical Bayes (BEB) is used to calculate the Bayesian posterior probability (BPP) of the codons sites under a positive selection [40]. In order to improve the accuracy of results, sites with BPP < 95% is regarded as positive selection sites. 2.4. Sliding windows analysis To explore further the distribution of positive selection sites on specific functional domains of C3 and C9 between fishes and mammals, we use the Nei and Gojobori method [41] to conduct the u values of sliding windows. Sliding windows are operated in the software SWAAP 1.0.2 [42] with windows and step sizes of 90 bp and 3 bp, respectively.
2.2. Phylogenetic tree construction 3. Results and discussion C9 gene sequences are used to construct the phylogenetic tree. All the gene sequences are aligned under codon model by MUSCLE software for its high accuracy and speed [31,32]. The protein domains of the amino acid sequences are characterized using the SMART program (http://smart.rembi-heidelberg.de/; [33]). The jModeltest software [34,35] is used to select the optimal substitution model before constructing the phylogenetic tree and the GTR þ I þ G model is regarded as the best-fit model by Bayesian information criterion (BIC). Phylogenetic trees are constructed with the Bayesian approach by MrBayes v3.2 [36] which is running 5,000,000 generation with 25% of trees burned. 2.3. Molecular evolutionary analysis Under neutrality, coding sequences are expected to present a ratio of non-synonymous substitutions (dN) over synonymous substitutions (dS). The u (dN/dS) ratio is a measure of natural selection acting on the protein. Simplistically, u ¼ 1, 1 are
3.1. The protein domains of C3 and C9 The protein domains of the amino acid sequences of C3 and C9 are characterized using the SMART program (http://smart.rembiheidelberg.de/; [33]). For C9 protein, it has different protein domains in aquatic and terrestrial animals. All fishes C9 proteins show a second TSP domain which is absent in mammalian counterparts (Fig. 1C and D). For C3 proteins, mammals and fishes share the same protein domains which include A2M_N, A2M_N_2, ANATO, A2M, Thiol-ester_cl, A2M_comp, A2M_recep and C345C (Fig. 1A and B). 3.2. Sliding window analysis along fish and mammal on C3 and C9 genes By the sliding window method, we separately calculate the values of u among fishes and mammals of C3 and C9 codons (Fig. 1). For C3 genes, we find that the values of u are larger than one at the
2042
S. Wang et al. / Fish & Shellfish Immunology 35 (2013) 2040e2045
Fig. 1. Sliding window analysis of variation in omega value along fish and mammal and the protein domains of C3 and C9. Windows and step sizes are 90 bp and 3 bp, respectively. A and B are C3 genes. Due to C3 protein, fish and mammal share the same protein domains. C and D are C9 genes. Fishes C9 proteins show a second TSP domain which is absent in mammalian counterparts.
A2M_N and ANATO domains in fishes. For C9 genes, the values of u which are larger than one occurred in MACPF domain for both mammals and fishes, also in TSP1 domain for fishes and EGF domain for mammals. Yang [37] pointed out that entire genes with u > 1 were generally rare, and that positive selection generally occurred in some structural domains. So we can speculate that positive selection sites may distribute in these domains. 3.3. Evolutionary process of C3 and C9 The Phylogenetic tree (Fig. 2) is constructed with thirty-five C9 genes which are supported with high posterior probabilities (pp ¼ 1.0) in most nodes. The tree of C9 genes is used to detect the evolution pattern in ancestral lineages of fishes and mammals for their huge different living environment and we estimated the ratio of u by a likelihood method. Firstly, the one-ratio model assumes that there has a unique u for all branches of the tree and the value of u is estimated to be 0.24 which is smaller than 1 (Table 1). The result reveals that the entire C9 genes undergo the purifying selection. Secondly, the free ratio-model allows different u values for each lineage among the tree compared with the one-ratio model, and the result shows that each branch has its own independent u value (p < 0.01, Table 1). Lastly, the branch-site models are used to detect whether positive selection sites in the ancestral lineages of mammals and fishes. Interestingly, the LRT tests based on the branch-site models for the ancestral lineages of mammals and
fishes suggests there is significant evidence of positive selection in ancestral lineages of fishes (2DlnL ¼ 13.70, p < 0.01, Table 1). As summarized in Table 1, we find three positive selection sites with high BPP of 0.96 in the ancestral lineages of fishes, but no positive selection sites are detected in ancestral lineages of mammals (Table 1). In addition, we use site model to detect the possible positive selection sites in the current C3 and C9 genes of fishes and mammals, respectively. The site models allow u to vary among codons and treat the u ratio for any site in the genes as a random variable from a statistical distribution [43,44]. Positive selection site occurs in some codons with u > 1 (in model M2a and M8). For mammals of C3 genes, no positive selection sites are detected by M2a model and M8 model (Table 2). But for fishes C3 genes, we detect six sites under positive selection with BPP values larger than 0.95 by M8 (304, 576, 831, 905, 971, 1408, Table 2). These results indicate that fishes and mammalian C3 genes experience different evolution pattern. To fishes C9 genes, we do not detect positive selection sites by M2a model and M8 model (Table 3). On the contrary, the BEB approach detects eight sites and fourteen sites under positive selection on mammals by M2a (204, 205, 206, 209, 210, 211, 212, 213, Table 3) and M8 (48, 160, 183, 204, 205, 206, 209, 210, 211, 212, 213, 339, 342, 356, Table 3), respectively. Besides, the LRT test statistic (2DlnL) of M1-M2a and M7-M8 comparison of mammal subset are 58.10 (p ¼ 0.00) and 72.67 (p ¼ 0.00). These results also indicate that there have different evolution pattern on C9 genes in mammals and fishes.
S. Wang et al. / Fish & Shellfish Immunology 35 (2013) 2040e2045
2043
Fig. 2. Phylogenetic tree of all C9 genes was constructed using MrBayes with Bayesian method. The GTR þ I þ G model was selected for the Bayesian analysis and the consensus tree was built after burning 125,000 trees from the 5,000,000 generations. Bayesian posterior probabilities values were indicated.
It is known that the vertebrate complement system can be activated through three overlapping pathways: the classical, alternative and lectin pathways [4,5]. C3 is the crossing of the three pathways in complement activation and acts as one of the bridges linking innate and acquired immunity [45e47]. In innate immunity, C3a and C3b, which are the activation products, have an essential role in phagocytosis, respiratory burst and inflammatory process [48]. In adaptive immunity, complement plays a fundamental role in mediating and enhancing humoral immunity. Due to complement activation, activated C3 becomes covalently bound to the surfaces of foreign microorganisms. Some of microorganisms are taken up by the complement receptors present on many antigenpresenting cells (APCs), including B lymphocytes [49,50]. These cells process antigens, then present them to T cells. They aid in mediating the antibody response to T cell-dependent antigens [8]. Lastly, the T cells produce antibody to kill the foreign microorganisms. But others are presented to MAC which can attach to the surface of membrane to create a pore in the cell membrane leading to cell lysis and death. Due to the crossing of three pathways of
complement activation, C3 becomes a target that microorganisms attack it. Besides, the aquatic environment contains countless kinds of bacteria and viruses comparing with terrestrial environment. So we detect six sites under positive selection on fishes C3 in our study (Table 2). For C3 genes, many fishes possess multiple forms of C3 which are the product of different genes and that differ in their structure and function [51,52]. The unique structural and functional diversity in C3 may improve its ability of recognizing a broader range of microorganisms, thereby allowing fish to expand their innate immune recognition capabilities [15]. Thus C3 can bind to more surfaces of foreign microorganisms and present them to antigen-presenting cells to process the antigen. Lastly, the T cells produce antibody to kill the foreign microorganisms. In terms of fishes C9, it may face little selective pressure on killing the foreign microorganisms. So we do not detect positive selection sites on fishes C9 genes. On the contrary, we do not find positive selection sites in mammalian C3 indicating that it faces fewer intense selective pressures than fishes. Besides, mammalian C3 has only one isoform which is encoded by one gene. Due to these reasons,
Table 1 Likelihood ratio tests of branch-models and branch-site models on C9 genes. Model Branch-model A: One-ratio B: Omega ¼ 1 C: Free-ratio Branch-site model 1: Null-fishes 2: fishes 3: Null-mammals 4: mammals a b c
Npa
Ln likelihood
Parameter estimates
70 69 137
12271.05 12752.86 12207.79
u ¼ 0.24 u ¼ 1.00
72 73 72 73
11889.35 11882.50 11891.32 11888.41
Variable u by branch
Numbers of parameters. Only the sites with BPP > 0.95 were shown. Twice the difference of ln [likelihood] between the two models compared.
Model compare
Positive selection sitesb
2DlnLc (p-value)
None B vs A C vs A
n/a
963.62 (p ¼ 0.00) 126.52 (p < 0.01)
1 vs 2
10,64,143
13.70 (p < 0.01)
3 vs 4
None
5.81 (p ¼ 0.02)
2044
S. Wang et al. / Fish & Shellfish Immunology 35 (2013) 2040e2045
Table 2 Site model tests on C3 genes in subsets of mammals and fishes. Model
Npa
Parameter estimates
Ln likelihood
Model compare
Positive selection sitesb
2DlnLc (p-value)
Data set: mammals M0:one-ratio M3: discrete
23 27
26235.51 25527.16
M3 vs M0
Not analyzed
1416.69 (p ¼ 0.00)
M1a: nearly neutral
24
M2a: positive selection
26
25527.62
M2 vs M1
None
0.00 (p ¼ 1.00)
M7: b M8: b and u
u ¼ 0.17 u0 ¼ 0.02, p0 ¼ 0.41, u1 ¼ 0.20, p1 ¼ 0.44, u2 ¼ 0.79, p2 ¼ 0.15 u0 ¼ 0.10, p0 ¼ 0.79, u1 ¼ 1.00, p1 ¼ 0.21 u0 ¼ 0.10, p0 ¼ 0.79, u1 ¼ 1.00, p1 ¼ 0.21, u2 ¼ 63.91, p2 ¼ 0.00
24 26
p ¼ 0.49, q ¼ 1.85 p0 ¼ 0.93, p1 ¼ 0.07, u ¼ 1.04, p ¼ 0.66, q ¼ 3.53
25539.95 25522.25
M8 vs M7
None
35.41 (p < 0.01)
Data set: fishes M0:one-ratio M3: discrete
37 41
46732.84 45362.15
M3 vs M0
Not analyzed
2741.39 (p ¼ 0.00)
M1a: nearly neutral
38
M2a: positive selection
40
45563.38
M2 vs M1
Not allowed
0.00 (p ¼ 1.00)
M7: b M8: b and u
u ¼ 0.31 u0 ¼ 0.05, p0 ¼ 0.39, u1 ¼ 0.36, p1 ¼ 0.44, u2 ¼ 1.16, p2 ¼ 0.17 u0 ¼ 0.15, p0 ¼ 0.67 u1 ¼ 1.00, p1 ¼ 0.33 u0 ¼ 0.15, p0 ¼ 0.67, u1 ¼ 1.00, p1 ¼ 0.04, u2 ¼ 1.00, p2 ¼ 0.29
38 40
45402.64 45342.50
M8 vs M7
304,576,831,905,971,1408
120.27 (p ¼ 0.00)
a b c
p ¼ 0.51, q ¼ 0.96 p0 ¼ 0.93, p1 ¼ 0.07 u ¼ 1.62, p ¼ 0.64, q ¼ 1.53
25527.62
45563.38
Numbers of parameters. Only the sites with BPP > 0.95 were shown. Twice the difference of ln [likelihood] between the two models compared.
mammalian C3 can not recognize a broader range of microorganisms and present more antigens to T cells comparing with fish, so that the mammalian C9 should face more pressures on killing the foreign microorganisms than fish. The result of site models on mammalian C9 also indicates that it faces intense selective
pressures. So we speculate that the relationship of C3 and C9 is complementary from evolutionary angle. In conclusion, different living environments lead to different evolutionary patterns on C3 and C9 in mammals and fishes. Besides, C3 and C9 experience different evolutionary patterns.
Table 3 Site model tests on C9 genes in subsets of mammals and fishes. Model compare
Positive selection sitesb
2DlnLc (p-value)
M3 vs M0
Not analyzed
1200.99 (p ¼ 0.00)
13557.92
M2 vs M1
204,205,206,209,210,211,212,213
58.10 (p ¼ 0.00)
13542.15 13505.82
M8 vs M7
48,160,183,204,205,206,209,210, 211,212,213,339,342,356
72.67 (p ¼ 0.00)
5180.27 5036.32
M3 vs M0
Not analyzed
287.91 (p ¼ 0.00)
5068.98
M2 vs M1
None
0.00 (p ¼ 1.00)
5038.38 5036.18
M8 vs M7
None
4.40 (p ¼ 0.11)
Model
Npa
Parameter estimates
Ln likelihood
Data set: mammals M0:one-ratio M3: discrete
49 53
14114.89 13514.40
M1a: nearly neutral
50
M2a: positive selection
52
M7: b M8: b and u
u ¼ 0.36 u0 ¼ 0.03, p0 ¼ 0.39, u1 ¼ 0.37, p1 ¼ 0.41, u2 ¼ 1.38, p2 ¼ 0.19 u0 ¼ 0.12, p0 ¼ 0.65, u1 ¼ 1.00, p1 ¼ 0.35 u0 ¼ 0.12, p0 ¼ 0.63, u1 ¼ 1.00, p1 ¼ 0.32, u2 ¼ 2.54, p2 ¼ 0.05
50 52
p ¼ 0.35, q ¼ 0.61 p0 ¼ 0.89, p1 ¼ 0.11, u ¼ 1.69, p ¼ 0.45,q ¼ 1.12
Data set: fishes M0:one-ratio M3: discrete
17 21
M1a: nearly neutral
18
M2a: positive selection
20
M7: b M8: b and u
u ¼ 0.22 u0 ¼ 0.03, p0 ¼ 0.40, u1 ¼ 0.30, p1 ¼ 0.45, u2 ¼ 0.98, p2 ¼ 0.14 u0 ¼ 0.12, p0 ¼ 0.68 u1 ¼ 1.00, p1 ¼ 0.32 u0 ¼ 0.12, p0 ¼ 0.68, u1 ¼ 1.00, p1 ¼ 0.23, u2 ¼ 1.00, p2 ¼ 0.00
18 20
a b c
p ¼ 0.50,q ¼ 1.26 p0 ¼ 0.99, p1 ¼ 0.01, u ¼ 5.94, p ¼ 0.51, q ¼ 1.36
13586.97
5068.98
Numbers of parameters. Only the sites with BPP > 0.95 were shown. Twice the difference of ln [likelihood] between the two models compared.
S. Wang et al. / Fish & Shellfish Immunology 35 (2013) 2040e2045
Acknowledgments This study was supported by National Nature Science Foundation of China (31370049) and Zhejiang Provincial Natural Science Foundation of China (LY12C03004). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.fsi.2013.10.018. References [1] Bordet J. Surle moded’action des serums preventifs. Ann Inst Pasteur 1986;10: 193e219. [2] Sunyer JO, Lambris JD. Complement. In: Encyclopedia of life sciences 1999. [3] Gasque P. Complement: a unique innate immune sensor for danger signals. Mol Immunol 2004;41:1089. [4] Nonaka M, Smith SL. Complement system of bony and cartilaginous fish. Fish Shellfish Immunol 2000;10:215e28. [5] Holland MCH, Lambris JD. The complement system in teleosts. Fish Shellfish Immunol 2002;12:399e420. [6] Kishore U, Reid KB. C1q: structure, function and receptors. Immunopharmacology 2000;49:159e70. [7] Petersen SV, Thiel S, Jensen L, Vorup-Jensen T, Koch C, Jensenius Jc. Control of the classical and the MBL pathway of complement activation. Mol Immunol 2000;37:803e11. [8] Carroll MC. The role of complement and complement receptors in induction and regulation of immunity. Annu Rev Immunol 1998;16:545e68. [9] Pangburn MK. Alternative pathway of complement. Methods Enzymol 1998;162:639e53. [10] Turner MW. The role of mannose-binding lectin in health and disease. Mol Immunol 2003;40:423. [11] Fujita T, Matsushita M, Endo Y. The lectin-complement pathwaydits role in innate immunity and evolution. Immunol Rev 2004;198:185. [12] Law SK, Dodds AW. The internal thioester and the covalent binding properties of the complement proteins C3 and C4. Protein Sci 1997;6:263e74. [13] Xu Y, Narayana SV, Volanakis JE. Structural biology of the alternative pathway convertase. Immunol Rev 2001;180:123e35. [14] Gros P, Milder FJ, Janssen BJ. Complement driven by conformational changes. Nat Rev Immunol 2008;8:48e58. [15] Sunyer JO, Zarkadis IK, Lambris JD. Complement diversity: a mechanism for generating immune diversity. Immunol Today 1998;19:519e23. [16] Dishaw LJ, Smith SL, Bigger CH. Characterization of a C3-like cDNA in a coral: phylogenetic implications. Immunogenetics 2005;57:535e48. [17] Zhu Y, Thangamani S, Ho B, Ding JL. The ancient origin of the complement system. Embol J 2005;24:382e94. [18] Fujito NT, Sugimoto S, Nonaka M. Evolution of thioester-containing proteins revealed by cloning and characterization of their genes from a cnidarian sea anemone, Haliplanella lineate. Dev Comp Immunol 2010;34:775e84. [19] Nonaka M, Yoshizaki F. Primitive complement system of invertebrates. Immunol Rev 2004;198:203e15. [20] Pinto MR, Melillo D, Giacomelli S, Sfyroera G, Lambris JD. Ancient origin of the complement system: emerging invertebrate models. Adv Exp Med Biol 2007;598:372e88. [21] Meng FX, Sun YN, Liu XZ, Wang JX, Xu TJ, Wang RX. Analysis of C3 suggests three periods of positive selection events and different evolutionary patterns between fish and mammals. PLoS ONE 2012;5:e37489. [22] Katagiri T, Hirono I, Aoki T. Molecular analysis of complement component C8beta and C9 cDNAs of Japanese flounder, Paralichthys olivaceus. Immunogenetics 1999;50:43e8. [23] Yeo GS, Elgar G, Sandford R, Brenner S. Cloning and sequencing of complement component C9 and its linkage to DOC-2 in the pufferfish Fugu rubripes. Gene 1997;200:203e11. [24] Uemura T, Yano T, Shiraishi H, Nakao M. Purification and characterization of the eighth and ninth components of carp complement. Mol Immunol 1996;33:925e32.
2045
[25] Chondrou MP, Londou AV, Zarkadis IK. Expression and phylogenetic analysis of the ninth complement component (C9) in rainbow trout. Fish Shellfish Immunol 2006;21:572e6. [26] Walport MJ. Complement. First of two parts. N Engl J Med 2001;344: 1058e66. [27] Volanakis JE. The complement system. Surv Immunol Res 1984;3:202e8. [28] Volanakis JE. Oveview of the complement system. In: Volanakis JE, Frank M, editors. The human complement system in health and disease 1998. New York. [29] Dankert JR, Esser AF. Bacterial killing by complement C9-mediated killing in the absence of C5b-8. Biochem J 1987;244:393e9. [30] Esser AF. Big MAC attack: complement proteins cause leaky patches. Immunol Today 1991;12:316e8. [31] Edgar RC. MUSCLE: a multiple sequence alignment method with reduced time and space complexity. BMC Bioinformatics 2004;5:113. [32] Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 2004;32:1792e7. [33] Letunic I, Copley RR, Pils B, Pinkert S, Schultz J, Bork P. SMART5: domains in the context of genomes and networks. Nucleic Acids Res 2006;34: 257e60. [34] Posada D. jModelTest: phylogenetic model averaging. Mol Biol Evol 2008;25: 1253e6. [35] Posada D. Selection of models of DNA evolution with jModelTest. Methods Mol Biol 2009;537:93e112. [36] Ronquist F, Huelsenbeck JP. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 2003;19:1572e4. [37] Yang Z. PAML 4: phylogenetic analysis by maximum likelihood. Mol Biol Evol 2007;24:1586e91. [38] Yang Z. PAML: a program package for phylogenetic analysis by maximum likelihood. Comput Appl Biosci 1997;13:555e6. [39] Yang Z, Nielsen R, Goldman N, Pedersen AM. Codon-substitution models for heterogeneous selection pressures at amino acid sites. Genetics 2000;155: 431e49. [40] Yang Z. Likelihood ratio tests for detecting positive selection and application to primate lysozyme evolution. Mol Biol Evol 1998;15:568e73. [41] Nei M, Gojobori T. Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol Biol Evol 1986;3: 418e26. [42] Pride DT. SWAAP 1.0.2: a tool for analyzing substitutions and similarity in multiple alignments; 2004. 1.0.2 ed. [43] Nielsen R, Yang Z. Likelihood models for detecting positively selected amino acid sites and applications to the HIV-1 envelope gene. Genetics 1998;148: 929e36. [44] Yang Z, Bielawski JP. Statistical methods for detecting molecular adaptation. Trends Ecol Evol 2000;15:496e503. [45] Nielsen CH, Lesile RG. Complement’s participation in acquired immunity. J Leukoc Biol 2002;72:249e61. [46] Carroll MC. The complement system in regulation of adaptive immunity. Nat Immunol 2004;5:981e6. [47] Morgan BP, Marchbank KJ, Longhi MP, Harris CL, Gallimore AM. Complement: central to innate immunity and briding to adaptive responses. Immunol Lett 2005;97:171e9. [48] Sahu A, Lambris JD. Structure and biology of complement protein C3, a connecting link between innate and acquired immunity. Immunol Rev 2001;180: 35e48. [49] Jacquier-Sarlin MR, Gabert FM, Villiers MB, Colomb MG. Modulation of antigen processing and presentation by covalently linked complement C3b fragment. Immunology 1995;84:164e70. [50] Villiers MB, Villiers CL, Jacquier-Sarlin MR, Gabert FM, Journet AM, Colomb MG. Covalent binding of C3b to tetanus toxin: influence on uptake/ internalization of antigen by antigen-specific and non-specific B cells. Immunology 1996;89:348e55. [51] Sunyer JO, Tort L, Lambris JD. Structural C3 diversity in fish: characterization of five forms of C3 in the diploid fish Sparus aurata. J Immunol 1997;158: 2813e21. [52] Nakao M, Mutsuro J, Obo R, Fujiki K, Nonaka M, Yano T. Molecular cloning and protein analysis of divergent forms of the complement component C3 from a bony fish, the common carp (Cyprinus carpino): presence of variants lacking the catalytic histidine. Eur J Immunol 2000;30:858e66.
Contents lists available at ScienceDirect
Fish & Shellfish Immunology journal homepage: www.elsevier.com/locate/fsi
Short communication
The evolutionary analysis on complement genes reveals that fishes C3 and C9 experience different evolutionary patterns Shanchen Wang, Rixin Wang, Tianjun Xu* Laboratory of Fish Biogenetics & Immune Evolution, College of Marine Science, Zhejiang Ocean University, Zhoushan, PR China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 August 2013 Received in revised form 14 October 2013 Accepted 21 October 2013 Available online 30 October 2013
Complement is a humoral factor of innate immunity and plays an essential role in altering the host of the presence of potential pathogens and clearing of invading microorganisms. The third complement component (C3) not only is regarded as the crossing of the three pathways of complement activation, but also serves one of the bridges linking innate and acquired immunity. The nine complement component (C9) can combine with C5b, C6, C7 and C8 to form MAC which bounds to the surface of microorganisms to kill them. The evidence of evolution on C3 genes which have multiple functions and plays central role in innate immunity was documented in our previous study. Now we were interested in the evolution of C9 genes which were the terminal complement components. For these reasons, we want to explore the evolutionary patterns of C9 and whether C3 and C9 experience different evolutionary patterns. In our study, we used the sliding window method to separately calculate the values of u among fishes and mammals of C3 and C9 codons. In order to detect the positive selection sites, we used the maximum likelihood (ML) method to study the evolutionary pattern on C3 and C9 genes. Positive selection sites were detected in mammalian C9 genes and no positive selection sites were detected in fishes C9 genes. However, no positive selection sites were detected in mammalian C3 genes and positive selection sites were detected in fishes C3 genes. The result indicated that C3 and C9 had different evolutionary patterns on mammals and fishes. In conclusion, different living environments lead to different evolutionary patterns on C3 and C9 in mammals and fishes. Besides, different complement components may have different evolutionary patterns on mammals and fishes. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Innate immunity Adaptive immunity C3 C9 Molecular evolution
1. Introduction The complement system was first described as a heat-sensitive factor in fresh serum that ‘complemented’ the effect of specific antibody in the lysis of bacteria and red blood cells [1]. Now, the complement system, which consists of more than thirty-five plasma and cell-surface proteins, is known to be a highly sophisticated host-defense system that plays key roles in innate and adaptive immunity [2,3]. Activation of the complement is initiated through any of three distinct pathways: the classical, alternative and lectin pathways [4,5]. The classical pathway is activated by immune complexes formed between antigens and antibodies [6,7]. The alternative pathway is directly activated by a large number of microorganisms [8,9]. The lectin pathway requires the interaction of lectins such as mannose-binding lectin (MBL) and ficolins, with sugar moieties based on the surface of microorganism [10,11].
* Corresponding author. Tel./fax: þ86 580 2550826. E-mail address: [email protected] (T. Xu). 1050-4648/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fsi.2013.10.018
Activation of the complement system through any the three pathways results in the activation of C3. The first complement component, C1, is Ca2þ dependent protein complex that consists of C1r and C1s bound to one molecule of C1q. When surface-bound IgM or IgG binds to C1q, the classical pathway is activated. Activated C1s cleaves C4 and C2 into C4a, C4b, C2a and C2b, respectively. The C4bC2a enzymatic complex is known as C3 convertase which cleaves C3 into C3a and C3b [12e14]. C3 belongs to A2M alpha-2-macroglobulin superfamily and plays a central role in complement activation which emerged over 700 million years ago [15]. C3b can become covalently bound to the cell surface. Once bound, C3b can form a complex with C4bC2a to give rise to C5 convertase. Then C5 is cleaved into C5a and C5b. C5a plays an important role in the inflammation process and C5b initiates the self-assembly of the MAC by subsequent binding of C6, C7, C8 and C9. MAC can create a pore in the cell membrane leading to cell lysis and death. With the deep study, complement molecules have been found and studied in a variety of organisms, principally in vertebrates. The homologs of C3 have been found in invertebrates [16e18]. These findings proved that the origin and evolution of the complement
S. Wang et al. / Fish & Shellfish Immunology 35 (2013) 2040e2045
system are traced to the earliest radiations of the animal kingdom [19,20]. Now the study on complement system is on evolutionary process rather than molecular structure and cloning. In our previous study, we revolve around the complement component to research. The evidence of evolution on C3 genes was reported. We found that three periods of positive selection events had happened on C3 genes during evolutionary history and C3 genes experienced different evolutionary patterns on fishes and mammals for their living environments [21]. Now, the evolution of C9 genes which are the terminal complement components assembling with C5b, C6, C7 and C8 to form MAC is attracted us. Besides, many fishes C9 genes were cloned and studied, for example, Paralichthys olivaceus [22], Fugu rubripes [23], Cyprinus carpio [24], Oncorhynchus mykiss [25] and so on. So we want to explore the evolutionary patterns and evolutionary mechanism of C9 genes. Besides, it is known that C3 not only is the crossing of the three pathways of complement activation, but also serves as one of the bridges linking innate and acquired immunity [26]. The activated C3 is able to covalently bind to foreign microorganisms, which are presented to the T cells to result in cellular immune response, or further lysed by the MAC. However, C9 is a single-chain glycoprotein that is involved in the formation of the MAC on the surface of target cells [27,28]. The MAC is assembled by C5b, C6, C7, C8 and C9, once assembled on the target surface it forms transmembrane channels resulting in cell lysis [29,30]. Thus in the history of the evolution, the question on the relationship of C3 and C9 was attracted us. So we are interested in the evolutionary patterns on C3 and C9. In our study, we used the maximum likelihood (ML) method to study the evolution pattern on C3 and C9, respectively. We want to explore the evolution pattern on C9 genes and test whether C3 and C9 have difference on evolution. 2. Materials and methods 2.1. Taxonomic coverage We derived nine C9 sequences from nine fish species, twenty-five C9 sequences from twenty-five mammals, one C9 sequence from amphibian; nineteen C3 sequences from ten fishes, twelve C3 sequences from mammals, together with two C3 sequences from invertebrates and one from bird (Table S1). All the C9 and C3 sequences in our study were derived from GenBank (http://www.ncbi.nlm.nih. gov/Genbank/) and Ensemble (http://www.ensemble.org/) database for evolutionary analyses.
2041
indicative of neutral evolution, purifying selection and positive selection. The phylogenetic tree is used to detect whether different environment affects the evolution pattern on aquatic and terrestrial organisms in C9. In order to investigate the evolutionary process of C3 and C9 genes in mammals and fishes, PAML 4 [37] is used to estimate the ratio of u. The PAML package includes a free-ratio model, site-specific model, a branch-specific model (two-ratio) and a branch-site model (model A) [38]. The free-ratio model is used to allow the u ratio to vary on each branch which is used to test whether this model fitted the data significantly better than the one ratio model which is used to identify the selective pressures in all genes by the likelihood ratio test (LRT). Branch-site model was used to detect positive selection on the ancestral foreground lineages. The site model which includes six models of codons substitution is applied to subset of fishes and mammals C3 and C9 sequences to detect the possible positive selection sites among these lineages. The following codons substitution models were utilized: M0 (one ratio) with assuming a unique value of u across all sites, M3 (discrete) which has three discrete classes of site with different u values for sites in the phylogeny, M1a (nearly neutral model) with two site classes (0 < u < 1, u ¼ 1) estimating the frequency p0 and p1, M2a (positive selection) which added a class of positively selected sites comparing with M1a, M7(b) which do not allow the positive selection with assuming the frequencies of different u values (0 < u < 1) to follow a beta distribution whose shape is controlled by two parameters (p and q), M8 (b & u) which is similar to M7 by adding a proportion p1 with u > 1 [34]. In all cases, significant differences between models can be evaluated by calculating twice the difference of the log-likelihood values (2DlnL) following a chi-squared distribution with the number of degrees of freedom equal to the difference in the number of free parameters between models [39]. The Bayes empirical Bayes (BEB) is used to calculate the Bayesian posterior probability (BPP) of the codons sites under a positive selection [40]. In order to improve the accuracy of results, sites with BPP < 95% is regarded as positive selection sites. 2.4. Sliding windows analysis To explore further the distribution of positive selection sites on specific functional domains of C3 and C9 between fishes and mammals, we use the Nei and Gojobori method [41] to conduct the u values of sliding windows. Sliding windows are operated in the software SWAAP 1.0.2 [42] with windows and step sizes of 90 bp and 3 bp, respectively.
2.2. Phylogenetic tree construction 3. Results and discussion C9 gene sequences are used to construct the phylogenetic tree. All the gene sequences are aligned under codon model by MUSCLE software for its high accuracy and speed [31,32]. The protein domains of the amino acid sequences are characterized using the SMART program (http://smart.rembi-heidelberg.de/; [33]). The jModeltest software [34,35] is used to select the optimal substitution model before constructing the phylogenetic tree and the GTR þ I þ G model is regarded as the best-fit model by Bayesian information criterion (BIC). Phylogenetic trees are constructed with the Bayesian approach by MrBayes v3.2 [36] which is running 5,000,000 generation with 25% of trees burned. 2.3. Molecular evolutionary analysis Under neutrality, coding sequences are expected to present a ratio of non-synonymous substitutions (dN) over synonymous substitutions (dS). The u (dN/dS) ratio is a measure of natural selection acting on the protein. Simplistically, u ¼ 1, 1 are
3.1. The protein domains of C3 and C9 The protein domains of the amino acid sequences of C3 and C9 are characterized using the SMART program (http://smart.rembiheidelberg.de/; [33]). For C9 protein, it has different protein domains in aquatic and terrestrial animals. All fishes C9 proteins show a second TSP domain which is absent in mammalian counterparts (Fig. 1C and D). For C3 proteins, mammals and fishes share the same protein domains which include A2M_N, A2M_N_2, ANATO, A2M, Thiol-ester_cl, A2M_comp, A2M_recep and C345C (Fig. 1A and B). 3.2. Sliding window analysis along fish and mammal on C3 and C9 genes By the sliding window method, we separately calculate the values of u among fishes and mammals of C3 and C9 codons (Fig. 1). For C3 genes, we find that the values of u are larger than one at the
2042
S. Wang et al. / Fish & Shellfish Immunology 35 (2013) 2040e2045
Fig. 1. Sliding window analysis of variation in omega value along fish and mammal and the protein domains of C3 and C9. Windows and step sizes are 90 bp and 3 bp, respectively. A and B are C3 genes. Due to C3 protein, fish and mammal share the same protein domains. C and D are C9 genes. Fishes C9 proteins show a second TSP domain which is absent in mammalian counterparts.
A2M_N and ANATO domains in fishes. For C9 genes, the values of u which are larger than one occurred in MACPF domain for both mammals and fishes, also in TSP1 domain for fishes and EGF domain for mammals. Yang [37] pointed out that entire genes with u > 1 were generally rare, and that positive selection generally occurred in some structural domains. So we can speculate that positive selection sites may distribute in these domains. 3.3. Evolutionary process of C3 and C9 The Phylogenetic tree (Fig. 2) is constructed with thirty-five C9 genes which are supported with high posterior probabilities (pp ¼ 1.0) in most nodes. The tree of C9 genes is used to detect the evolution pattern in ancestral lineages of fishes and mammals for their huge different living environment and we estimated the ratio of u by a likelihood method. Firstly, the one-ratio model assumes that there has a unique u for all branches of the tree and the value of u is estimated to be 0.24 which is smaller than 1 (Table 1). The result reveals that the entire C9 genes undergo the purifying selection. Secondly, the free ratio-model allows different u values for each lineage among the tree compared with the one-ratio model, and the result shows that each branch has its own independent u value (p < 0.01, Table 1). Lastly, the branch-site models are used to detect whether positive selection sites in the ancestral lineages of mammals and fishes. Interestingly, the LRT tests based on the branch-site models for the ancestral lineages of mammals and
fishes suggests there is significant evidence of positive selection in ancestral lineages of fishes (2DlnL ¼ 13.70, p < 0.01, Table 1). As summarized in Table 1, we find three positive selection sites with high BPP of 0.96 in the ancestral lineages of fishes, but no positive selection sites are detected in ancestral lineages of mammals (Table 1). In addition, we use site model to detect the possible positive selection sites in the current C3 and C9 genes of fishes and mammals, respectively. The site models allow u to vary among codons and treat the u ratio for any site in the genes as a random variable from a statistical distribution [43,44]. Positive selection site occurs in some codons with u > 1 (in model M2a and M8). For mammals of C3 genes, no positive selection sites are detected by M2a model and M8 model (Table 2). But for fishes C3 genes, we detect six sites under positive selection with BPP values larger than 0.95 by M8 (304, 576, 831, 905, 971, 1408, Table 2). These results indicate that fishes and mammalian C3 genes experience different evolution pattern. To fishes C9 genes, we do not detect positive selection sites by M2a model and M8 model (Table 3). On the contrary, the BEB approach detects eight sites and fourteen sites under positive selection on mammals by M2a (204, 205, 206, 209, 210, 211, 212, 213, Table 3) and M8 (48, 160, 183, 204, 205, 206, 209, 210, 211, 212, 213, 339, 342, 356, Table 3), respectively. Besides, the LRT test statistic (2DlnL) of M1-M2a and M7-M8 comparison of mammal subset are 58.10 (p ¼ 0.00) and 72.67 (p ¼ 0.00). These results also indicate that there have different evolution pattern on C9 genes in mammals and fishes.
S. Wang et al. / Fish & Shellfish Immunology 35 (2013) 2040e2045
2043
Fig. 2. Phylogenetic tree of all C9 genes was constructed using MrBayes with Bayesian method. The GTR þ I þ G model was selected for the Bayesian analysis and the consensus tree was built after burning 125,000 trees from the 5,000,000 generations. Bayesian posterior probabilities values were indicated.
It is known that the vertebrate complement system can be activated through three overlapping pathways: the classical, alternative and lectin pathways [4,5]. C3 is the crossing of the three pathways in complement activation and acts as one of the bridges linking innate and acquired immunity [45e47]. In innate immunity, C3a and C3b, which are the activation products, have an essential role in phagocytosis, respiratory burst and inflammatory process [48]. In adaptive immunity, complement plays a fundamental role in mediating and enhancing humoral immunity. Due to complement activation, activated C3 becomes covalently bound to the surfaces of foreign microorganisms. Some of microorganisms are taken up by the complement receptors present on many antigenpresenting cells (APCs), including B lymphocytes [49,50]. These cells process antigens, then present them to T cells. They aid in mediating the antibody response to T cell-dependent antigens [8]. Lastly, the T cells produce antibody to kill the foreign microorganisms. But others are presented to MAC which can attach to the surface of membrane to create a pore in the cell membrane leading to cell lysis and death. Due to the crossing of three pathways of
complement activation, C3 becomes a target that microorganisms attack it. Besides, the aquatic environment contains countless kinds of bacteria and viruses comparing with terrestrial environment. So we detect six sites under positive selection on fishes C3 in our study (Table 2). For C3 genes, many fishes possess multiple forms of C3 which are the product of different genes and that differ in their structure and function [51,52]. The unique structural and functional diversity in C3 may improve its ability of recognizing a broader range of microorganisms, thereby allowing fish to expand their innate immune recognition capabilities [15]. Thus C3 can bind to more surfaces of foreign microorganisms and present them to antigen-presenting cells to process the antigen. Lastly, the T cells produce antibody to kill the foreign microorganisms. In terms of fishes C9, it may face little selective pressure on killing the foreign microorganisms. So we do not detect positive selection sites on fishes C9 genes. On the contrary, we do not find positive selection sites in mammalian C3 indicating that it faces fewer intense selective pressures than fishes. Besides, mammalian C3 has only one isoform which is encoded by one gene. Due to these reasons,
Table 1 Likelihood ratio tests of branch-models and branch-site models on C9 genes. Model Branch-model A: One-ratio B: Omega ¼ 1 C: Free-ratio Branch-site model 1: Null-fishes 2: fishes 3: Null-mammals 4: mammals a b c
Npa
Ln likelihood
Parameter estimates
70 69 137
12271.05 12752.86 12207.79
u ¼ 0.24 u ¼ 1.00
72 73 72 73
11889.35 11882.50 11891.32 11888.41
Variable u by branch
Numbers of parameters. Only the sites with BPP > 0.95 were shown. Twice the difference of ln [likelihood] between the two models compared.
Model compare
Positive selection sitesb
2DlnLc (p-value)
None B vs A C vs A
n/a
963.62 (p ¼ 0.00) 126.52 (p < 0.01)
1 vs 2
10,64,143
13.70 (p < 0.01)
3 vs 4
None
5.81 (p ¼ 0.02)
2044
S. Wang et al. / Fish & Shellfish Immunology 35 (2013) 2040e2045
Table 2 Site model tests on C3 genes in subsets of mammals and fishes. Model
Npa
Parameter estimates
Ln likelihood
Model compare
Positive selection sitesb
2DlnLc (p-value)
Data set: mammals M0:one-ratio M3: discrete
23 27
26235.51 25527.16
M3 vs M0
Not analyzed
1416.69 (p ¼ 0.00)
M1a: nearly neutral
24
M2a: positive selection
26
25527.62
M2 vs M1
None
0.00 (p ¼ 1.00)
M7: b M8: b and u
u ¼ 0.17 u0 ¼ 0.02, p0 ¼ 0.41, u1 ¼ 0.20, p1 ¼ 0.44, u2 ¼ 0.79, p2 ¼ 0.15 u0 ¼ 0.10, p0 ¼ 0.79, u1 ¼ 1.00, p1 ¼ 0.21 u0 ¼ 0.10, p0 ¼ 0.79, u1 ¼ 1.00, p1 ¼ 0.21, u2 ¼ 63.91, p2 ¼ 0.00
24 26
p ¼ 0.49, q ¼ 1.85 p0 ¼ 0.93, p1 ¼ 0.07, u ¼ 1.04, p ¼ 0.66, q ¼ 3.53
25539.95 25522.25
M8 vs M7
None
35.41 (p < 0.01)
Data set: fishes M0:one-ratio M3: discrete
37 41
46732.84 45362.15
M3 vs M0
Not analyzed
2741.39 (p ¼ 0.00)
M1a: nearly neutral
38
M2a: positive selection
40
45563.38
M2 vs M1
Not allowed
0.00 (p ¼ 1.00)
M7: b M8: b and u
u ¼ 0.31 u0 ¼ 0.05, p0 ¼ 0.39, u1 ¼ 0.36, p1 ¼ 0.44, u2 ¼ 1.16, p2 ¼ 0.17 u0 ¼ 0.15, p0 ¼ 0.67 u1 ¼ 1.00, p1 ¼ 0.33 u0 ¼ 0.15, p0 ¼ 0.67, u1 ¼ 1.00, p1 ¼ 0.04, u2 ¼ 1.00, p2 ¼ 0.29
38 40
45402.64 45342.50
M8 vs M7
304,576,831,905,971,1408
120.27 (p ¼ 0.00)
a b c
p ¼ 0.51, q ¼ 0.96 p0 ¼ 0.93, p1 ¼ 0.07 u ¼ 1.62, p ¼ 0.64, q ¼ 1.53
25527.62
45563.38
Numbers of parameters. Only the sites with BPP > 0.95 were shown. Twice the difference of ln [likelihood] between the two models compared.
mammalian C3 can not recognize a broader range of microorganisms and present more antigens to T cells comparing with fish, so that the mammalian C9 should face more pressures on killing the foreign microorganisms than fish. The result of site models on mammalian C9 also indicates that it faces intense selective
pressures. So we speculate that the relationship of C3 and C9 is complementary from evolutionary angle. In conclusion, different living environments lead to different evolutionary patterns on C3 and C9 in mammals and fishes. Besides, C3 and C9 experience different evolutionary patterns.
Table 3 Site model tests on C9 genes in subsets of mammals and fishes. Model compare
Positive selection sitesb
2DlnLc (p-value)
M3 vs M0
Not analyzed
1200.99 (p ¼ 0.00)
13557.92
M2 vs M1
204,205,206,209,210,211,212,213
58.10 (p ¼ 0.00)
13542.15 13505.82
M8 vs M7
48,160,183,204,205,206,209,210, 211,212,213,339,342,356
72.67 (p ¼ 0.00)
5180.27 5036.32
M3 vs M0
Not analyzed
287.91 (p ¼ 0.00)
5068.98
M2 vs M1
None
0.00 (p ¼ 1.00)
5038.38 5036.18
M8 vs M7
None
4.40 (p ¼ 0.11)
Model
Npa
Parameter estimates
Ln likelihood
Data set: mammals M0:one-ratio M3: discrete
49 53
14114.89 13514.40
M1a: nearly neutral
50
M2a: positive selection
52
M7: b M8: b and u
u ¼ 0.36 u0 ¼ 0.03, p0 ¼ 0.39, u1 ¼ 0.37, p1 ¼ 0.41, u2 ¼ 1.38, p2 ¼ 0.19 u0 ¼ 0.12, p0 ¼ 0.65, u1 ¼ 1.00, p1 ¼ 0.35 u0 ¼ 0.12, p0 ¼ 0.63, u1 ¼ 1.00, p1 ¼ 0.32, u2 ¼ 2.54, p2 ¼ 0.05
50 52
p ¼ 0.35, q ¼ 0.61 p0 ¼ 0.89, p1 ¼ 0.11, u ¼ 1.69, p ¼ 0.45,q ¼ 1.12
Data set: fishes M0:one-ratio M3: discrete
17 21
M1a: nearly neutral
18
M2a: positive selection
20
M7: b M8: b and u
u ¼ 0.22 u0 ¼ 0.03, p0 ¼ 0.40, u1 ¼ 0.30, p1 ¼ 0.45, u2 ¼ 0.98, p2 ¼ 0.14 u0 ¼ 0.12, p0 ¼ 0.68 u1 ¼ 1.00, p1 ¼ 0.32 u0 ¼ 0.12, p0 ¼ 0.68, u1 ¼ 1.00, p1 ¼ 0.23, u2 ¼ 1.00, p2 ¼ 0.00
18 20
a b c
p ¼ 0.50,q ¼ 1.26 p0 ¼ 0.99, p1 ¼ 0.01, u ¼ 5.94, p ¼ 0.51, q ¼ 1.36
13586.97
5068.98
Numbers of parameters. Only the sites with BPP > 0.95 were shown. Twice the difference of ln [likelihood] between the two models compared.
S. Wang et al. / Fish & Shellfish Immunology 35 (2013) 2040e2045
Acknowledgments This study was supported by National Nature Science Foundation of China (31370049) and Zhejiang Provincial Natural Science Foundation of China (LY12C03004). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.fsi.2013.10.018. References [1] Bordet J. Surle moded’action des serums preventifs. Ann Inst Pasteur 1986;10: 193e219. [2] Sunyer JO, Lambris JD. Complement. In: Encyclopedia of life sciences 1999. [3] Gasque P. Complement: a unique innate immune sensor for danger signals. Mol Immunol 2004;41:1089. [4] Nonaka M, Smith SL. Complement system of bony and cartilaginous fish. Fish Shellfish Immunol 2000;10:215e28. [5] Holland MCH, Lambris JD. The complement system in teleosts. Fish Shellfish Immunol 2002;12:399e420. [6] Kishore U, Reid KB. C1q: structure, function and receptors. Immunopharmacology 2000;49:159e70. [7] Petersen SV, Thiel S, Jensen L, Vorup-Jensen T, Koch C, Jensenius Jc. Control of the classical and the MBL pathway of complement activation. Mol Immunol 2000;37:803e11. [8] Carroll MC. The role of complement and complement receptors in induction and regulation of immunity. Annu Rev Immunol 1998;16:545e68. [9] Pangburn MK. Alternative pathway of complement. Methods Enzymol 1998;162:639e53. [10] Turner MW. The role of mannose-binding lectin in health and disease. Mol Immunol 2003;40:423. [11] Fujita T, Matsushita M, Endo Y. The lectin-complement pathwaydits role in innate immunity and evolution. Immunol Rev 2004;198:185. [12] Law SK, Dodds AW. The internal thioester and the covalent binding properties of the complement proteins C3 and C4. Protein Sci 1997;6:263e74. [13] Xu Y, Narayana SV, Volanakis JE. Structural biology of the alternative pathway convertase. Immunol Rev 2001;180:123e35. [14] Gros P, Milder FJ, Janssen BJ. Complement driven by conformational changes. Nat Rev Immunol 2008;8:48e58. [15] Sunyer JO, Zarkadis IK, Lambris JD. Complement diversity: a mechanism for generating immune diversity. Immunol Today 1998;19:519e23. [16] Dishaw LJ, Smith SL, Bigger CH. Characterization of a C3-like cDNA in a coral: phylogenetic implications. Immunogenetics 2005;57:535e48. [17] Zhu Y, Thangamani S, Ho B, Ding JL. The ancient origin of the complement system. Embol J 2005;24:382e94. [18] Fujito NT, Sugimoto S, Nonaka M. Evolution of thioester-containing proteins revealed by cloning and characterization of their genes from a cnidarian sea anemone, Haliplanella lineate. Dev Comp Immunol 2010;34:775e84. [19] Nonaka M, Yoshizaki F. Primitive complement system of invertebrates. Immunol Rev 2004;198:203e15. [20] Pinto MR, Melillo D, Giacomelli S, Sfyroera G, Lambris JD. Ancient origin of the complement system: emerging invertebrate models. Adv Exp Med Biol 2007;598:372e88. [21] Meng FX, Sun YN, Liu XZ, Wang JX, Xu TJ, Wang RX. Analysis of C3 suggests three periods of positive selection events and different evolutionary patterns between fish and mammals. PLoS ONE 2012;5:e37489. [22] Katagiri T, Hirono I, Aoki T. Molecular analysis of complement component C8beta and C9 cDNAs of Japanese flounder, Paralichthys olivaceus. Immunogenetics 1999;50:43e8. [23] Yeo GS, Elgar G, Sandford R, Brenner S. Cloning and sequencing of complement component C9 and its linkage to DOC-2 in the pufferfish Fugu rubripes. Gene 1997;200:203e11. [24] Uemura T, Yano T, Shiraishi H, Nakao M. Purification and characterization of the eighth and ninth components of carp complement. Mol Immunol 1996;33:925e32.
2045
[25] Chondrou MP, Londou AV, Zarkadis IK. Expression and phylogenetic analysis of the ninth complement component (C9) in rainbow trout. Fish Shellfish Immunol 2006;21:572e6. [26] Walport MJ. Complement. First of two parts. N Engl J Med 2001;344: 1058e66. [27] Volanakis JE. The complement system. Surv Immunol Res 1984;3:202e8. [28] Volanakis JE. Oveview of the complement system. In: Volanakis JE, Frank M, editors. The human complement system in health and disease 1998. New York. [29] Dankert JR, Esser AF. Bacterial killing by complement C9-mediated killing in the absence of C5b-8. Biochem J 1987;244:393e9. [30] Esser AF. Big MAC attack: complement proteins cause leaky patches. Immunol Today 1991;12:316e8. [31] Edgar RC. MUSCLE: a multiple sequence alignment method with reduced time and space complexity. BMC Bioinformatics 2004;5:113. [32] Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 2004;32:1792e7. [33] Letunic I, Copley RR, Pils B, Pinkert S, Schultz J, Bork P. SMART5: domains in the context of genomes and networks. Nucleic Acids Res 2006;34: 257e60. [34] Posada D. jModelTest: phylogenetic model averaging. Mol Biol Evol 2008;25: 1253e6. [35] Posada D. Selection of models of DNA evolution with jModelTest. Methods Mol Biol 2009;537:93e112. [36] Ronquist F, Huelsenbeck JP. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 2003;19:1572e4. [37] Yang Z. PAML 4: phylogenetic analysis by maximum likelihood. Mol Biol Evol 2007;24:1586e91. [38] Yang Z. PAML: a program package for phylogenetic analysis by maximum likelihood. Comput Appl Biosci 1997;13:555e6. [39] Yang Z, Nielsen R, Goldman N, Pedersen AM. Codon-substitution models for heterogeneous selection pressures at amino acid sites. Genetics 2000;155: 431e49. [40] Yang Z. Likelihood ratio tests for detecting positive selection and application to primate lysozyme evolution. Mol Biol Evol 1998;15:568e73. [41] Nei M, Gojobori T. Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol Biol Evol 1986;3: 418e26. [42] Pride DT. SWAAP 1.0.2: a tool for analyzing substitutions and similarity in multiple alignments; 2004. 1.0.2 ed. [43] Nielsen R, Yang Z. Likelihood models for detecting positively selected amino acid sites and applications to the HIV-1 envelope gene. Genetics 1998;148: 929e36. [44] Yang Z, Bielawski JP. Statistical methods for detecting molecular adaptation. Trends Ecol Evol 2000;15:496e503. [45] Nielsen CH, Lesile RG. Complement’s participation in acquired immunity. J Leukoc Biol 2002;72:249e61. [46] Carroll MC. The complement system in regulation of adaptive immunity. Nat Immunol 2004;5:981e6. [47] Morgan BP, Marchbank KJ, Longhi MP, Harris CL, Gallimore AM. Complement: central to innate immunity and briding to adaptive responses. Immunol Lett 2005;97:171e9. [48] Sahu A, Lambris JD. Structure and biology of complement protein C3, a connecting link between innate and acquired immunity. Immunol Rev 2001;180: 35e48. [49] Jacquier-Sarlin MR, Gabert FM, Villiers MB, Colomb MG. Modulation of antigen processing and presentation by covalently linked complement C3b fragment. Immunology 1995;84:164e70. [50] Villiers MB, Villiers CL, Jacquier-Sarlin MR, Gabert FM, Journet AM, Colomb MG. Covalent binding of C3b to tetanus toxin: influence on uptake/ internalization of antigen by antigen-specific and non-specific B cells. Immunology 1996;89:348e55. [51] Sunyer JO, Tort L, Lambris JD. Structural C3 diversity in fish: characterization of five forms of C3 in the diploid fish Sparus aurata. J Immunol 1997;158: 2813e21. [52] Nakao M, Mutsuro J, Obo R, Fujiki K, Nonaka M, Yano T. Molecular cloning and protein analysis of divergent forms of the complement component C3 from a bony fish, the common carp (Cyprinus carpino): presence of variants lacking the catalytic histidine. Eur J Immunol 2000;30:858e66.
