View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Archivio istituzionale della ricerca - Università di Padova

Invited Review ACTA ZOOLOGICA BULGARICA Acta zool. bulg., 67 (2), 2015: 305-317

New Insights into the Molecular Evolution of Metazoan

Ri g e r s Ba k i u 1*, Gi a n f r a n c o Sa n t o v i t o 2

1 Department of and Fisheries, Agricultural University of Tirana, Koder Kamez, 1029 Tirana, Albania; E-mail: [email protected] 2 Department of Biology, University of Padova, 35121 Padova, Italy

Abstract: Peroxiredoxins (Prx) are enzymes present in all biological kingdoms, from bacteria to . The oxi- dised active site cysteine of Prx can be reduced by a cellular thiol, thus enabling Prx to function as a . Peroxiredoxins have been object of an increasing interest for its pivotal role in cell defence and as conserved markers for circadian rhythms in metabolism across all three phylogenetic domains (Eukarya, Bacteria and Archaea). Metazoan cells express six Prx isoforms that are localised in various cellular compartments. Using bioinformatics tools, based on Bayesian approach, we analysed the phylo- genetic relationships among metazoan Prxs, with the aim to acquire new data on the molecular evolution of these . molecular evolution analyses were performed by the application of Mr. Bayes and HyPhy software to the coding and sequences of deuterostomes and protostomes. The obtained results confirmed that the molecular evolution of metazoan Prx was peculiar and suggested that the positive selection may had operated for the evolution of these proteins and a purifying selection was present during this process.

Keywords: peroxiredoxin; metazoans; molecular evolution.

Introduction

An unusual antioxidant protein, now called perox- is specifically oxidised by H2O2 to cysteine sulfenic iredoxin (Prx, EC 1.11.1.15), was initially identified acid (Cys–SOH), which is resolved by reaction with on the basis of its capacity to protect proteins from Cys170–SH of the adjacent monomer. Theseis results oxidative damage caused by reactive oxygen in the formation of a disulfide link, Cys47–S-S–Cys170

(ROS), catalysing the reduction of H2O2, to water (Ch a e et al. 1994). The conserved Cys residue cor- and alcohol in the presence of dithiotreitol (DTT). responding to Cys47 of yeast Prx was later referred

Analysis of Prx purified from yeast revealed that to as the peroxidatic Cys (CP) to reflect its sensitiv- it did not contain conventional redox centres such ity to oxidation by peroxides, and the conserved Cys as metals, heme, flavin, or selenocysteine, being residue corresponding to Cys170 was designated the o o d very different with respect to other known antioxi- resolving Cys (CR) (W et al. 2003). dant (Ki m et al. 1988; Ch a e et al. 1994). Later, it On the basis of the presence or absence of the

has been found that (i) Prxs are present in all bio- CR residue, Prxs are classified into typical 2-Cys, logical kingdoms, from bacteria to animals; (ii) two atypical 2-Cys, and 1-Cys Prx subfamilies (Rh e e cysteine residues, corresponding to Cys47 and Cys170 & Wo o , 2011). cells express six isoforms of yeast Prx, are highly conserved among Prx fam- of Prxs: isoforms from 1 to 4 belong to the typical ily members; (iii) Prxs are homodimers arranged in 2-Cys Prx group; Prx5 belongs to the class of 2-Cys a head-to-tail orientation; and (iv) Cys47–SH of Prx enzymes. Isoform 6 is the only one belonging to

*Corresponding author

305 Bakiu R., G. Santovito

1-Cys Prxs (Fisher, 2011). Prx1 is mainly localised Metazoa is one of great eukaryote kingdoms. in the cytosol, nucleus and peroxisomes, but it has An increasingly well resolved Proterozoic fossil been found also in serum (Im m e n s c h u h et al. 2003). record documents the presence of most of the major Prx2 is present in cytosol and nucleus and it has been taxa of eukaryotes, including the rhodophytes, stra- shown to bond the cell membrane (Ch a et al. 2000). menopiles, alveolates and green plants during the late Prx3 is located exclusively in the mitochondria (Ca o Mesoproterozoic to early Neoproterozoic (about one et al. 2007). Prx4 has been found in both cytosol and billion years ago). A coincident rise in acritarch diver- endoplasmic reticulum and contains a leader peptide sity, combined with molecular phylogeny evidence that is believed to be essential for protein secretion for rapid cladogenesis, points to a major radiation of (Ok a d o -Ma t s u m o t o et al. 2000). Prx5 is localised in eukaryote groups at this time, sometimes referred to cytosol, mitochondria and peroxisomes (Ca o et al. as the “big bang” of eukaryotic evolution (Co n w a y 2007). Prx6 is located in cytosol, vesicles and lyso- Mo rr i s et al. 1987). The most fundamental division somes (So r o k i n a et al. 2011). Prx localisation is mul- within the bilateral metazoans is the protostome- tifarious, being dependent on the cell type but also on deuterostome branch. Protostomes include at least the environmental conditions (Rh e e et al. 2012). annelids, arthropods, molluscs and platyhelminths. Many studies suggest that Prxs are more than They have spiral cleavage and usually mosaic de- just simple antioxidant enzymes. For example Prx6 velopment, are schizocelic, form the mouth at (or also acts as phospholipase A2 (Ch e n et al. 2000). near) the site of the blastopore and mesoderm from Other evidences suggest that Prx oxidation also al- a mesentoblast that is usually 4d. Deuterostomes in- lows them to function as molecular chaperones (Ja n g clude at least echinoderms and , the latter et al. 2004) and regulates the cell cycle (Ph a l e n et group including the back-boned animals. They have al. 2006). A very interesting proposed role is en- radial cleavage and usually regulative development, compassed by the floodgate hypothesis (Wo o d et al. are enterocelic, form the mouth away from the blas- 2003; Wo o et al. 2010), in which active Prxs normal- topore and mesoderm from endodermal cells along a l e n t i n e ly keep H2O2 low (i.e. a closed floodgate) but, under the archeneteron (V et al. 1997). signalling conditions that causes loss of function via Approximately 2.5 billion years ago, photo- overoxidation in a localised region of the cell, allow- synthetic bacteria acquired the capacity to photodis- ing H2O2 to build up locally (i.e, be released by an sociate water, leading to the geologically rapid ac- open floodgate) for signalling purposes (Ha l l et al. cumulation of molecular oxygen during the Great 2009, Ha n s c h m a n n et al. 2013). Oxidation Event (GOE), when anaerobic life un- The abundance of Prx is high: it can account derwent a catastrophic decline (Ed g a r et al. 2012). for up to 1% of all soluble cellular proteins (Wo o d Organisms that survived the transition to an aerobic et al. 2003). Furthermore, typical 2-Cys Prxs are the environment were those that respired and/or evolved largest and most widely distributed subfamily (So i t o oxygen. All oxygen-utiliszing organisms acquired et al. 2011). These Prxs are moonlighting proteins: constitutive antioxidant defensesdefences (both at high H2O2 concentrations they act as holdases, small molecules and enzymes), detoxifying and whereas when the rate of ROS formation is low they scavenging the ROS that are continuously produced are thioredoxin-dependent (Ja n g et al. as a by-product of aerobic metabolism (Acw o r t h et 2004). The dual functions of typical 2-Cys Prxs, al. 1997, Ha l l i w e l l & Gu t t e r i d g e 1999). Among modulating ROS concentrations and preventing pro- them, Prx has been object of an increasing inter- tein aggregation, may play pivotal roles in cellular est for its pivotal role in cell defensedefence and as response to pathogens and external stress (Ja n g et conserved marker for circadian rhythms in metabo- al. 2004). The importance of Prxs is unarguable, as lism. Ed g a r and colleagues (2012) studied the cy- transgenic/knockout mouse models overexpressing cles of Prx oxidation-reduction in Eukarya, Bacteria or deficient in most highly expressed Prxs has dem- and Archaea and proposed that all these organisms onstrated a decrease in genome stability and acceler- have cellular rhythms sharing a common molecular ated aging, and an overexpression of these proteins origin. In fact, it has been proposed that the ability is associated with various problems related to can- to survive cycles of may have con- cers treatment (Ha m i l t o n et al. 2012). Recent stud- tributed a selective advantage from the beginnings ies show Prxs expressed in tumortumour cells play of aerobic life. Twenty-four hours cycles of Prx positive roles in their progression and/or metastasis oxidation–reduction in all domains were observed in transplanted animals. Different functions of Prxs in both Archaea, Bacteria and Eukaryota supporting are required for their progression/metastasis in vivo the hypothesis that cellular rhythms share a common depending on tumortumour types (Is h i i et al. 2012). molecular origin (Ed g a r et al. 2012). However, only

306 New Insights into the Molecular Evolution of Metazoan Peroxiredoxins a few studies have been dedicated to the evolution of tion criterion (Akaike Information Criterion-AIC Prxs (Co p l e y et al. 2004; Kn o o p s et al. 2007; Pé r e z - and Corrected Akaike Information Criterion-cAIC). Sá n c h e z et al. 2011). Our study aims and we decided Models of nucleotide substitution allow for the cal- to extend the research on Prx evolution performing culation of probabilities of change between nucle- phylogenetic analysis on the most studied group of otides along the branches of a phylogenetic tree. The eucariotic organisms (Eukarya), the animals, using use of a particular substitution model may change Bayesian approach. Fortunately, there is sufficient the outcome of the phylogenetic analysis (Le m m o n information about many characterised nucleotide and Mo r i a r t y , 2004). Statistical model selection has and amino acid sequences of metazoan Prxs. Using become an essential step for the estimation of phyl- bioinformatics tools we analysed the phylogenetic ogenies from DNA sequence alignments. ProtTest3 relationships among metazoan Prxs, with the aim was used for the selection of the best-fit model of an- to acquire new data on the molecular evolution of alysed protein evolution (Da rr i b a et al. 2011). The these proteins. program ProtTest is one of the most popular tools for selecting models of amino acid replacement, a Material and Methods routine step in phylogenetic analysis. One hundred twenty-two candidate models of amino acid replace- Amino acid and mRNA coding sequences of ment and three types of criteria (Akaike Information different Prxs isoforms from protostomes were found Criterion-AIC, Corrected Akaike Information in the NCBI database (Table 1). These were Japanese Criterion-cAIC and Bayesian Information Criterion- spineless cuttlefish Sepiella maindroni Rochebrune, BIC) were used in these statistical analyses. The Prx 1884, cockscomb pearl mussel Cristaria plicata cDNA and amino acid sequences phylogenetic trees Leach, 1815, Pacific abalone Haliotis discus discus were built using the Bayesian inference (BI) method Reeve, 1846, Sydney rock oyster Saccostrea glom- implemented in Mr. Bayes 3.2 (Ro n q u i s t et al. 2012). erata (Gould, 1850), mud crab Scylla paramamosain Four independent runs, each one with four simulta- Estampador, 1949, Chinese mitten crab Eriocheir sin- neous Markov Chain Monte Carlo (MCMC) chains, ensis H. Milne-Edwards, 1853, lugworm Arenicola were performed for 1,000,000 generations sampled marina (Linnaeus, 1758) and southern house mos- every 1000 generations. FigTree software (http:// quito Culex quinquefasciatus Say, 1823. The follow- tree.bio.ed.ac.uk/software/figtree) was used to dis- ing deuterostomes were selected: rhesus macaque play the annotated phylogenetic trees. In order to de- Macaca mulatta (Zimmermann, 1780), Sumatran tect the presence of positive or negative selection on orangutan Pongo abelii Lesson, 1827, human Homo Prx molecular evolution we used statistical methods sapiens Linnaeus, 1758, European cattle Bos tau- implemented in HyPhy package (Ko s a k o v s k y Po n d rus Linnaeus, 1758, wild boar Sus scrofa Linnaeus, et al. 2005); SLAC, FEL and REL methods are able 1758, brown rat Rattus norvegicus Linnaeus, 1758, to detect the presence of possible positive selection house mouse Mus musculus Linnaeus, 1758, chicken and Mixed Effects Model of Evolution (MEME) Gallus gallus domesticus (Linnaeus, 1758), rock pi- program is able to detect even sites under episodic geon Columba livia Linnaeus, 1758, African clawed diversifying selection (Mu rr e l l et al. 2012). frog Xenopus laevis Daudin, 1802, western clawed frog Xenopus tropicalis (Gray, 1864), striped beakfish fasciatus (Temminck & Schlegel, 1844), Results and discussion zebrafish Danio rerio (F. Ha m i l t o n , 1822), gilthead Metazoan Prxs mRNA sequences were aligned us- seabream Sparus aurata Linnaeus, 1758,, channel ing T-Coffee software in combined libraries of local catfish Ictalurus punctatus (Rafinesque, 1818), rain- and multiple alignments, which are known to induce bow trout Oncorhynchus mykiss (Walbaum, 1792), high accuracy and performance in sequence align- Atlantic salmon Salmo salar Linnaeus, 1758, and ments. The obtained alignment was 1222 residues southern bluefin tuna Thunnus maccoyii (Castelnau, (nucleotides and gaps) long and the mean score val- 1872). All respective sequences were aligned us- ue was 59, indicating that the alignment was reliable ing T-Coffee multiple sequence alignment software (No t r e d a m e et al. 2000). jModelTest 0.1.1 software package (No t r e d a m e et al. 2000). determined that GTR+G model was the best-fit mod- jModelTest (Po s a d a 2008) was used to carry out el of Prxs cDNA sequence evolution with a gamma statistical selection of the best-fit models of nucle- shape value (four rate categories) of 0.72 using AIC otide substitution to analyse Prx molecular evolution. and cAIC statistical criteria (-lnL= 16365.654). Analyses were performed using 88 candidate models Phylogenetic relationships between all these differ- of nucleotide substitution and two types of informa- ent Prxs were determined using the most powerful

307 Bakiu R., G. Santovito

Table 1. Prx sequences used for Bayesian phylogeny, their NCBI accession numbers and the respective references Nucleotide accession Protein accession Species Isoform References number number Arenicola marina Prx6 Lo u m a y e et al., 2008 DQ059567 AAY96294 Bos taurus Prx6 Sa l m e r i et al., 2012 NM_174643 NP_777068 Columba livia Prx6 Ga o , 2012 JQ364950 AFD04441 Cristaria plicata Prx6 Pe i et al., 2010 HQ199304 ADN06076 Culex quinquefasciatus Prx6 At k i n s o n et al., 2007 XM_001861490 XP_001861525 Danio rerio Prx1 Co x et al., 2014 NM_001013471 NP_001013489 Danio rerio Prx2 Li u et al., 2013 NM_001002468 NP_001002468 Danio rerio Prx3 Lu et al., 2014 NM_001013460 NP_001013478 Danio rerio Prx4 Mu k a i g a s a et al., 2012 NM_001089425 NP_001082894 Danio rerio Prx5 Lu et al., 2014 NM_001024406 NP_001019577 Danio rerio Prx6 Mu k a i g a s a et al., 2012 NM_200805 NP_957099 Eriocheir sinensis Prx6 Mu et al., 2009 EU626070 ACF35639 Gallus gallus Prx6 Ca l d w e l l et al., 2005 NM_001039329 NP_001034418 Haliotis discus discus Prx6 Ni k a p i t i y a et al., 2006 EF103356 ABO26614 Homo sapiens Prx6 St r a u s b e r g et al., 2002 BC053550 AAH53550 Ictalurus punctatus Prx6 Ye h and Kl e s i u s , 2007 DQ779284 ABG77029 Macaca mulattaPrx6 Predicted NM_001266085 NP_001253014 Mus musculus Prx6 Pa c i f i c i et al., 2014 NM_007453 NP_031479 Oncorhynchus mykiss Prx6 Predicted NM_001165132 NP_001158604 Oplegnathus fasciatus Prx6 De Zo y s a et al., 2012 GQ903768 ADJ21808 Pongo abelii Prx6 Predicted NM_001132889 NP_001126361 Rattus norvegicus Prx6 Pa u l a et al., 2013 NM_053576 NP_446028 Saccostrea glomerata Prx6 Gr e e n et al., 2009 FJ626708 ACQ73550 Salmo salar Prx1 Lo o et al., 2012 NM_001140823 NP_001134295 Salmo salar Prx5 An d r e a s s e n a n d Ho y h e i m , 2013 BT149950 AGH92554 Scylla paramamosain Prx6 Fu et al., 2008 FJ429110 ACJ53746 Sepiella maindroni Prx6 So n g et al., 2010 HQ662844 AEI52300 Sparus aurata Prx1 Pe r e z -Sa n c h e z et al., 2011 GQ252679 ADI78064 Sparus aurata Prx2 Pe r e z -Sa n c h e z et al., 2011 GQ252680 ADI78065 Sparus aurata Prx3 Pe r e z -Sa n c h e z et al., 2011 GQ252681 ADI78066 Sparus aurata Prx4 Pe r e z -Sa n c h e z et al., 2011 GQ252682 ADI78067 Sparus aurata Prx5 Pe r e z -Sa n c h e z et al., 2011 GQ252683 ADI78068 Sparus aurata Prx6 Pe r e z -Sa n c h e z et al., 2011 GQ252684 ADI78069 Sus scrofa Prx6 Li u et al., 2011 NM_214408 NP_999573 Thunnus maccoyii Prx2 Su t t o n et al., 2010 EU093980 ABW88997 Xenopus (Silurana) tropicalis Prx6 Kl e i n et al., 2002 NM_001011325 NP_001011325 Xenopus laevis Prx6 Sh a f e r et al., 2011 NM_001089200 NP_001082669 statistical method of BI. We compute a majority rule (four rate categories) of 1.0 using all statistical cri- tree for all the trees sampled during the MCMC. We teria: AIC, cAIC and BIC (-lnL= -7241.91). In Fig. decided to use BI, because this method is much faster 2 is shown the best phylogeny generated by the ap- than the Maximum Likelihood (ML) for big datasets plication of BI method into the Prxs amino acid se- (Do u a d y et al. 2003). The best phylogeny generated quences. by the BI method is depicted in 1. In both cladrograms, three main clusters were Amino acid sequences of Prxs were aligned present: one including all typical Prx 2-Cys (isoforms and a 322 residue-long alignment was obtained. The 1, 2, 3 and 4), Prx5 isoforms (atypical 2-Cys Prxs) obtained alignment was better than that previously and the third grouping Prx6 isoforms (1-Cys). This achieved for nucleotide sequences, because its score distribution confirmed the previously obtained re- (86) was significantly better. ProtTest3 was used for sults using non-Bayesians methods (Pé r e z -Sá n c h e z the evolution best-fit model determination. WAG+G et al. 2011). In particular, the phylogenetic relation- resulted the best model, with a gamma shape value ships among the typical 2-Cys isoforms were com-

308 New Insights into the Molecular Evolution of Metazoan Peroxiredoxins - unrestricted

+ 0.013 β q-value 0.03823 0.05457 0.04247 0.03891 0.03903 0.01258 0.00808 0.00941 0.04105 0.01301 0.00781 0.03878 0.04037 0.03677 0.04499 – number of false of number –

0.003 0.006 0.0037 0.0016 0.0003 0.0028 0.0002 0.0001 0.0037 0.0003 0.0025 0.0033 0.0017 0.0016 p-value 3.00E-05 8.00E-05 q-value

1 1 1 1 1 1 1 1

1 1 1 0.83372 0.89837 0.72555 0.41649 0.98855 Pr[β=β+] MEME

lineage-specific and

β+ - β 3.0294 2.4262 2.0368 2.4091 4.8352 1.7028 2.6154 3.9012 1.8403 5.1242 3.1483 3.1601 1.5164 2.2162 1.5918 5.26E+00

Pr[β=β-] 6.00E-09 6.00E-09 6.00E-09 6.00E-09 1.66E-01 1.02E-01 6.00E-09 2.74E-01 6.00E-09 5.84E-01 6.00E-09 6.00E-09 6.00E-09 6.00E-09 1.15E-02 6.00E-09

0 0 0 0 0 0 0 0

0 0 0 β- 0.1248 0.2133 0.1115 0.0637 0.1948

0 0 α 0 0 0 0 0 0

0 0 0 0.083 0.1248 0.2133 0.1115 0.1948 – the number of false positive tests; positive false of number the – tor p-value 462273 64824.5 27229.4 6560730 1422990 2.45E+05 1.10E+11 13193000 30126900 1.74E+07 6.78E+07 3.26E+07 2.25E+03 2.041E+10 1.169E+09 220018000 934304000 777602000 Bayes Fac - 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 - Prob ability Posterior Posterior 1.00E+00 1.00E+00 1.00E+00 REL dS] ized E[dN- 1.45676 1.81062 1.56329 1.37872 1.50176 2.60861 1.35513 1.73989 1.69322 1.37862 1.42271 2.17128 2.03312 1.37357 1.35918 1.38826 1.36025 2.47379 α – number of synonymous substitutions for site; Normal - dS or E[dN] 2.0552 1.7522 1.6027 1.4364 1.7637 1.58E+00 1.47E+00 2.79E+00 1.81E+00 1.45E+00 1.52E+00 2.29E+00 2.12E+00 1.59E+00 1.43E+00 1.53E+00 1.43E+00 2.81E+00 0.093 0.101 E[dS] 0.1185 0.2445 0.1889 1.1835 0.0813 0.1145 0.0748 0.0705 0.0665 0.0976 0.0825 0.2155 0.0703 0.1444 0.0723 0.3335 empirical Bayes factors (ratio of posterior and prior odds of having ω =dN/dS > 1 at a given site) – two and p-value 0.00219 0.00119 0.00422 0.00135 7.05E-03 2.35E-03 1.90E-04 6.80E-05 4.10E-05 6.90E-05 3.52E-03 1.90E-04 3.70E-05 4.79E-03 1.10E-03 2.28E-03 1.32E-03 4.30E-04 ized dN-dS 0.21001 0.22375 0.38065 Normal - 0.351151 0.247727 0.290003 0.265428 0.286676 0.483288 0.336468 0.331806 0.236881 0.388245 0.289126 0.192282 0.254535 0.199977 0.510288 10.6 FEL dN/dS 20.113 23.687 32.625 10.814 Infinite Infinite Infinite Infinite Infinite Infinite 4.00E+08 4.00E+08 8.00E+08 5.00E+08 4.00E+08 3.00E+08 8.00E+08 posterior probabilities dN 1.9976 2.9797 2.5821 2.1403 2.4136 3.8971 1.6935 2.7132 2.6756 1.9101 1.8043 3.2297 3.0694 4.1148 2.3314 1.5505 2.2616 1.6126 0 0 0 0 0 0 dS 0.099 0.1482 0.2436 0.1019 0.2091 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 dN – number o nonsynonymous substitutions for site; p-value 0.00137 0.009462 0.001973 0.003961 0.009654 0.006305 0.007901 0.007809 0.003479 0.000602 0.000802 0.000899 0.003694 0.002922 0.008575 0.004778 0.008069 0.002773 ized SLAC dN-dS 1.0767 1.13119 1.11827 1.30723 1.02408 1.57071 1.17026 1.21023 1.14687 1.74533 1.32839 Normal - 0.758611 0.754975 0.856242 0.907304 0.746854 0.935665 0.753941 Positively selected codons identified by using computational techniques (SLAC-Single Ancestor Likelihood Counting, FEL-Fixed Effects Likelihood, REL-Ran dN-dS 8.7161 9.0526 6.0459 6.11164 9.15713 6.93142 10.5823 6.14108 8.29008 7.34477 12.7151 9.47341 9.79703 9.28409 7.57435 6.10328 14.1287 10.7535 16 106 124 129 139 149 159 162 185 210 220 227 236 260 261 263 269 308 Codon positive significant tests; – non-synonymous substitutions for site; dom Effects Likelihood and MEME-Mixed Effects Model of Evolution) and their calculated statistics: calculated their and Evolution) of Model Effects MEME-Mixed and Likelihood Effects dom Table S1. Table measures for determining whether a site is under positive selection

309 Bakiu R., G. Santovito

0.5198 0.5422 q-value Bayes 4.93E-01 8.06E-01 4.96E-01 5.08E-01 Factor 11019.9 75796.9 E. sinensis , E.

– number of number – dS 0.0209 0.0315 p-value 3.52E-02 1.44E-02 4.65E-02 3.17E-02 1 1 1 1 1

0.999825 0.998794 6.45E-01 Pr[β=β+] S. paramamosain , S. Posterior Probability Posterior

β+ 1.90262 1.69764 2.68E+00 2.37E+00 1.37E+00 1.67E+00 MEME REL A. marina , A.

Pr[β=β-] 6.00E-09 6.00E-09 3.55E-01 6.00E-09 6.00E-09 6.00E-09 -5.18972 -0.892806 0 0

β- C. plicata , C. Normalized E[dN-dS] , 4.89E-10 4.99E-09 4.99E-09 4.99E-09 0.21 0.28 E[dN] 0 0

α 5.00E-09 5.00E-09 5.00E-09 5.00E-09 X. tropicalis X. 1.098 5.465 E[dS] – number of nonsynonymous substitutions for site; for substitutions nonsynonymous of number – dN tor 127.76 140.889 253.813 221.996 134.566 240.347 99.9964 103.083 231.242 130.579 Bayes Fac - p-value 0.00095 4.21E-06 9.93E-01 9.96E-01 9.95E-01 9.93E-01 9.96E-01 9.90E-01 9.90E-01 9.96E-01 9.92E-01 9.92E-01 Posterior Posterior Probability dN-dS -0.148523 -0.771473 REL Normalized 0.4983 0.511949 0.498974 0.436025 0.440821 0.451779 0.652494 0.440754 0.444394 0.493914 E[dN-dS] Normalized FEL 0 0.07 dN/dS – the number of false positive tests; positive false of number the – E[dN] 5.11E-01 5.15E-01 4.49E-01 4.54E-01 4.65E-01 6.72E-01 5.15E-01 5.31E-01 4.58E-01 4.58E-01 0 (ratio of posterior and prior odds of having ω =dN/dS < 1 at a given site) – two measures two – site) given a at 1 < =dN/dS ω having of odds prior and posterior of (ratio factors Bayes empirical dN 0.46277 p-value and E[dS] 0.01968 0.016347 0.012936 0.013549 0.016986 0.012953 0.016732 0.019351 0.017199 0.013299 dS 1.1976 6.6837 p-value 2.45E-02 4.94E-02 4.81E-02 2.82E-02 3.77E-02 1.28E-02 2.80E-02 2.76E-02 4.87E-02 4.45E-02 1.1976 p-value 0.00026 dN-dS S. glomerata. Computational techniques (FEL-Fixed Effects Likelihood, REL-Random Effects Likelihood and MEME-Mixed Effects 0.89455 0.56883 1.24154 0.691165 0.601366 0.886091 0.892736 0.955191 0.668627 0.620555 Normalized posterior probabilities posterior FEL dN/dS SLAC 3.46E+08 2.20E+08 2.33E+08 3.43E+08 2.68E+08 4.81E+08 3.46E+08 3.70E+08 2.59E+08 2.40E+08 -1.40252 -0.597586 Normalized dN-dS dN S. maindroni and 1.7315 1.2942 1.20115 1.10103 1.16401 1.71512 1.33782 2.40312 1.72799 1.84887 dN-dS dS Negatively selected codons identified by using computational techniques (SLAC-Single Ancestor Likelihood Counting, FEL-Fixed Effects Likelihood and REL- -11.358 -4.8376 Positively selected codons identified using the data subset including Prx6 coding sequences from sequences coding Prx6 including subset data the using identified codons selected Positively 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 20 163 96 112 108 210 220 126 143 147 199 127 Codon Codon for determining whether a site is under negative selection synonymous substitutions for site; for substitutions synonymous Random Effects Likelihood) and their calculated statistics: calculated their and Likelihood) Effects Random Table S2. Table H. discus discus , Model of Evolution) were applied and calculated statistics are reported Table S3. Table

310 New Insights into the Molecular Evolution of Metazoan Peroxiredoxins E. 135611 252936 11972.5 1.22E+06 4.51E+09 1.31E+07 5.23E+03 7.58E+06 1.32E+03 2.85E+05 5.04E+06 1.63E+03 8.58E+06 1.63E+05 1.83E+03 9.84E+08 1.69E+03 1.09E+08 1.85E+04 6.85E+03 1.88E+07 4.00E+04 1.16E+03 1.04E+07 2.08E+03 8.18E+04 Bayes Factor S. paramamosain , 1 1 1 1 1 1 1 1 1 0.99981 0.99941 0.999993 0.999996 0.999999 0.999917 0.999245 0.999997 0.999389 0.999994 0.999457 0.999946 0.999855 0.999975 0.999146 0.999523 0.999988 Posterior Posterior Probability A. marina , REL , -2.5516 -2.18475 -1.64936 -3.37202 -1.98922 -3.33257 -3.67627 -1.92782 -3.15637 -1.72424 -2.93308 -1.69159 -1.83574 -1.86608 -1.79844 -1.84857 -2.74328 -1.67556 -2.24469 -3.31459 -1.13071 -1.90421 -1.59868 -1.77969 -2.28174 -1.79268 E[dN-dS] Normalized C. plicata , E[dN] 0.02491 0.026502 0.027609 0.019916 0.027414 0.036982 0.051906 0.021085 0.333878 0.025746 0.021624 0.027609 0.024408 0.024909 0.027592 0.035383 0.028729 0.054331 0.204275 0.032041 0.036989 0.042866 0.035435 0.052531 0.020371 0.343866 X. tropicalis E[dS] 1.7192 2.21125 1.67697 3.39193 2.01664 3.36955 3.72818 1.94891 3.49025 1.74999 2.95471 1.86015 1.89099 1.82604 1.88395 2.30211 2.76819 1.70428 2.29902 3.51887 1.16275 2.58859 1.94708 1.63412 1.83222 2.13655 p-value 0.00464 0.00092 4.19E-05 3.13E-05 1.27E-03 5.83E-03 6.03E-04 3.34E-04 6.35E-03 4.66E-05 0.000451 4.80E-03 0.000758 7.18E-04 0.001043 1.07E-03 5.98E-03 0.000163 2.24E-04 5.15E-03 0.001214 9.69E-03 6.36E-04 0.000169 2.32E-03 6.59E-03 -641.38 -2.2849 -1.53117 -3.46156 -2.04555 -213.165 -2.78656 -7.63185 -2.61438 -8.24837 -786.334 -1.98971 -1.90533 -2.32698 -2.09704 -6.81587 -1.74044 -6.95794 -16.3374 -7.26305 -1.23758 -1.24787 -2.67551 -134.623 -2.90222 -0.940304 Normalized dN-dS FEL 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 dS dN/ 0.017 0.006 0.049 dN 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 2.76E-01 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 2.00E-01 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 5.00E-09 2.90E-01 dS 3.8513 5.9078 6.70022 3.95938 412.604 5.39369 4.50411 14.7723 1241.46 5.06041 16.2418 2.96374 1522.03 3.68796 4.42266 4.05905 13.1928 3.36881 13.4678 31.8233 1.82006 14.0584 2.39547 2.41539 5.17873 260.578 Computational techniques (SLAC-Single Likelihood Ancestor S. Counting, glomerata. FEL-Fixed Likelihood Computational and Effects techniques REL- (SLAC-Single Likelihood p-value 0.01249 0.004115 0.002558 0.037037 0.004216 0.004394 0.004115 0.020774 0.030038 0.018305 0.004708 0.012346 0.018149 0.012346 0.037037 0.024259 0.037508 0.016891 0.023387 0.034967 0.012346 0.039348 0.033926 0.012346 0.017937 0.031831 and SLAC S. maindroni -1.8856 -1.8662 -1.8856 -1.1595 -1.8772 -2.75765 -1.13136 -2.33447 -2.30243 -1.50848 -1.74432 -1.52594 -1.85826 -2.25015 -1.50848 -1.50848 -1.13136 -1.86796 -1.40377 -1.42627 -1.97627 -1.50848 -1.13136 -1.49415 -1.50848 -1.80609 Normalized dN-dS dN-dS -4.4566 -5.33772 -2.18987 -4.51862 -3.64979 -2.91983 -3.37632 -2.95361 -3.59685 -4.35539 -2.91983 -3.61222 -2.91983 -2.18987 -3.61563 -3.64979 -2.71714 -2.76069 -3.82529 -2.24434 -2.91983 -2.18987 -2.89208 -2.91983 -3.63352 -3.49587 Negatively selected codons identified by using the data subset including Prx6 coding sequences from H. discus discus , 11 10 12 13 15 20 25 35 36 38 39 41 43 44 46 48 49 50 59 63 69 71 72 75 78 178 Codon Random Effects Likelihood) were applied and calculated statistics are reported Random Effects Table S4. Table sinensis ,

311 Bakiu R., G. Santovito patible with the hypothesis exposed by Co p l e y and lection) on protein coding is often difficult to colleagues (2004) which suggested an evolutionary identify because selection is frequently transient or model based on the appearance of 3 groups of Prxs episodic, i.e. it affects only a subset of lineages. To (1-2, 3 and 4) through the loss of N-terminal exten- verify the selection type in Prx evolution we used sion in Prx3s and the acquisition of C-terminal ex- existing computational techniques (SLAC, FEL and tension in Prx4s. REL) implemented in the HyPhy package. These The deuterostomes Prxs were equally posi- techniques are designed to identify sites subject to tioned and the respective branches were supported pervasive selection (a large proportion of positively by the highest posterior probability values. The selected sites) but may fail to recognise sites where only exception was the Prx6 of X. tropicalis, which selection is episodic. For this reason we used MEME in both trees emerged far from the sequence of X. method that is able to identify instances of both epi- laevis. These results were peculiar. They might sug- sodic and pervasive positive selection at individual gest the appearance of Prx6 at different times, site level. The obtained results (Table S1) indicated through independent duplication events of the ances- that nearly 5% of the 308 codons were positively se- tor, even after the speciation event that led to the dif- lected and Table S2less than 1% of them were nega- ferentiation of Chordata phylum. Based on the cur- tively selected (Table S2). These results suggested rently available data, it was not possible to confirm that Prx genes were more susceptible to positive se- that the hypothetical multiple appearance of Prx6 lection than to negative one. could be a distinguishing feature of Amphibians or It is known that if all existing gene sequences might indeed have happened also in other classes. had not been screened for recombination, selection Whatever the case, this duplication event could not analyses of alignments with recombinants in them be recent but should have occured at least 416-360 using a single tree could generate misleading re- million years ago, and might be typical of Prx6. sults (Ko s a k o v s k y Po n d et al. 2005). Thus, we used In fact, among the various isoforms, only Prx6 are GARD program (Ko s a k o v s k y Po n d et al. 2006) known to be represented by multiple sub-isoforms, to identify possible breakpoints in the Prx gene such as in Drosophila melanogaster (DPX-6005 and sequences. One breakpoint has been found (-lnL DPX-2540, Ra d y u k et al. 2001) and Ciona intesti- = 31879), but Kishino Hasegawa (KH) tests indi- nalis (unpublished personal data). cated it as statistically non-significant breakpoint in The distribution of sequences was each level of significance. However, the calculated less consistent. All protostome Prx6 sequences were mean substitution value was 2.89 substitutions per grouped except for the protein of C. quinquefascia- site, revealing that there was a huge amount of di- tus that emerged as a single branch (Fig. 1). This vergence among all the analysed sequences. In this position was the same in the phylogenetic analysis case all the possible positive or negative selection performed on amino acid sequences, but in this case tests could be non-reliable. In order to verify the re- it resulted to be the sister group of invertebrate Prxs sults we carried out the selection tests for a subset of (Fig. 2). Instead, the sequence of A. marina seems to sequences that were not so divergent. We included be more related to vertebrate Prxs. Probably, the dis- in this data subset only the Prx sequences that were cordance emerged from the comparison of the two responsible for the emerged discordance between phylogenetic topologies may be linked to a possible the two phylogenetic topologies. The Prx6 coding differences in substitution rates, which are usually sequences of X. tropicalis, C. plicata, A. marina, caused by positive and/or negative selection. S. paramamosain, E. sinensis, H. discus discus, Evolutionary biologists typically have invoked S. maindroni and S. glomerata were aligned using two types of selective forces shaping the evolution T-Coffee software. We obtained a good score (92). of species. One is the purifying selection, which The calculated mean substitutions value was 0.77 favours the conservation of existing phenotypes. substitutions per site, which meant that the diver- The other is the positive selection (also known as gence among the subset sequences was low. Thus, Darwinian selection), which promotes the emer- we carried out the previously used selection tests gence of new phenotypes. Positive selection can (Tables S3 & S4). Additionally, for this data sub- leave a set of telltale signatures in the genes, such set GARD application found one breakpoint (-lnL as the rapid divergence of functional sites between = 8983.58), but Kishino Hasegawa (KH) tests as- species (diversifying selection) and the depression signed it as statistically non-significant breakpoint of polymorphism within species (Kr e i t m a n et al. in each level of significance. No positive selection 2000, Ya n g & Bi e l a w s k i 2000, Ba m s h a d & Wo o d i n g results were obtained by using the SLAC method 2003). The imprint of natural selection (positive se- (Table S3).

312 New Insights into the Molecular Evolution of Metazoan Peroxiredoxins

Fig. 1. Phylogenetic relationships among metazoan Prxs using coding sequences and BI (Bayesian Inference) method (arithmetic mean = -29057.97; harmonic mean = -29083.98). Posterior probability values higher than 50% are indi- cated on each node. The scale for branch length (2.0 substitution/site) is shown below the tree

We compared the results presented in Table S4 tebrates, the number of negatively selected codons with the crystal structure of Prx1 from S. mansoni (about 12% of the 224 codons) was higher than the (SmPrx1, PDB code 3ztlD, solved by X-ray, reso- positively selected ones (approximately 5%, Tables lution 3.00Å) characterised by Saccoccia and col- S3 & S4). leagues (2012). In the active site of SmPrx1 low- All these results confirmed that the molecular molecular-weight dimer, residues from 47 to 50, evolution of metazoan Prx was peculiar and suggest- 48 including the CP (Cys ), form the first turn of the α2 ed that the positive selection may have operated into helix. The compact hydrophobic pocket is formed by the evolution of these proteins and a purifying selec- Tyr40, Pro41, Ala42, Thr45 and Pro49 that, together with tion has been present during this process. Probably, 124 49 Arg , surround the CP (Saccoccia et al. 2012). Pro the natural selection was responsible for the devia- is highly conserved in typical 2-Cys Prxs (Ho f m a n n tion from the evolution pathway of Prx genes char- et al. 2002) and reduces the propensity of the first acteristic of the protostome isoforms. turn of the α2 helix to retain its secondary structure. Further analyses are needed to verify if the Our results indicated that only codons 41, 48 and 49 positively selected codons, identified by bioinfor- were negatively selected (Table S4), confirming that matic approach, really codify essential amino acids. some (but not all) of the important for peroxidase ac- A possible verification could be done by site-specific tivity amino acids were conserved during the evolu- mutagenesis of those nucleotides that could structur- tion of Prxs. While these data confirmed the sound- ally and functionally affect the enzymatic activity of ness of our results, on the other hand it brought out these proteins. some questions on the absence of conservation for Purifying selection is important for the evo- the other amino acids that were important for cata- lution of a , because it can help the be- lytic activity. In this subset of sequences from inver- longing genes to maintain their optimal function.

313 Bakiu R., G. Santovito

Fig. 2. Phylogenetic relationships among metazoan Prxs using amino acid sequences and BI (Bayesian Inference) method (arithmetic mean = -12412.89; harmonic mean = -12437.68). Posterior probability values higher than 50% are indicated on each node. The scale for branch length (2.0 substitution/site) is shown below the tree

However, positive selection is an important source in reproduction (abalone sperm lysin, sea urchin of evolutionary innovation and is a major force un- bindin, proteins in mammals) and proteins that ac- derlying the adaptation of species to a new envi- quire new functions after gene duplication (Yang & ronment (Ko s i o l et al. 2008). Many proteins have Bi e l a w s k i 2000). We suggest to include Prxs among been found under positive selection, such as those those proteins, as their diversification allowed the involved in immunity (MHC, immunoglobulin VH, cells of all organisms to acquire a powerful defence class 1 chitinas), proteins or pheromones involved system against the risk of oxidative stress.

References

Acw o r t h I. N., Mc Ca b e D. R. and Ma h e r T. J. 1997. The analysis W., Sh u m w a y M., Si s k P., St o l t e C., Ze n g Q., Ei s e n s t a d t of fre radicals, their reaction products, and antioxidants. In: E., Fr a s e r -Li g g e t t C., St r a u s b e r g R., Ga l a g a n J., Bi rr e n Ba s k i n S.I., Sa l e m H editors, Oxidants, Antioxidants, and B. and Co l l i n s F. H. 2007. The Broad Institute Genome Free Radicals. Washington DC: Taylor and Francis, 23-77. Sequencing Platform Annotation of Culex pipiens quin- An d r e a s s e n R. and Ho y h e i m B. 2013. Full length cDNAs in salmo quefasciatus. Unpublished. salar pre-smolt and tissue specific libraries. Unpublished. Bamshad M. & Wooding S.P. 2003. Signatures of natural selec- At k i n s o n P. W., He m i n g w a y J., Ch r i s t e n s e n B. M., Hi g g s S., tion in the . – Nature Reviews Genetics, Ko d i r a C., Ha n n i ck L., Me g y K. O’Le a r y S., Pe a r s o n 4: 99-111. M., Ha a s B. J., Ma u c e l i E., Wo r t m a n J. R., Le e N. H., Ca l d w e l l R. B., Ki e r z e k A. M., Ar a k a w a H., Be z z u b o v Y., Za i m Gu i g o R., St a n k e M., Al v a r a d o L., Am e d e o P., An t o i n e J., Fi e d l e r P., Ku t t e r S., Bl a g o d a t s k i A., Ko s t o v s k a D., C. H., Ar e n s b u r g e r P., Bi d w e l l S. L., Cr a w f o r d M., Ca- Ko t e r M., Pl a c h y J., Ca r n i n c i P., Ha y a s h i z a k i Y. and m a r o F., De v o n K., En g e l s R., Ha m m o n d M., Ho w a r t h Bu e r s t e d d e J. M. 2005. Full-length cDNAs from chicken C., Ko e h r s e n M., La w s o n D., Mo n t g o m e r y P., Ne n e V., bursal lymphocytes to facilitate gene function analysis. – Nu s b a u m C., Pu i u D., Ro m e r o -Se v e r s o n J., Se v e r s o n D. Genome Biology, 6: R6.

314 New Insights into the Molecular Evolution of Metazoan Peroxiredoxins

Ca o Z., Li n d s a y J. G. and Is a a c s N. W. 2007. Mitochondrial Publications. peroxiredoxins. – Subcellular Biochemistry, 44: 295-315. Ha m i l t o n R.T., Wa l s h M.E. and Va n Re m m e n H. 2012. Mouse Ch a M. K., Yu n C. H. and Ki m I. H. 2000. Interaction of human Models of Oxidative Stress Indicate a Role for Modulating thiol-specific antioxidant protein 1 with erythrocyte plasma Healthy Aging. – Journal of Clinical and Experimental membrane. – Biochemistry, 39: 6944-6950. Pathology, S4: 1–14. Ch a e H. Z., Ch u n g S. J. and Rh e e S. G. 1994. Thioredoxin- Ha n s c h m a n n E-A, Go d o y J.R., Be r n d t C., Hu d e m a n n C. and dependent peroxide reductase from yeast. – Journal of Li l l i g C.H. 2013. Thioredoxins, glutaredoxins, and perox- Biological Chemistry, 269: 27670-27678. iredoxins--molecular mechanisms and health significance: Ch e n J. W., Do d i a C., Fe i n s t e i n S. I., Ja i n M. K. and Fi s h e r A. from cofactors to antioxidants to redox signaling. – Anti- B. 2000. 1-Cys peroxiredoxin, a bifunctional enzyme with oxidants and Redox Signaling, 19:1539-605. and phospholipase A2 activities. – Ho f m a n n B., He c h t H.J. and Fl o h é L. 2002. Peroxiredoxins. – Journal of Biological Chemistry, 275: 28421-28427. Biological Chemistry, 383: 347–364. Co n w a y Mo rr i s S., Pe e l J. S., Hi g g i n s A. K., So p e r N. J. and Im m e n s c h u h S., Ba u m g a r t -Vo g t E., Ta n M., Iw a h a r a S., Ra- Da v i s N. C. 1987. A burgess shale-like fauna from the lower m a d o r i G. and Fa h i m i H.D. 2003. Differential cellular and cambrian of north greenland. – Nature, 345: 802-805. subcellular localization of heme-binding protein 23/perox- Co p l e y S. D., No v a k W. R. P. and Ba b b i t t P. C. 2004. Divergence iredoxin I and heme oxygenase-1 in rat liver. – Journal of of Function in the Thioredoxin Fold Suprafamily: Evidence Histochemistry and Cytochemistry. 51: 1621–1631. for Evolution of Peroxiredoxins from a Thioredoxin-like Is h i i T., Wa r a b i E. and Ya n a g a w a T. 2012. Novel roles of Ancestor. – Biochemistry, 43: 13981-13995. peroxiredoxins in inflammation, and innate im- Co x A. G., Sa u n d e r s D. C., Ke l s e y P. B. Jr, Co n w a y A. A., Te s - munity. – Journal of Clinical Biochemistry and Nutrition, m e n i t s k y Y., Cr o c e K. J., No r t h T. E. and Go e s s l i n g W. 50: 91-105. 2014. S-nitrosothiol signaling regulates liver development Ja n g H.H., Le e K.O., Ch i Y.H., Ju n g B.G., Pa rk S.K., Pa rk J.H., and improves outcome following toxic liver injury. – Cell Le e J.R., Le e S.S., Mo o n J.C., Yu n J.W., Ch o i Y.O., Ki m Reports, 6: 56-69. W.Y., Ka n g J.S., Ch e o n g G-W, Yu n D-J, Rh e e S.G., Ch o Da rr i b a D., Ta b o a d a G.L., Do a l l o R. and Po s a d a D. 2011. Prot- M.J. and Lee S.Y. 2004. Two enzymes in one; two yeast Test 3: fast selection of best-fit models of protein evolution. peroxiredoxins display oxidative stress-dependent switch- – Bioinformatics, 27: 1164-1165. ing from a peroxidase to a molecular chaperone function. De Zo y s a M., Ry u J. H., Ch u n g H. C., Ki m C. H., Ni k a p i t i y a C., – Cell, 117: 625–635. Oh C., Ki m H., Sa r a n y a Re v a t h y K., Wh a n g I. and Le e Ki m K.W., Ki m I.H., Le e K.Y., Rh e e S.G. and St a d t m a n E.R. J. 2012. Molecular characterization, immune responses 1988. The isolation and purification of a specific protec- and DNA protection activity of rock bream (Oplegnathus tor protein which inhibits enzyme inactivation by a thiol/

fasciatus), peroxiredoxin 6 (Prx6). – Fish Shellfish Im- Fe(III)/O2 mixed-function oxidation system. – Journal of munology, 33: 28-35. Biological Chemistry, 263: 4704–4711. Do u a d y C. J., De l s u c F., Bo u c h e r Y., Do o l i t t l e W. F. and Do u - Kl e i n S.L., St r a u s b e r g R.L., Wa g n e r L., Po n t i u s J., Cl i f t o n z e r y E.J.P. 2003. Comparison of Bayesian and maximum S.W. and Ri c h a r d s o n P. 2002. Genetic and genomic tools likelihood bootstrap measures of phylogenetic reliability. for Xenopus research: The NIH Xenopus initiative. – De- – Molecular Biology Evolution, 20: 248-254. velopmental Dynamics, 225: 384-391. Ed g a r R. S., Gr e e n E. W., Zh a o Y., v a n Oo i j e n G., Ol m e d o M., Kn o o p s B., Lo u m a y e E. and Va n De r Ee ck e n V. 2007. Evolu- Qi n X., Xu Y., Pa n M., Va l e k u n j a U. K., Fe e n e y K. A., tion of the peroxiredoxins. – Subcellular Biochemistry, Ma y w o o d E. S., Ha s t i n g s M. H., Ba l i g a N. S., Me rr o w 44: 27–40. M., Mi l l a r A.J., Jo h n s o n C. H., Ky r i a c o u C. P., O’Ne i l l Ko s a k o v s k y Po n d S.L. and Fr o s t S.D.W. 2005. Not so different J. S. and Re d d y A. B. 2012. Peroxiredoxins are conserved after all: a comparison of methods for detecting amino acid markers of circadian rhythms. – Nature, 485: 459-464. sites under selection. – Molecular Biology and Evolution, Fi s h e r A. B. 2011. Peroxiredoxin 6: a bifunctional enzyme with 22: 1208–1222. glutathione peroxidase and phospholipase A2 activities. – Ko s a k o v s k y Po n d S.L., Fr o s t S.D.W. and Mu s e S.V. 2005. Antioxidants and Redox Signaling, 15: 831-844. HyPhy: hypothesis testing using phylogenies. – Bioinfor- Fu M. J., Zo u Z. H., Zh a n g Z. P. and Wa n g Y. L. 2008. Molecu- matics, 21: 676–679. lar characterization and expression profile of glutathione Ko s a k o v s k y Po n d S.L., Po s a d a D., Gr a v e n o r M.B., Wo e l k C.H. peroxidase in the developing ovary and testis of Scylla and Fr o s t S.D.W. 2006. Automated phylogenetic detection paramamosain. Unpublished. of recombination using a genetic algorithm. – Molecular Ga o P. F. 2012. Molecular Cloning and cDNA Sequence Analysis Biology and Evolution, 23: 1891–1901. of Peroxiredoxin of Pigeon. Unpublished. Ko s i o l C., Vi n a r T., d a Fo n s e c a R.R., Hu b i s z M.J., Bu s t a m a n t e Gr e e n T. J., Di x o n T. J., De v i c E., Ad l a r d R. D. and Ba r n e s C.D., Ni e l s e n R. and Si e p e l A. (2008). Patterns of positive A. C. 2009. Differential expression of genes encoding selection in six mammalian genomes. – PLOS Genetics, anti-oxidant enzymes in Sydney rock oysters, Saccostrea 4: e1000144. glomerata (Gould) selected for disease resistance. – Shell- Kr e i t m a n M. 2000. Methods to detect selection in populations fish Immunology, 26: 799-810. with applications to the human. -Annual Review of Genom- Ha l l A., Ka rp l u s P. A. and Po o l e L. B. 2009. Typical 2-Cys ics and Human Genetics, 1: 539–559. peroxiredoxins – structures, mechanisms and functions. – Le m m o n A.R. and Mo r i a r t y E.C. 2004. The importance of proper FEBS Journal, 276: 2469–2477. model assumption in Bayesian phylogenetics. – Systematic Ha l l i w e l l B. and Gu t t e r i d g e J. M. C. 1999. Free Radicals in Biology, 53: 265-277. Biology and Medicine, 3rd ed. Oxford: Oxford Science Li u C., Xu H., La m S.H. and Go n g Z. 2013. Selection of reliable

315 Bakiu R., G. Santovito

biomarkers from PCR array analyses using relative distance Pe r e z -Sa n c h e z J., Be r m e j o -No g a l e s A., Ca l d u c h -Gi n e r J.A., computational model: methodology and proof-of-concept Ka u s h i k S. and Si t j a -Bo b a d i l l a A. 2011. Molecular study. – PLoS ONE, 8: E83954. characterization and expression analysis of six peroxire- Li u Y., Wu W.J., Zu o B., Re n Z.Q. and Xi o n g Y.Z. 2011. Polymor- doxin paralogous genes in gilthead sea bream (Sparus phism in coding region of pig PRDX6 gene and its genetic aurata): Insights from fish exposed to dietary, pathogen effects analysis. – Yi Chuan, 33: 743-748. and confinement stressors. – Fish Shellfish Immunology, Lo o G.H., Su t t o n D.L. and Sc h u l l e r K.A. 2012. Cloning and 31: 294-302. functional characterisation of a peroxiredoxin 1 (NKEF A) Pé r e z -Sá n c h e z J., Be r m e j o -No g a l e s A., Ca l d u c h -Gi n e r J.A., cDNA from Atlantic salmon (Salmo salar) and its expres- Ka u s h i k S. and Si t j à -Bo b a d i l l a A. 2011. Molecular sion in fish infected withNeoparamoeba perurans. – Fish characterization and expression analysis of six peroxire- Shellfish Immunology, 32: 1074-1082. doxin paralogous genes in gilthead sea bream (Sparus Lo u m a y e E., An d e r s e n A.C., Cl i pp e A., De g a n d H., Du b u i s s o n aurata): insights from fish exposed to dietary, pathogen M., Za l F., Mo r s o m m e P., Re e s J.F. and Kn o o p s B. 2008. and confinement stressors. – Fish Shellfish Immunology, Cloning and characterization of Arenicola marina perox- 31: 294–302. iredoxin 6, an annelid two-cysteine peroxiredoxin highly Ph a l e n T.J., We i r a t h e r K., De m i n g P.B., An a t h y V., Ho w e A.K., homologous to mammalian one-cysteine peroxiredoxins. v a n d e r Vl i e t A., Jö n s s o n T.J., Po o l e L.B. & He i n t z N.H. – Free Radical Biology and Medicine, 45: 482-493. 2006. Oxidation state governs structural transitions in Lu A., Hu X., Wa n g Y., Sh e n X., Li X., Zh u A., Ti a n J., Mi n g Q. peroxiredoxin II that correlate with cell cycle arrest and and Fe n g Z. 2014. iTRAQ analysis of gill proteins from recovery. – Journal of Cell Biology, 175: 779–789. the zebrafish (Danio rerio) infected with Aeromonas hydro- Po s a d a D. 2008. jModelTest: phylogenetic model averaging. – phila. – Fish Shellfish Immunology, 36: 229-239. Molecular Biology and Evolution, 25: 1253–1256. Marchini J.F., Brown K.K., Stamler J.S., Colagiovanni Ra d y u k S.N., So h a l R.S. and Orr W.C. 2001. Thioredoxin D.B., Rosenthal G.J., peroxidases can foster cytoprotection or cell death in re- Mu C., Zh a o J., Wa n g L., So n g L., Zh a n g H., Li C., Qi u L. and sponse to different stressors: over- and under-expression of Ga i Y. 2009. Molecular cloning and characterization of thioredoxin peroxidase in Drosophila cells. – Biochemistry peroxiredoxin 6 from Chinese mitten crab Eriocheir sin- Journal, 371: 743–752. ensis. – Fish Shellfish Immunology, 26:821-827. Rh e e S.G. and Wo o H.A. 2011. Multiple functions of peroxiredox- Mu k a i g a s a K., Ng u y e n L.T., Li L., Na k a j i m a H., Ya m a m o t o M. ins: peroxidases, sensors and regulators of the intracellular and Ko b a y a s h i M. 2012. Genetic evidence of an evolu- messenger H2O2, and protein chaperones. – Antioxidants tionarily conserved role for Nrf2 in the protection against and Redox Signaling, 15: 781–794. oxidative stress. – Molecular and Cellular Biology, 32: Rh e e S.G., Wo o H.A., Ki l I.S. and Ba e S.H. 2012. Peroxiredoxin 4455-4461. functions as a peroxidase and a regulator and sensor of Mu rr e l l B., We r t h e i m J.O., Mo o l a S., We i g h i l l T., Sc h e f f l e r local peroxides. – Journal of Biological Chemistry, 287: K. and Ko s a k o v s k y Po n d S.L. 2012. Detecting individual 4403–4410. sites subject to episodic diversifying selection. – PLOS Ro n q u i s t F., Te s l e n k o M., v a n d e r Ma rk P., Ay r e s D.L., Da r - Genetics, 8: e1002764. l i n g A., Hö h n a S., La r g e t B., Li u L., Su c h a r d M.A. and Ni k a p i t i y a C., Ka n g H.-S. and Le e J. 2006. Novel gene discovery Huelsenbeck J.P. 2012. MrBayes 3.2: efficient bayesian from Haliotis discus discus by normalized cDNA library phylogenetic inference and model choice across a large analysis. Unpublished. model space. – Systematic Biology, 61: 539–542. No t r e d a m e C., Hi g g i n s D.G. and He r i n g a J. 2000. T-Coffee: A Sa cc o cc i a F., Di Mi cc o P., Bo u m i s G., Br u n o r i M., Ko u t r i s I., novel method for fast and accurate multiple sequence align- Mi e l e A.E., Mo r e a V., Sr i r a t a n a P., Wi l l i a m s D.L., Be l - ment. – Journal of Molecular Biology, 302: 205–217. l e l l i A. and An g e l u cc i F. 2012. Moonlighting by different stressors: crystal structure of the chaperone species of a Ok a d o -Ma t s u m o t o A., Ma t s u m o t o A., Fu j i i J. and Ta n i g u c h i N. 2000. Peroxiredoxin IV is a secretable protein with heparin- 2-Cys peroxiredoxin. – Cell, 20: 429–439. binding properties under reduced conditions. – Journal of Sa l m e r i M., Mo t t a C., Ma s t r o j e n i S., Am o d e o A., An f u s o C.D., Biochemistry, 127: 493–501. Gi u r d a n e l l a G., Mo r e l l o A., Al b e r g h i n a M., To s c a n o M.A. and Lu p o G. 2012. Involvement of PKCalpha-MAPK/ Pa c i f i c i F., Arr i g a R., So r i c e G.P., Ca p u a n i B., Sc i o l i M.G., ERK-phospholipase A(2) pathway in the Escherichia coli Pa s t o r e D., Do n a d e l G., Be l l i a A., Ca r a t e l l i S., Co pp o l a invasion of brain microvascular endothelial cells. – Neu- A., Fe rr e l l i F., Fe d e r i c i M., Sc o n o cc h i a G., Te s a u r o M., roscience Letters, 511: 33-37. Sb r a cc i a P., De l l a -Mo r t e D., Gi a cc a r i A., Or l a n d i A. and La u r o D. 2014. Peroxiredoxin 6, a novel player in the Sh a f e r M.E., Wi l l s o n J.A. and Da m j a n o v s k i S. 2011. Expres- pathogenesis of diabetes. – Diabetes, 63:3210-3220. sion analysis of the peroxiredoxin gene family during early development in Xenopus laevis. – Pa u l a F.M., Fe rr e i r a S.M., Bo s c h e r o A.C. and So u z a K.L. 2013. Modulation of the peroxiredoxin system by Patterns, 11: 511-516. in insulin-producing RINm5F cells: down-regulation of So i t o L., Wi l l i a m s o n C., Kn u t s o n S.T., Fe t r o w J.S., Po o l e L.B. PRDX6 increases susceptibility of beta cells to oxidative and Ne l s o n K.J. 2011. PREX: PeroxiRedoxin classifica- stress. – Molecular and Cellular Endocrinology, 374: tion indEX, a database of subfamily assignments across the 56-64. diverse peroxiredoxin family. – Nucleic Acids Research, 39: 332–337. Pe i P., We n C. and Hu B. 2010. Molecular cloning, characteriza- tion and expression analysis of peroxiredoxin 6 from Cris- So n g W., Wa n g C., Sh a o C., Mu C., Zh a n g Y. and Zh a n P. 2010. taria plicata and the antioxidant activity of its recombinant Molecular cloning and characterization of a novel peroxire- protein. Unpublished. doxin 6 protein from Sepiella maindroni. Unpublished.

316 New Insights into the Molecular Evolution of Metazoan Peroxiredoxins

So r o k i n a E.M., Fe i n s t e i n S.I., Zh o u S. and Fi s h e r A.B. 2011. National Academy of Sciences U.S.A., 99: 16899-16903. Intracellular targeting of peroxiredoxin 6 to lysosomal Su t t o n D.L., Lo o G.H., Me n z R.I. and Sc h u l l e r K.A. 2010 organelles requires MAPK activity and binding to 14-3- Cloning and functional characterization of a typical 2-Cys 3ε. – American Journal of Physiology-Cell Physiology,300: peroxiredoxin from southern bluefin tuna (Thunnus mac- 1430–1441. coyii). – Comparative Biochemistry and Physiology Part B: St r a u s b e r g R.L., Fe i n g o l d E.A., Gr o u s e L.H., De r g e J.G., Biochemistry and Molecular Biology, 156: 97-106. Kl a u s n e r R.D., Co l l i n s F.S., Wa g n e r L., Sh e n m e n C.M., Va l e n t i n e J.W. 1997. Cleavage patterns and the topology of meta- Sc h u l e r G.D., Al t s c h u l S.F., Ze e b e r g B., Bu e t o w K.H., zoan tree of life. – Proceedings of the National Academy Sc h a e f e r C.F., Bh a t N.K., Ho pk i n s R.F., Jo r d a n H., Mo o r e of Sciences U. S. A., 94: 8001–8005. T., Ma x S.I., Wa n g J., Hs i e h F., Di a t c h e n k o L., Ma r u s i n a Wo o H.A., Yi m S.H., Sh i n D.H., Ka n g D., Yu D.Y. and Rh e e S.G. K., Fa r m e r A.A., Ru b i n G.M., Ho n g L., St a p l e t o n M., 2010. Inactivation of peroxiredoxin I by phosphorylation So a r e s M.B., Bo n a l d o , M.F., Ca s a v a n t T.L., Sc h e e t z allows localized H2O2 accumulation for cell signaling. T.E., Br o w n s t e i n M.J., Us d i n T.B., To s h i y u k i S., Ca r n i n c i -Cell, 140: 517–528. P., Pr a n g e C., Ra h a S.S., Lo q u e l l a n o N.A., Pe t e r s G.J., Wo o d Z.A., Po o l e L.B. and Ka rp l u s P.A. 2003. Peroxiredoxin Ab r a m s o n R.D., Mu l l a h y S.J., Bo s a k S.A., McEw a n P.J., evolution and the regulation of hydrogen peroxide signal- Mc Ke r n a n K.J., Ma l e k J.A., Gu n a r a t n e P.H., Ri c h a r d s ing. – Science, 300: 650–653. S., Wo r l e y K.C., Ha l e S., Ga rc i a A.M., Ga y L.J., Hu l y k Wo o d Z.A., Sc h r ö d e r E., Ro b i n Ha rr i s J. and Po o l e L.B. 2003. S.W., Vi l l a l o n D.K., Mu z n y D.M., So d e r g r e n E.J., Lu X., Structure, mechanism and regulation of peroxiredoxins. – Gi b b s R.A., Fa h e y J., He l t o n E., Ke t t e m a n M., Ma d a n A., Trends in Biochemical Sciences, 28: 32–40. Ro d r i g u e s S., Sa n c h e z A., Wh i t i n g M., Ma d a n A., Yo u n g Ya n g Z. and Bi e l a w s k i J.P. 2000. Statistical methods for detecting A.C., Sh e vc h e n k o Y., Bo u f f a r d G.G., Bl a k e s l e y R.W., molecular adaptation. – Trends in Ecology and Evolution, To u c h m a n J.W., Gr e e n E.D., Di ck s o n M.C., Ro d r i g u e z 15: 496–503. A.C., Gr i m w o o d J., Sc h m u t z J., My e r s R.M., Bu t t e r f i e l d Y.S., Kr z y w i n s k i M.I., Sk a l s k a U., Sm a i l u s D.E., Sc h n e rc h Ye h H.-Y. and Kl e s i u s P.H. 2007. cDNA cloning, characteriza- A., Sc h e i n J.E., Jo n e s S.J. and Ma rr a M.A. 2002. Genera- tion, and expression analysis of channel catfish (Ictalurus tion and initial analysis of more than 15,000 full-length punctatus Rafinesque, 1818) Peroxiredoxin 6 gene.– Fish human and mouse cDNA sequences. – Proceedings of the Physiology and Biochemistry 33: 233-239.

Received: 18.03.2015 Accepted: 13.08.2015

317