Polymorphisms in the FOXO gene are associated with growth traits in the Sanmen breeding population of the razor clam Sinonovacula constricta

Shumei Xie,Donghong Niu,b,Knyun Wei,Zhiguo Dong,Jile Li,c,*

aKey Laboratory of Exploration and Utilization of Aquatic Genetic Resources,Ministry of Education,Shanghai Ocean University,Shanghai,201306,China

bNational Demonstration Center for Experimental Fisheries Science Education,Shanghai Ocean University,Shanghai,201306,China

cShanghai Engineering Research Center of Aquaculture,Shanghai,201306,China

dCo-Innovation Center of Jiangsu Marine Bio-industry Technology,Lianyungang,222001,China

Keywords:Sinonovacula constricta Single-nucleotide polymorphism(SNP)FOXO Growth traits

ABSTRACT We identi fied the FOXO gene and five single-nucleotide polymorphisms(SNPs)in its coding region in the Sanmen breeding population of the razor clam Sinonovacula constricta.The ScFOXO gene encodes a protein of 625 amino acids in length that contains a conserved DNA-binding domain.Real-time polymerase chain reaction analysis showed that ScFOXO mRNA is widely expressed in adult tissues,with higher expression in the siphons and gills.Of the SNPs,c.879G＞C(Val293Val)and c.1725A＞G(Ser575Ser)are synonymous mutations,and c.543C＞T(Phe181Leu),c.848A＞G(Tyr283Cys),and c.1625G＞C(Gly542Phe)are non-synonymous mutations.The SNPs all show signi ficant associations with total body weight,shell length,shell width,and shell height(SH)in this population.Polymorphic parameter analysis showed that all SNPs except c.848A＞G(Tyr283Cys)exhibit complete linkage.The SNP c.1625G＞C(Gly542Phe)was used for genetic linkage mapping and demonstrates the strongest linkage(logarithm of the odds[LOD]value:41.46)with the marker 96616(37.02 cM)in the linkage group 9.The LOD value of this marker in relation to growth traits ranges from 1.04 to 1.53;only markers with a LOD value≥3 were considered to be linked with growth traits in the razor clam linkage map.In summary,ScFOXO may be a growth trait-related gene and may represent a marker for Sanmen breeding group selection.

1.Introduction

Single-nucleotide polymorphisms(SNPs)are abundant and randomly distributed in the genome,and can be used as markers to construct high-density genetic linkage maps.SNP markers have been proven to be directly related to phenotype through linkage with functional genes(Lindblad-Toh et al.,2000).Most SNPs are harmless,but some mutations in functional genes can be phenotypic,pathogenic,or even deadly as a result of changes in amino acid coding or expression level(Ferrer-Costa,Orozco,&de la Cruz,2002;Syvanen,2001;Yates&Sternberg,2013).

Growth traits are classic polygenic traits,reflecting the combined in fluence of multiple genetic factors.However,in the domestic dog,sequence variation of the insulin growth factor 1(IGF-1)gene has been shown to be a major contributor to body size(Sutter et al.,2007).IGF-1 is also known to in fluence both cell size and cell number in mice and humans(Agrogiannis,Sifakis,Patsouris,&Konstantinidou,2014;Lupu,Terwilliger,Lee,Segre,&Efstratiadis,2001;Randhawa&Cohen,2005).For species of fish and shell fish of economic importance in which the genes that determine growth traits have not yet been discovered,research on the associations between growth-related genes and growth traits are of great signi ficance for marker-assisted group selection breeding(Hershberger,2006).

FOXO is a conserved transcription factor negatively regulated by the insulin/IGF-1 signaling pathway that belongs to Class O(“Other”)of the FOX superfamily.There are four FOXO family members in vertebrates,two FOXO orthologs in the sea squirt Ciona intestinalis and sea urchin Actiniaria,and only one FOXO ortholog in Drosophila melanogaster and Caenorhabditis elegans(Huang and Tindall.2007;Wang et al.,2013).The FOXO family exhibits a high degree of sequence conservation within the DNA-binding domain and three protein kinase B(PKB)-phosphorylation sites(RXRXX[S/T])(Huang and Tindall.2007).Also,a five-amino acid insertion(SNSSA)betweenα-helices 2 and 3 of the DNA-binding domain is the difference between the FOXO subset and other FOX-related family members(Puig,Marr,Ruhf,&Tjian,2003).FOXO is inactivated after phosphorylation by the insulin-activated PKB,where upon it binds with 14-3-3 proteins(Brunet et al.,2001).Critical roles for FOXO proteins in increased resistance to various cellular stresses and growth limitation have been reported in D.melanogaster,C.elegans,and mammals(Jünger et al.,2013;Mattila,Bremer,Ahonen,Kostiainen,&Puig,2009,Martin A).To the best of our knowledge,there are no published reports on the molluscan FOXO,except for the deduced protein sequences of FOXO in Crassostrea gigas(XP_011414359.1),Crassostrea hongkongensis(KU052832.1),Mizuhopecten yessoensis(XM_021521691.1),and Octopus bimaculoides(XM_014933417.1),which are available from the National Center for Biotechnology Information database.Here,we provide evidence on FOXO in the razor clam Sinonovacula constricta.

The razor clam Sinonovacula constricta is a nutritious shell fish of economic importance in China that is also distributed in Japan,Vietnam,and Korea.The razor clam is favored by Chinese farmers;however,improved parent clams are needed to increase production.Sanmen County,on the northeast coast of Taizhou City,Zhejiang Province,China,is an important base for the breeding and cultivation of high-quality razor clams.A high-density genetic linkage map was constructed using SNP sites in the razor clam,which indicated that five linkage groups and 16 quantitative trait loci(QTLs)contribute to growth traits(Niu et al.,2017).However,de ficient genomic resources have increased the dif ficulty of finding the determining genes,despite the fact that the genetic distance in the razor clam linkage map averages just 0.32cM(Niu et al.,2017).

In this study,we used reverse genetics to locate the functional gene FOXO in the linkage map and to examine its linkage with SNP markers in QTLs for growth-related traits.We characterized FOXO in the razor clam(ScFOXO),identi fied SNPs in its coding region,and genotyped the mapping family(Niu et al.,2017)and the Sanmen breeding population using Sanger sequencing.JoinMap®4.0(Maliepaard,Jansen,&Van Ooijen,1997;Ooijen&Voorrips,2002)was used to locate ScFOXO in the razor clam linkage map according to linkage inheritance,where a logarithm of the odds(LOD)score exceeding 3 is considered suf ficient evidence of linkage(Jones,2000;Ott,2011).In addition,the correlation between SNPs and growth traits was veri fied in the Sanmen breeding population and mapping family.

2.Materials and methods

2.1.Sample collection

The DNA samples of the mapping family were stored in our lab(Niu et al.,2017).In total,200 individuals were randomly collected from a population hatched from the same batch and housed in the same outdoor nursery pond at Donghang Farm,Sanmen County,Taizhou City,Zhejiang Province,China.Growth traits comprising shell width(SW),shell length(SL),and shell height(SH)were measured using a vernier caliper,and total body weight(TW)was measured using an electronic balance.Mantle tissues were dissected and stored in 75%ethanol at-20°C until DNA extraction.Also,the siphon tissues of six smaller and six larger individuals were harvested and stored in RNA Storage Liquid(Tiangen Bio Inc.,Beijing,China)in three 1.5-mL tubes.

Healthy 1-year-old razor clams of a similar size were collected and maintained in 16‰saline solution with circulating aeration and food for 1 week.Subsequently,seven tissues,comprising the foot,mantle,siphon,gill,digestive gland,gonad,and hemolymph tissues of nine individuals were obtained,frozen immediately in liquid nitrogen,and kept at-80°C for RNA extraction.

2.2.DNA and RNA isolation and first-strand cDNA synthesis

ATIANamp Marine Animals DNA Kit(Tiangen)was used for DNA extraction according to the manufacturer's instructions.DNA integrity was checked by electrophoresis on a 1%agarose gel.Total RNA was extracted using an RNAsimple Total RNA Kit(Tiangen)according to the manufacturer's instructions.First-strand cDNA was generated using a PrimeScript™RT Reagent Kit(Takara Bio Inc.,Kusatsu,Japan)from 1μg of total RNA according to the manufacturer's protocol.

2.3.Sequence veri fication,genotyping,and location

A sequence alignment of the FOXO gene was obtained from the transcriptome library(Niu et al.,2016a),and the coding sequences were con firmed in both directions using the F1 and R1 primers(Table 1)via the Sanger method.We also identi fied SNPs in the coding region of the ScFOXO gene by amplifying the first-strand cDNA mix obtained from the six smaller and six larger individuals from the population group.We aligned the FOXO sequence against the simpli fied genome library(data not published),and designed the F2 and R2 primers(Table 1)to amplify the partial genome sequence of ScFOXO for genotyping the population group.Polymerase chain reaction(PCR)was performed using the following parameters:2min at 94°C;35 cycles of 30 s at 94°C,30 s at 50°C,and30 sat 72°C;and a finalextension stepfor 10 minat 72°C.SNPs were detected by direct sequencing using the Sanger method and genotyped using the software BioEdit 7.0.5(Tippmann,2004).

The F3 and R3 primers(Table 1)were designed to amplify and genotype one of the SNP sites in the map family(Niu et al.,2016a),and JoinMap®4.0 was used to group and locate the ScFOXO gene(Maliepaard et al.,1997;Ooijen&Voorrips,2002).

2.4.Quantitative analysis of ScFOXO mRNA expression

Quantitative real-time PCR(qRT-PCR)was performed using a Bio-Rad CFX96 Real-Time PCR Detection System(Bio-Rad Laboratories,Inc.,Hercules,CA,USA)and a SYBR®Rescript qRT-PCR Kit(Takara)to analyze the mRNA expression of ScFOXO.The primer pairs F4-R4 and 18sF-18sR(Table 1)were used to amplify the ScFOXO and 18S rRNA genes,respectively(Niu et al.,2016b).The ampli fication ef ficiencies of ScFOXO and 18S rRNA were 105%and 99%,respectively,based on Real-Time PCR Miner analysis.The fluorescence intensities of the tissue products were assessed in triplicate,and the reactions contained the following components:10μL of 2×SYBR®Premix Ex Taq™ II(Takara);6.8μL of nucleasefree water;50 ng of cDNA;and 0.8μL of each gene-speci fic primer(10μM).Three technical replicates were conducted for each sample.Gene expression levels in different tissues were measured by normalizing the cycle-threshold(CT)values of ScFOXO to those of 18S rRNA via the relative quanti fication method calculated according to the equation 2-ΔΔCT.Variation analysis was performed using one-way analysis of variance in SPSS 18.0(West,2009),and a P value of＜0.05 was considered signi ficant.

2.5.Bioinformatics and statistical analyses

An open reading frame finder(http://www.ncbi.nlm.nih.gov/gorf/gorf.html)was used to obtain the amino acid sequences.Mature domains were identi fied using SMART(http://smart.embl-heidelberg.de).The amino acid sequences of human(Homo sapiens), fly(D.melanogaster),oyster(C.gigas),and razor clam(S.constricta)FOXO orthologs were aligned using BioEdit 7.0.5(Tippmann,2004).A phylogenetic tree was constructed using the neighbor-joining method in MEGA 7.0(Tamura et al.,2011).

Table 1 List of the primers used in this study.

JoinMap®4.0 was used to locate the ScFOXO gene in the razor clam genetic linkage map(Niu et al.,2017;Ooijen&Voorrips,2002;Ott,2011).Genetic parameters including observed heterozygosity,expected heterozygosity,polymorphism information content(PIC),effective allele number,and Hardy-Weinberg equilibrium were calculated using Popgene32(Yeh,1998).The online software SHEsis(http://analysis.bio-x.cn/myAnalysis.php)was used to analyze the linkage disequilibrium(LD)of different SNP sites(Shi&He,2006).Associations between SNP genotypes and the measured traits(SL,SH,SW,and TW)were examined using a general linear model(GLM;Duncan's method)in SPSS 18.0(West,2009).A P value of＜0.05 was considered statistically signi ficant.

3.Results

3.1.Sequence analysis and expression pro filing of ScFOXO in adult tissues

The deduced ScFOXO protein contained 625 amino acid residues encodedby1878nucleotides(GenBankaccession number:MG999524).Structural analysis of the ScFOXO protein showed high conservation in the DNA-binding domain,and DNA-binding sites were predicted at the following locations:105Tyr(Y),106Phe(F),120Asn(N),123Arg(R),124His(H),and 142Lys(K).Also,three PKB phosphorylationsites(RXRXX[S/T])wereconserved in ScFOXO,but the speci fic five-amino acid SNSSA insertionof theFOXO subsetwas changed to NTSSA(Fig.1).Phylogenetic analysis(Fig.2)showed that ScFOXO is closer to C.gigas FOXO(CgFOXO)and forms a branch,whereas in vertebrates,FOXO3,FOXO6,FOXO4,and FOXO1 forms a branch.ScFOXO mRNA was widely expressed in adult tissues,with higher expression in the siphons,followed by the gills,foot,digestive gland,and other tissues(Fig.3).

3.2.Identi fication of polymorphisms in the ScFOXO gene

Five SNPs were identi fied in the coding region of the ScFOXO gene by direct sequencing,and named based on locus-mutation type and amino acid alterations(Fig.4).Of them,c.879G＞C(Val293Val)and c.1725A＞G(Ser575Ser)were synonymous mutations,and c.543C＞T(Phe181Leu),c.848A＞G(Tyr283Cys),and c.1625G＞C(Gly542Phe)were non-synonymous mutations.

3.3.Polymorphic parameters of single-nucleotide polymorphism sites and their associations with growth traits

In total,152 individuals from the population group were successfully ampli fied and sequenced,but we failed to genotype one of the individuals at the c.848A＞G(Tyr283Cys)site.Analysisof polymorphic parameters(Table 2)indicated that the population obeyed the Hardy-Weinberg equilibrium(P＞0.05);whereas only c.848A＞G(Tyr283Cys)was a low polymorphic site(PIC＜0.25),the other sites were moderately polymorphic(0.25＜PIC ＜0.5).LD analysis showed complete linkage in all SNPs except c.848A＞G(Tyr283Cys)(Table 3).The potential associations between the genotypes and growth traits are shown in Table 4;all five SNPs showed signi ficant correlations with the measured growth traits.In particular,individuals with complete linkage sites of c.543C＞T,c.879G＞C,c.1625G＞C,and c.1725A＞G of the CC-GG-GG-AA type were signi ficantly(P＜0.05)bigger and heavier than those of the TC-GC-GC-GAandTT-CC-CC-GGtypes.Inaddition,AG-type c.848A＞G(Tyr283Cys)individuals tended to be of greater size and weight(P＜0.05)than GG-type individuals.Also,in the mapping family,individuals with the AA-type c.1725A＞G(Ser575Ser)site were bigger and heavier(P＜0.05)than those with the GG-type site(Table 5).

3.4.The location of ScFOXO in the genetic linkage map of Sinonovacula constricta

ScFOXO was mapped to linkage group 9(LG9)of the razor clam genetic linkage map(Niu et al.,2017)based on the genotype of the SNP site c.1625G＞C(Gly542Phe).The marker that exhibited the strongest link(LOD:41.46)to ScFOXO was 96616(LG9;37.02 cM),with the markers demonstrating the following most probable order:marker 35543-marker 39621-marker 96616-ScFOXO-marker 58022-marker 25778-marker 24795.The LOD value of marker 96616 in contribution to TW was 1.31,in contribution to SH was 1.04,in contribution to SL was 1.53,and in contribution to SW was 1.32.

4.Discussion

Most traits of economic importance in fish and shell fish are quantitative in nature and determined by multiple genes as con firmed by genome-wide QTL mapping(Guo,Li,Wang,&Kong,2012;Li&He,2014;Niu et al.,2017;Ren et al.,2016;Wang,Li,&Zhang,2016).Construction of a razor clam genetic linkage map by speci fic-locus ampli fied fragment sequencing identi fied 16 QTLs related to growth traits(Niu et al.,2017).Because the determining genes have not been discovered,the location of growth-related genes in the linkage map and identi fication of SNPs signi ficantly associated with growth traits are important for marker-assisted group selection breeding.In this study,we found that five SNPs in the ScFOXO gene were signi ficantly associated with growth traits in the Sanmen breeding population and mapping family(Niu et al.,2017),that ScFOXO was located in the growth-trait linkage group LG9,and that ScFOXO exhibited the strongest linkage with the marker 96616(37.02cM)(Niu et al.,2017).ScFOXO is the first functional gene to be located in the razor clam genetic linkage map,and may represent a candidate marker for group selection breeding in Sanmen County.

Fig.1.Multiple alignment of the amino acid sequences of FOXO3 in Homo sapiens and FOXO in Drosophila melanogaster,Crassostrea gigas,and Sinonovacula constricta.Fully conserved sequences are shaded in gray,the DNA-binding sites(105Tyr[Y],106Phe[F],120Asn[N],123Arg[R],124His[H],and 142Lys[K])are indicated with asterisks,the fiveamino acid insertion(NTSSA)is underlined,and the conserved protein kinase B-phosphorylation motifs(RXRXX[S/T])are boxed.

Three PKB-phosphorylation motifs (RXRXX [S/T])were conserved in the DNA-binding domain of ScFOXO(Huang and Tindall.2007).The five-amino acid insertion SNSSA,which is the difference between the FOXO subset and other FOX-related family members,was changed to NTSSA(Puig et al.,2003).Also,phylogenetic analyses showed that ScFOXO clustered with CgFOXO,providing evidence that FOXO in the razor clam belongs to the FOXO subset.

FOXO is a conserved transcription factor controlled both downstream target genes responsible for growth and upstream feedback targets in the insulin signaling pathway(Puig et al.,2003).The wide expression of ScFOXO mRNA in the tissues of adult razor clams suggests that the gene has multiple functions in cellular homeostasis and metabolism.FOXOs in mammals and worms are also broadly expressed in several tissue types,and are critical effectors of cellular homeostasis and metabolism(Kim&Webb,2017).In the razor clam,the siphons and gills are the primary organs that interact with the environment.Also,the siphons,gills,and foot possess rhythmic movement ability,and the higher expression of ScFOXO in these organs may reflect their potential function in stress responses,particularly in the response to enemy invasion and oxygen resistance(Brunet et al.,2004;Kops et al.,2002).Under stress conditions,FOXOs activate growth-inhibitory genes and restrict the growth and differentiation of cells(Harvey et al.,2008;Nowak,Gupta,&Stocker,2018).

Fig.2.Phylogenetic tree of the FOXO family.ScFOXO is indicated with a solid black triangle.The phylogenetic tree was constructed via a ClustalW-generated multiple amino-acid sequence alignment using the neighbor-joining method in the MEGA7 package.The topologic stability of the trees was evaluated using 10,000 bootstrap replications,and the bootstrap values(%)are indicated by the numbers at the nodes.

Fig.3.Relative expression levels of ScFOXO in adult tissues.The y-axis showsΔΔCT values normalized against the endogenous control(18S rRNA).Gene expression levels in different tissues are expressed relative to that in the hemolymph.Different letters indicate statistical signi ficance(P＜0.05).

Five SNPs were detected in an approximately 1.2-kb sequence of ScFOXO,which is a signi ficantly higher frequency than that identi fied in a previous study of one SNP in 1 kb(Vignal,Milan,SanCristobal,&Eggen,2002),implying that the frequency of SNPs is high in the ScFOXO gene.Also,in an analysis of polymorphic parameters in the group,only the c.848A＞G(Tyr283Cys)site showed low polymorphism(PIC＜0.25),whereas the other four sites were moderately polymorphic(0.25＜PIC＜0.5)and exhibited completed linkage.SNPs are typical markers of two alleles with the highest polymorphism,reflected in a PIC of 0.5(Hubert,Bussey,Higgins,Curtis,&Bowman,2009).All five SNPs were true SNPs in this population.Of them,three were non-synonymous mutations and switched the aromatic amino acid Phe to the non-polar fatty acid amino acid Leu in c.543C＞T(Phe181Leu),the aromatic amino acid Tyr to the uncharged amino acid Cys in c.848A＞G(Tyr283-Cys),and the non-polar fatty acid amino acid Gly to the aromatic amino acid Phe in c.1625G＞C(Gly542Phe).Non-synonymous substitutions alter the amino acid composition of a protein,leading to structural changes and variations in protein interactions and functions(Moen et al.,2008;Yates&Sternberg,2013).A single Glu to Val mutation in theβ-globin chain results in sickle cell anemia,whereas epidermal growth factor receptor mutations in non-small cell lung cancer are correlated with the clinical response to therapy(Paez et al.,2004).Analysis of the associations between genotypes and growth traits indicated that razor clams with 181Phe and 542Gly or 283Tyr in the ScFOXO protein tended to be bigger and heavier(P＜0.05),possiblyas a resultof altered protein interactions or DNA-binding af finities relative to wild-type ScFOXO protein(Yates&Sternberg,2013).In fact,except for c848A＞G,the SNPs showed the same association results with growth traits as if they are completely linked.The association of these loci with growth may be because they are closely linked with a locus affecting growth(Sonesson&Meuwissen,2009).

Fig.4.Sequencing maps of five single-nucleotide polymorphisms in the FOXO gene from the mixed cDNA of Sinonovacula constricta.Single-nucleotide polymorphisms(SNPs)in the ScFOXO gene were identi fied by direct sequencing of the products ampli fied from the first-strand cDNA mix of six smaller and six larger individuals from the Sanmen breeding population group.Five SNPs were identi fied,which were named based on locus-mutation type and amino acid alterations.

ScFOXO was mapped to LG9 of the razor clam genetic linkage map,and exhibited the strongest linkage(LOD:41.46)to marker 96616(37.02cM).The LOD scores tended to increase,and the map distances(recombination rates)between loci became smaller(Jones,2000;Ott,2011),indicating that ScFOXO is inherited with marker 96616.The LOD value of marker 96616 related to the contribution toTW ranged from 1.04 to 1.53(Niu et al.,2017).There are five QTLs located in LG9 when the LOD score threshold is set at 3(Niu et al.,2017),indicating that ScFOXO is not the major gene responsible for growth traits in the map family.However,the associations between ScFOXO and growth traits were signi ficant in the mapping family and the Sanmen breeding population.Also,it was clear that the genetic map constructed using an F1 cross population was imperfect in terms of its completeness and accuracy.Thus,although ScFOXO is not a major gene underlying growthtraits,it is a growth trait-related gene,and can be used as a marker for Sanmen breeding group selection.

Table 3 Linkage disequilibrium analysis of five single-nucleotide polymorphism sites in the FOXO gene of Sinonovacula constricta.

Table 4 Association analysis of five single-nucleotide polymorphism sites associated with growth traits in the FOXO gene of Sinonovacula constricta.

Table 5 Association analysis of c.1725A＞G(Ser575Ser)sites in the FOXO gene of Sinonovacula constricta associated with the growth traits of the mapping family.

In summary,we have provided the first report of the FOXO gene in the razor clam.We identi fied SNPs in the coding region of the gene and analyzed their association with growth traits of the Sanmen breeding population.Furthermore,we successfully located ScFOXO in the razor clam genetic linkage map.The findings of this study suggest that ScFOXO represents a marker for Sanmen group selection breeding.

Acknowledgments

This work was supported by the National Natural Science Foundation of China[grant number 31472278],the National High Technology Research and Development Program of China(863 Program)[grant number 2012AA10A400-3],and the Shanghai Universities Knowledge Service Platform[grant number ZF1206].