Screening for candidate genes related with histological microstructure, meat quality and carcass characteristic in pig based on RNA-seq data.
Analyzing transcriptomes using next generation sequencing methods (RNA-seq) has been useful in various applications: estimation of gene expression; identification of novel genes or transcripts; and detection of genomic variants, gene fusion and RNA editing [1,2]. According to numerous studies, calling variants from RNA-seq data has many advantages. First, RNA sequencing is less cost-intensive compared to genome sequencing, and often, genomic variants can be identified from the existing RNA-seq data . Furthermore, the comparison of single nucleotide polymorphism (SNP) panels obtained from animals differing in the levels of given traits can be useful in searching for the genetic background of analyzed traits. To date, the most efficient and reliable methods of SNP calling from RNA-seq data have been validated [4,5].
In pigs, an RNA-seq approach enabled Jung et al  to detect nonsynonymous single nucleotide variations potentially related to enzyme activity in porcine liver. Moreover, the authors showed that selected SNPs were associated with pork quality and could be used in marker-assisted selection in future pig breeding programs. Transcriptome sequencing of liver tissue collected from four different pig breeds enabled the identification of non-synonymous SNPs within genes coding for enzymes that are critical during post-mortem metabolic process in muscle . Using RNA-seq data, Piorkowska et al  compared distributions of genetic variants detected in two breeds with extreme fatness parameters and identified genes involved in fatty acid metabolism and lipid storage, which can be potential genetic markers associated with meat quality in pigs. Conversely, Fisher et al  detected non-synonymous mutations with a possible influence on porcine reproduction traits using transcriptome data obtained after RNA sequence from oviduct and testis tissue.
It has been well-established that percentage number of fiber types in muscle determines metabolic properties and water holding capacity of muscle tissue, as well as marbling, intramuscular fat (IMF) level, meat color and texture [10-12]. In our previous study performed on Large White pigs, we identified genes differentially expressed in muscle contingent on histological microstructure . According to the differential expression and biological function, several genes were selected with possible effect on variation of fiber-type distribution. Thus, the aim of the present study was to identify genetic variants based on RNA-seq data, obtained via transcriptome sequencing of muscle tissue of pigs differing in muscle histological structure. The second aim was to select the most interesting variants, based on transcriptome-wide association study, and to verify the theirs effect on production traits and histological microstructure in a larger pig population.
MATERIALS AND METHODS
The SNPs in coding and untranslated regions (UTR) were identified based on RNA-seq data obtained previously by Ropka-Molik et al . Whole transcriptome sequencing was performed on longissimus lumborum samples collected from 11 Large White pigs that differed in the histological structure of this muscle. Pigs (unrelated sows) were maintained under the same housing and feeding conditions and slaughtered at final body weight of 100 kg ([+ or -]2.5 kg). Immediately after slaughter, muscle samples were collected and stored at -20[degrees]C in RNAlater solution (Ambion Inc., Austin, TX, USA) for RNA-seq application and at -80[degrees]C for histological analysis. The details of the animals used, the RNA-seq experiment and histological analysis were described in a previous study focused on differential expression analysis .
SNPs identification and association study based on RNA-seq data
Quality control of the raw RNA-seq data was performed using FastQC (v0.11.4; Illumina, San Diego, CA, USA). Adapters, poly-A sequences, and reads with a quality score lower than 20 and a fragment size reads lower than 36 bp were removed using the Flexbar tool (v.2.5GNU General Public License version 3.0 [GPLv3]). The cleaned reads were mapped to a Sus scrofa reference genome (Sscrofa10.2) using STAR aligner . Next, reads containing Ns (N cigar elements) were split (CIGAR string), and base quality score recalibration, INDEL realignment, duplicate removal, and identification of SNPs and INDELs (small insertions and deletions) were performed using Picard and GATK software packages . The filtering parameters were as follows: FisherStrand value>50, QualByDepth<2.0, RMS mapping quality<40, quality<40 and a SnpCluster filter. The genome annotation of identified polymorphisms was established using SnpEff 4.1b and Variant Effect Predictor (EMBL-European Bioinformatics Institute, Hinxton, Cambridge, UK) , and for downstream association analysis, only SNPs located within coding regions were selected (5'prime UTR variants; missense variants; synonymous variants; 3'prime UTR variants; noncoding transcript exon variants).
The association analysis for additive (fixed) effect of SNP was carried out for percentage and diameter of the I, IIA, and IIB types of fibers. The whole procedure was implemented in R-project. SNPs were considered significant if they met the following criteria: frequency (chisq p<0.05), a minimum of 9 observations, and a significant association analysis (p<0.05).
The differential expressed genes (DEGs) identified by RNAseq, and genes with SNPs (based on whole coding sequence analysis) significantly associated with percentage and/or diameter of I, IIA, and IIB types of fibers were mapped to quantitative trait loci (QTL) regions. The QTLs positions were taken from Animal Quantitative Trait Loci Database (Animal QTLdb) , which corresponds to the genome release used in this study (Sscrofta10.2, Ensemble database). We used R script  prepared by us to conduit the whole procedure. Based on RNA-seq results, 355 DEGs were mapped to QTLs regions. In contrast, based on the transcriptome-wide association study, 222 polymorphisms correlated with at least two histological traits and located within 149 genes were selected to the subsequent analysis.
Candidate genes--association study on a larger population
For an association study, four genes were selected based on high gene expression differences during RNA-seq analysis and their location within QTLs related to percentage of fibers I, IIA, and IIB (troponin T2 [TNNT2]; forkhead box O1 [FOXO1]). Two additional polymorphisms within two genes related to analyzed histological properties (transforming acidic coiled-coil-containing protein 2 [TACC2]; delta 4-desaturase, sphingolipid 1 [DEGS1]) were selected as significant based on the transcriptome-wide association study. For all genes, genotypes were assessed using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) method (TNNT2; TACC2; DEGS1) and polymerase chain reaction-artificially created restriction site (PCR-ACRS) method (FOXO1). The details of the analyses were presented in Table 1.
For the downstream association study, a larger pig population comprising 944 animals representing five breeds (Landrace n = 306, Large White n = 345, Pietrain n = 80, Duroc n = 61, native Pulawska breed n = 152) was used. All pigs (sows) were unrelated at least three generations back and were maintained under the same housing and feeding conditions according to Pig Test Station procedures. In the Test Station pigs are slaughtered, dissected and after carcass evaluation, meat is standard intended for consumption, thus research does not require the approval of Animal Experimentation committee. Pigs were fattened from 30 kg up to 100 ([+ or -]2.5) kg according to test procedures described previously by Ropka-Molik et al  and were slaughtered and dissected. For all animals, the following carcass characteristic traits were measured: weight of loin (kg), loin "eye" area (cm) and lean meat percentage (%). Furthermore, several meat quality traits were measured for selected animals: IMF content (the Soxhlet method)  and meat color (redness, a*; yellowness, b*; and lightness, L* parameters; Minolta Chroma Meter CR-310; Osaka, Japan) (699 pigs). The determination of meat color was performed using Chroma Meter CR-310 with a pulse xenon arc lamp D65 and 50 mm orifice. The a*, b*, and L* parameters were assessed 24 h post mortem on muscle tissue samples (longissimus lumborum; 150 to 200 g). The analysis was performed in two measurement points for each sample.
For selected animals, in total 293 (Landrace n = 110, Large White n = 70, Duroc n = 20, Pulawska pigs n = 93), longissimus muscle samples were collected immediately after slaughter and histological profile--percentage and diameter of the I, IIA, and IIB types of fibers--were estimated according to the procedure described previously by Ropka-Molik et al . The animals used to histological analysis were randomly selected from population included in association study. Tissue samples (frozen at -80[degrees]C) were cut on 10 [micro]m thick slides using a cryostat (Slee MEV, Mainz, Germany) and then fibers were detected using NADPH dehydrogenase (diaphorase) method. Myosin heavy chains (MHCs) were detected immunohistochemically with NCL-MHC antibodies (Leica, Novocasta, Leider Lane, Buffalo Grove, IL, USA) and NovoLink Polymer Detection Systems (Zalesie Gorne, Poland) (Leica, Novocasta, USA) . The obtained preparations were analyzed in terms of the percentage content of types of muscle fibers (I, IIA, IIB) and their diameter using the Nikon Eclipse 50i light microscope and MultiScan image analysis software (ver. 18.03; Sony, Tokyo, Japan). In each histological sample, the analysis covered a minimum of 300 fibers of each types. The percentage of fiber types was calculated based on 10 bundles measurements.
The statistical analysis for all selected genes was performed using the general linear model (GLM) procedure in SAS software (SASv. 8.02; Wadowice, Poland) and the model used was:
[Y.sub.ijk] = [mu]+[f.sub.i]+[h.sub.j]+[alpha]([x.sub.ij])+[e.sub.ijk]
Where, [Y.sub.ijk], the trait measured; [mu], the overall mean for the trait; [f.sub.i], the fixed effect of j genotype group of analyzed gene; [h.sub.j], the fixed effect of k breed; [alpha]([x.sub.ij]), covariate with weight of right half-carcass for selected slaughter traits; [e.sub.ijk], random error.
The correlation coefficients between histological features and carcass characteristics and meat quality traits were calculated using Pearson correlation method (SAS software, SASv. 8.02, Poland).
SNP identification and transcriptome-wide association analysis
Based on RNA-seq data, 65,535 SNPs were identified. After filtering according to the criteria outlined above, 51,590 SNPs were selected for use in the association study. The significant association of SNPs and percentage of I, IIA, and IIB fiber types were identified (p<0.05; 560, 251, and 837; respectively) (Table 2). We identified the panel of SNPs significantly related to the diameter of each fiber type. Few polymorphisms affected fiber I and IIB diameters compared to those that were associated with percentage of the same fiber types (Table 3). Our results show that the most numerous polymorphisms were 3'UTR variants, while the least numerous variants in each histological trait (both parentage and diameter of all fiber types) were 5'UTR variants and missense mutations (Tables 2, 3). The list of all SNPs significantly associated with histological traits is presented in Supplementary Table S1 available online (goo.gl/VAHcUs).
Several polymorphisms/genes influenced more than one analyzed histological trait, simultaneously affecting the percentage and the diameter of the fibers. The identified panel of genes was the basis for selecting significant SNPs for subsequent GLM analysis performed on a larger population. From such genes, we selected the TACC2, DEGS1, and FOXO1 genes, which were related to multiple histological properties.
Genes mapping to QTLs
Based on the pig QTL database (pigQTLdb), DEGs (n = 355) identified by the RNA-seq, were located within QTLs related with fiber percentage and diameter (78 and 65, respectively). The most interesting locus seems to be on chromosome 14, at which 23 genes located in regions affecting percentage of I and IIA fibers were identified. DEGs localized in QTLs associated with percentage of fiber type IIB were identified mainly on chromosome 2 (SSC2) and SSC10 (Supplementary Table S2). The most represented QTLs related with muscle fiber diameter were identified on SSC12 and SSC15; SSC2, SSC14 and SSC15; SSC2 and SSC14 (diameter of fibers I, IIA, and IIB; respectively) (Supplementary Table S3). On chromosome 14, 13 DEGs associated with diameter of both fibers IIB and IIA were mapped to four QTL regions (Supplementary Table S3).
After the transcriptome-wide association study, 149 genes were selected (222 SNPs) and mapped to QTL regions. Results showed that chromosome 14, on which four QTLs were identified (related with percentage of fiber type I: QTLs #2822 and #2824; and IIB fiber type: #2823 and #7044), was the most overrepresented (Table 4). Also, on SSC10, four QTLs were identified, but these were represented by low numbers of genes (#7012 and #7026, percentage of fiber type I; #7031 and #7034, percentage of fiber type IIA) (Table 4). In the case of diameter of fiber, the higher numbers of genes were detected on SSC14 and SSC3 within QTL regions associated with diameter of fiber types IIA and IIB (Table 5).
Association of selected SNPs with carcass characteristic and pork quality traits
Based on findings of other authors indicating relations of histological structure and meta quality and carcass traits [10-12], we also performed an association study on a large number of pigs. For this analysis four genes were selected: TNNT2; FOXO1; TACC2; DEGS1. These genes were previously identified as differentially expressed between muscle characterized by different histological properties; were localized within QTL regions associated with percentage of fiber types (FOXO1, IIB and IIA; TNNT2 and DEGS1, I and IIA), and contained SNPs potentially related with muscle microstructure. Moreover, subsequent GLM analysis showed significant relationships of FOXO1, DEGS1, and TNNT2 genes with loin 'eye' area, FOXO1 with loin weight, as well as FOXO1 and TACC2 with lean meat percentage. Furthermore, the significant (p<0.01) effect of polymorphisms within DEGS1, TACC2, and TNNT2 genes on IMF content was detected (Table 6). Only a small extent of SNPs was associated with pork color. TACC2 polymorphism showed an effect on value of a* color parameter and TNNT2 polymorphism influenced the b* color parameter (Table 6).
Association of selected SNPs with histological microstructure of muscle tissue
The association study performed on a large number of animals (293 samples of longissimus muscle) confirmed significant effect of FOXO1 and TACC2 genes on histological characteristics in pigs. FOXO1 was associated with diameter of fiber types IIA and IIB (p<0.01) (Table 7). The lowest diameter of both fiber types was detected in pigs with GG genotype, while the highest diameter was observed in heterozygote pigs. Conversely, a missense variant analyzed in the TACC2 locus showed a significant effect on diameter of fibers I and IIA and percentage of fiber type IIB. Our results showed that pigs with the AA genotype were characterized by the lowest diameter of fibers I and IIA, as well as the lowest percentage of fiber type IIB (Table 7).
Furthermore, our result confirmed also significant correlation between selected histological traits and carcass characteristic. The diameters of all fiber types were positively, significant correlated with weight of loin. Diameters of fibers I and IIA was also correlated with lean meat percentage, while diameter of fibers I and IIB with meat color parameter, yellowness (b*) (Table 8). Percentage of fiber types was only corelated with loin 'eye' area (percentage of fiber I and IIB) (Table 8). For the other analyzed traits, the significant correlations were not obtained.
Currently, much focus is placed on improving meat quality traits in pig breeding programs while maintaining satisfactory levels of meat and fatness characteristics . Research has been performed on refining pork quality through different methods including environmental conditions (maintenance, nutrition), animal transport, slaughter procedures, and genetics. One of the factors determining pork quality is fiber types distribution in muscle tissue [23,24], which can be affected by genetic background [25,26]. However, the genetic basis of muscle fiber structure has not been well understood.
The present study is a continuation of previous research concerning identification of genes determining muscle fiber structure in pigs . The differential expression analysis was performed on pigs representing one breed (Large White) to avoid an influence of breed factor on gene expression profile. The comparison of global gene expression profiles enabled us to detect the panel of genes that can play a role in formation of muscle fiber such as TNNT2, FOXO1, myosin XVIIIB, FERM domain containing 4B, and scavenger receptor class a member 5 . The SNP calling analysis is a next step facilitating the search for candidate genes and establishing the exact genetic basis of phenotypic traits. In our research, the wide-transcriptome association study, performed as a preliminary analysis, allowed us to select the most important genetic variants potentially related with histological profile in muscle, the effects of which were validated on a large pig population.
Mapping both DEGs and SNPs significantly correlated with muscle microstructure to the known QTLs showed that a high number of genes were localized on SSC14. The most over-represented were QTL regions associated with percentage of fiber type I (ID#2822 and #2824), IIB (ID#2823 and #7024), and diameter of fibers IIA (ID#7038) and IIB (ID#7018). These results were consistent with previous findings of Wimmers et al  (QTLs #2822; #2823; #2824) and Estelle et al  (QTLs #7018; #7024; #7038), who mapped QTLs associated with size and percentage of fibers on SSC14 and SSC15. These genome regions also span QTLs related with meat quality traits: pork color--redness and lightness , as well as fatness traits--abdominal fat weight and IMF content . This may indicate that identified genes can affect some production traits in pigs.
Genome location was one of the selection criteria of genes for the subsequent GLM analysis. The association study carried out on a larger group of animals confirmed the significant effect of scrutinized polymorphisms in FOXO1 and TACC2 genes on some histological characteristics. A prior study  indicated differential expression of the FOXO1 gene between pigs characterized by various percentages of fiber types I and IIB. The results obtained by transcriptome-wide association analysis, supplemented with information on genes localization in QTLs regions, were confirmed for both FOXO1 and TACC2. The most interesting gene seems to be TACC2 coding for transforming acidic coiled-coil protein 2. The RNAseq approach enabled us to detect several polymorphisms within the TACC2 gene. For the association study, the missense variant within TACC2 locus was selected (p.Arg2613Cys) due to the highly potential results on protein structure, as well as biological function.
Our results show that investigated SNP in TACC2 gene affected diameter of both fiber types I and IIA and percentage amount of fiber IIB. The same polymorphism was also significantly related with lean meat percentage as well as IMF content. In the cell cycle, TACC2 protein is involved in G2/M progression  by playing a key role in microtubule dynamics. It has been established that the TACC2 gene is a critical factor controlling mitotic spindle dynamics during the normal course of the mitosis process . To date, scarce research has investigated the involvement of the TACC2 gene in fiber type specification or muscle growth, even though this gene is highly expressed in skeletal and cardiac muscle. Our study suggested that transforming acidic coiled-coil protein 2 could play an important role during muscle fiber differentiation, resulting in modification of the number and size of muscle fibers. In turn, the differences in histological features are related to muscle mass and/or meat quality traits. The fiber composition in muscle is related with fatty acid and glycogen content due to the type of energy sources. The fiber type I is characterized by high fatty acid storage and low glycogen content, whereas the opposite trend is observed for the fiber type IIB. The observed missense variant changing arginine to cysteine in two different TACC2 isoforms (ENSSSCP00000029 474.1 :p.Arg795Cys; ENSSSCP00000025150.1:p.Arg2613Cys) may affect function of synthesized proteins. Our results confirmed previous findings that percentage amount of type fiber IIB is inversely related with IMF . According to the TACC2 missense variant, pigs with the AA genotype had significantly lower numbers of fiber type IIB and higher IMF content.
Conversely, our results indicated significant association between a FOXO1 mutation and diameter of fibers IIA and IIB and carcass characteristics such as weight of loin, loin 'eye' area and lean meat percentage. Furthermore, the significant correlation was observed between diameter of all fiber types and weight of loin. A previous study  showed differential expression of the FOXO1 gene depends on histological microstructure of muscle tissue. Moreover, our current research, which confirmed a significant relationship of a 3'UTR variant in the FOXO1 gene with fiber diameter, suggests that the analyzed SNP can affect FOXO1 transcript levels and histological profile as a result. Schiaffino and Mammucari  showed that FOXO proteins can regulate skeletal muscle growth via interaction with the IGF1-Akt/PKB pathway. Moreover, the research performed on transgenic mice confirmed that over-expression of the FOXO1 gene leads to muscle atrophy and increases transcript level of cathepsin L . Researchers demonstrated that skeletal muscle of mice overexpressing FOXO1 was characterized by decrease size of both fiber types I and II, likewise reducing the number of fiber type I. Subsequently, the up-regulation of the cathepsin L gene can indicate increasing protein degradation processes in muscle tissues. The FOXO1 gene has been considered as a main genetic factor closely associated with muscle fiber type specification and skeletal muscle differentiation . Our results confirmed that in pigs, the FOXO1 gene is related to muscle mass, as well as size of fiber types IIA and IIB, and should be investigated in the future for relation with important production traits.
Our results did not show the significant association of DEGS1 and TNNT2 genes with histological traits. However, polymorphisms identified in both genes affected loin 'eye' area and lean meat content in carcasses. This suggests that detected SNPs can be related with muscle growth processes but do not affect histological structure of muscle tissue at the same time. Furthermore, there is also the possibility of occurrence of another polymorphism within the DEGS1 and TNNT2 loci (in promoter or intron regions), which are associated with gene expression, and as a result, with cell size and fiber type distribution.
The use of RNA-seq data enabled us to identify a panel of SNPs within coding and UTRs regions with potential effects on histological microstructure. The genes detected by differential expression and the transcriptome-wide association analyses enabled us to select candidate genes potentially related to distribution of fiber types in porcine muscle tissue. These results can inform future studies evaluating the use of selected SNPs as genetic markers related to muscle histological profile and production traits in pig breeding.
CONFLICT OF INTEREST
We certify that there is no conflict of interest with any financial organization regarding the material discussed in the manuscript.
The study was supported by statutory activity of the National Research Institute of Animal Production (project no 01-014.1).
[1.] Wang Z, Gerstein M, Snyder M. RNA-Seq: a revolutionary tool for transcriptomics. Nat Rev Genet 2009;10:57-63.
[2.] Ramaswami G, Zhang R, Piskol R, et al. Identifying RNA editing sites using RNA sequencing data alone. Nat Methods 2013;10:128-32.
[3.] Piskol R, Ramaswami G, Li JB. Reliable identification of genomic variants from RNA-seq data. Am J Hum Genet 2013;93: 641-51.
[4.] Canovas A, Rincon G, Islas-Trejo A, Wickramasinghe S, Medrano JF. SNP discovery in the bovine milk transcriptome using RNA-Seq technology. Mamm Genome 2010;21:592-8.
[5.] Lopez-Maestre H, Brinza L, Marchet C, et al. SNP calling from RNA-seq data without a reference genome: identification quantification differential analysis and impact on the protein sequence. Nucleic Acids Res 2016;44:e148.
[6.] Jung WY, Kwon SG, Son M, et al. RNA-Seq approach for genetic improvement of meat quality in pig and evolutionary insight into the substrate specificity of animal carbonyl reductases. PLoS One 2012;7:e42198.
[7.] Lee YH, Cho ES, Kwon EJ, et al. Non-synonymous SNP in the ApoR gene associated with pork meat quality. Biosci Biotechnol Biochem 2011;75:2018-20.
[8.] Piorkowska K, Zukowski K, Tyra M, Ropka-Molik K. Detection of genetic variants different between Polish Landrace and Pulawska pigs by means of RNA-seq analysis. Anim Genet 2018;49:215-25.
[9.] Fischer D, Laiho A, Gyenesei A, Sironen A. Identification of reproduction-related gene polymorphisms using whole transcriptome sequencing in the Large White pig population. G3 (Bethesda) 2015;5:1351-60.
[10.] Joo ST, Kim GD, Hwang YH, Ryu YC. Control of fresh meat quality through manipulation of muscle fiber characteristics. Meat Sci 2013;95:828-36.
[11.] Bereta A, Tyra M, Ropka-Molik K, et al. Histological profile of the longissimus dorsi muscle in Polish Large White and Polish Landrace pigs and its effect on loin parameters and intramuscular fat content (IMF). Ann Anim Sci 2014;14:955-66.
[12.] Shen LY, Luo J, Lei HG, et al. Effects of muscle fiber type on glycolytic potential and meat quality traits in different Tibetan pig muscles and their association with glycolysis-related gene expression. Genet Mol Res 2015;14:14366-78.
[13.] Ropka-Molik K, Bereta A, Zukowski K, et al. Transcriptomic gene profiling of porcine muscle tissue depending on histological properties. Anim Sci J 2017;88:1178-88.
[14.] Dobin A, Davis CA, Schlesinger F, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 2013;29:15-21.
[15.] McKenna A, Hanna M, Banks E, et al. The genome analysis toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res 2010;20:1297-303.
[16.] Cingolani P, Platts A, Wang LL, et al. A program for annotating and predicting the effects of single nucleotide polymorphisms SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly 2012;6:80-92.
[17.] R Core Team. R: A language and environment for statistical computing [internet]. Vienna, Austria: R Foundation for Statistical Computing; 2017. Available from: https://www.R-project. org/
[18.] Hu ZL, Park CA, Reecy JM. Developmental progress and current status of the Animal QTLdb. Nucleic Acids Res 2016; 44:D827-33.
[19.] Ropka-Molik K, Dusik A, Piorkowska K, et al. Polymorphisms of the membrane-associated ring finger 4 ubiquitin protein ligase gene (MARCH4) and its relationship with porcine production traits. Livest Sci 2015;178:18-26.
[20.] Oczkowicz M, Tyra M, Ropka-Molik K, Mucha A, Zukowski K. Effect of IGF2 intron3-g. 3072G>A on intramuscular fat (IMF) content in pigs raised in Poland. Livest Sci 2012;149: 301-4.
[21.] Wojtysiak D, Kaczor U. Effect of g.2728G>A and g.3996 T>C polymorphisms at the leptin gene locus on microstructure and physicochemical properties of longissimus lumborum muscle of Polish Landrace pigs. Folia Biol 2011;59:77-82.
[22.] Zak G, Pieszka M. Improving pork quality through genetics and nutrition. Ann Anim Sci 2009;9:327-39.
[23.] Karlsson AH, Klont RE, Fernandez X. Skeletal muscle fibres as factors for pork quality. In: Quality of meat and fat in pigs as affected by genetics and nUTRition. Wageningen, The Netherlands: Wageningen Academic Publishers; 2000. pp. 47-67.
[24.] Wu F, Zuo JJ, Yu QP, et al. Effect of skeletal muscle fibers on porcine meat quality at different stages of growth. Genet Mol Res 2015;14:7873-82.
[25.] Nii M, Hayashi T, Mikawa S, et al. Quantitative trait loci mapping for meat quality and muscle fiber traits in a Japanese wild boarxLarge White intercross. J Anim Sci 2005;83:308-15.
[26.] Wimmers K, Fiedler I, Hardge T, et al. QTL for microstructural and biophysical muscle properties and body composition in pigs. BMC Genet 2006;7:15.
[27.] Estelle J, Gil F, Vazquez JM, et al. A quantitative trait locus genome scan for porcine muscle fiber traits reveals overdominance and epistasis. J Anim Sci 2008;86:3290-9.
[28.] de Koning DJ, Harlizius B, Rattink AP, et al. Detection and characterization of quantitative trait loci for meat quality traits in pigs. J Anim Sci 2001;79:2812-9.
[29.] Jiao S, Maltecca C, Gray KA, Cassady JP. Feed intake average daily gain feed efficiency and real-time ultrasound traits in Duroc pigs: II. Genomewide association. J Anim Sci 2014;92: 2846-60.
[30.] Takayama K, Horie-Inoue K, Suzuki T, et al. TACC2 is an androgen-responsive cell cycle regulator promoting androgen-mediated and castration-resistant growth of prostate cancer. Mol Endocrinol 2012;26:748-61.
[31.] Schiaffino S, Mammucari C. Regulation of skeletal muscle growth by the IGF1-Akt/PKB pathway: insights from genetic models. Skelet Muscle 2011;1:4.
[32.] Kamei Y, Miura S, Suzuki et al. Skeletal muscle FOXO1 (FKHR) transgenic mice have less skeletal muscle mass down-regulated Type I (slow twitch/red muscle) fiber genes and impaired glycemic control. J Biol Chem 2004;279:41114-23.
[33.] Xu M, Chen X, Chen D, Yu B, Huang Z. FoxO1: a novel insight into its molecular mechanisms in the regulation of skeletal muscle differentiation and fiber type specification. Oncotarget 2017;8:10662-74.
Katarzyna Ropka-Molik (1), *, Anna Bereta (2), Kacper Zukowski (2), Mirostaw Tyra (2), Katarzyna Piorkowska (1), Grzegorz Zak (2), and Maria Oczkowicz (1)
* Corresponding Author: Katarzyna Ropka-Molik
Tel: +48-666-081-208; Fax: +48-12-422-80-65, E-mail: firstname.lastname@example.org
(1) Department of Genomics and Animal Molecular Biology, National Research Institute of Animal Production, Krakowska 1, Balice 32-083, Poland
(2) Department of Animal Genetics and Breeding, National Research Institute of Animal Production, Krakowska 1, Balice 32-083, Poland
Submitted Sept 21, 2017; Revised Dec 13, 2017; Accepted Mar 5, 2018
Table 1. Detailed information about the analyzed genes and SNPs along with the methods used Gene Accession number SNP SNP variant TNNT2 ENSSSCG00000023031 c.294T>C splice region variant/ (rs330309370) synonymous variant FOXO1 ENSSSCG00000009370 c. *1395A>G 3' prime UTR variant (rs337841578) TACC2 ENSSSCG00000028063 c.7837C >T missense variant p.Arg2613Cys (rs80840610) DEGS1 ENSSSCG00000024484 c. *250T > C 3' prime UTR variant (rs344988843) Gene Accession number Primers Endonuclease TNNT2 ENSSSCG00000023031 F CACGCCTCCTTCTGTTCATC HpyCH4IV R CAGCTCCTCCTCCTCCTTCT FOXO1 ENSSSCG00000009370 F TCAGCAAGTTTTCCCTCTGTaCT HpyCH4III R CCCCAGCTTAACTTTCCAGA TACC2 ENSSSCG00000028063 F GCTGAGGTGTTCGTTGTGAA AluI R TGCACGATTGACGCTTTTAG DEGS1 ENSSSCG00000024484 F AGACGAGCCCCTCGTACAAT HpyCH4IV R GAACTCCCAGGCCTTTTCA Gene Accession number Obtained alleles (bp) TNNT2 ENSSSCG00000023031 T: 334 C: 255, 79 FOXO1 ENSSSCG00000009370 A: 154 G: 131, 23 TACC2 ENSSSCG00000028063 C: 200, 121, 56, 16 T: 200, 89, 56, 32, 16 DEGS1 ENSSSCG00000024484 T: 361 C: 211, 150 SNP, single nucleotide polymorphism; TNNT2, troponin T2; FOXO1, forkhead box O1; TACC2, transforming acidic coiled-coil-containing protein 2; DEGS1, delta 4-desaturase, sphingolipid 1. Table 2. Number of identified polymorphisms and selected genes significantly associated with percentage of all fiber types Fiber SNPs Gene (N) Genetic variants type I 560 630 5 prime UTR variant Synonymous variant Missense variant 3 prime UTR variant Non coding transcript exon variant, non coding transcript variant IIA 251 299 5 prime UTR variant Synonymous variant Missense variant 3 prime UTR variant Non coding transcript exon variant, non coding transcript variant IIB 837 921 5 prime UTR variant Synonymous variant Missense variant 3 prime UTR variant Non coding transcript exon variant, non coding transcript variant Fiber SNPs Genes type (N) I 19 MAFF; SCO2; MRTO4; SDHB; CD151; MRTO4 150 MYH7; COL5A1; SETX; TNNI1;ACE; MYH3; SLC15A4; SLC25A37;ATIC; CAST; CNOT11;CHD7 66 SLC16A5; MYO18B; FBXO7; DECR1; PINK1; SLC16A5; HEBP2 300 SLC17A5; MYO1E; TGFB2; DEGS1; SLC16A6; RPTOR; TACO1; BMP1; BMP6; MTR; TMX2; ILF2; GLO1 25 TXLNA; SSH2; MYH7; GPX3; RPIA;ACE; SGCA IIA 10 TMED7; UQCC3; GTF2F1 80 MYBPC1; PER1; ITGB7; BIN1; TRRAP; ITGB7; LIN7B; SUN2 31 TACC2; PROB1; MLH1; HCR; PISD; SERPINB1 127 CASP2; BMPER; MAFB; VIPR1; SLC9A3R1; COL4A2; PPP1R9B; CRP 3 TACC2; ZCCHC6 IIB 33 SLC2A12; MARC2;AQP3;ACO1; MYH13; GOT1; TACC2; MYLK2; 261 TACC2;APP; PLCD1; MYH13; TSN; GOT1; MYLK2; BRAF; SLC43A2; COL4A1; MARC2 91 TACC2; MARCHS; LTF; MTO1; CAST; FYCO1; LIFR 439 DEGS1; FOXO1; AKT3; FYN; RPTOR; SCARA5; BMPR2; SRFBP1; TAGLN2; TGFB1; GLO1 13 TACC2; SSH2; LAT; ARNT; KLRB1; TNFRSF21 N, number of identified genes; SNP, single nucleotide polymorphisms. Table 3. Number of identified polymorphisms and selected genes significantly associated with diameter of all fiber types Fiber SNPs Gene Genetic variants type (N) I 91 112 5 prime UTR variant Synonymous variant Missense variant 3 prime UTR variant Non coding transcript exon variant, non coding transcript variant IIA 139 162 5 prime UTR variant Synonymous variant Missense variant 3 prime UTR variant Non coding transcript exon variant, non coding transcript variant IIB 33 39 5 prime UTR variant Synonymous variant Missense variant 3 prime UTR variant Fiber SNPs Fiber type type (N) I 6 MYL9; TMEM258; MAP2K2 29 TACC2; GPX3; CA3; HP; NOB1; MYH7; PIK3C3;ATAT1 14 TACC2; FBXO7; BCAR1; HSF4; ELL2; TMEM5 38 ESR1; ERG; ITGB1; TACC2; MARCH6; TTL; CHD1L; HP; HYI; ATAT1 4 TACC2; SGCA IIA 5 CAPN3; PFDN4; MAFF; NOL1U; FGFR1OP2; INTS3 42 ACE; TACC2; ERBB2; NRBF2; HPS6; E2F6; SUN2; ILF2; RYR1; ENO1; AARS; 15 TACC2; SIRT3;ACAT2; FLAD1; RFX5; C4A; GSTM3 72 FOXO1; ITGB1; STX7; OAS2; CD82; HOXC6; PRR3; SLC29A1; STK38; RPTOR; 5 TACC2;ACE; SGCA; CD8A IIB 1 TMEM258 15 TRRAP; HP; HK1; SLC35B1; NXN;ADAM17; LUM 3 TACC2; BCAR1; HEATR5A 14 GOT1; OTULIN; GPAM; YBX3; SMUG1; IDH3A; ZSCAN31 N, number of identified genes; SNPs, single nucleotide polymorphisms. Table 4. Genes obtained after transcriptome-wide association study mapped to QTL regions related to percentage of fiber types I, IIA, and IIB I fiber type SSC QTL ID RGP1 1 2794 IKBKAP 2794 MARC2 10 7012 7026 C1orf115 7012 7026 DEGS1 7012 7026 RABIF 7012 7026 C9orf64 7012 7026 ANAPC7 14 2822 2824 PPTC7 2822 2824 AP1B1 2822 2824 MTR 2822 2824 ARV1 2822 2824 RAB4A 2822 2824 SUPV3L1 2822 2824 4022 EXOSC1 2822 2824 GOT1 2822 2824 IIA fiber type SSC QTL ID CCDC28A 1 7015 KCNQ5 7015 APIP 2 4025 RPS13 4025 7016 DHPS 4025 7016 MARC2 10 7034 C1orf115 7034 DEGS1 7034 RABIF 7034 C9orf64 7031 NRP1 7031 VPS36 11 7030 MAP3K14 12 7033 TUBG2 7033 NKIRAS2 7033 ZNF385C 7033 CNP 7033 TTC25 7033 IIB fiber type SSC QTL ID VPS36 11 7020 7023 7042 WRAP53 12 7021 7036 DHRS7C 7021 7036 ANAPC7 14 2823 7024 PPTC7 2823 7024 AP1B1 2823 7024 MTR 2823 7024 ARV1 2823 7024 RAB4A 2823 7024 SUPV3L1 2823 7024 4045 EXOSC1 2823 GOT1 2823 FAM45A 7044 TACC2 7044 QTL, quantitative trait loci. SSC, chromosome 2. Table 5. Genes obtained after transcriptome-wide association study mapped to QTL regions related to percentage of fiber types I, IIA, and IIB I fiber type SSC QTL ID SCRN2 12 2819 2821 UTP18 2819 2821 TOM1L1 2819 2821 WRAP53 2819 2821 FASTKD1 15 2832 HNRNPA3 2832 IIA fiber type SSC QTL ID SLF1 2 2797 CAST 2797 ANAPC7 14 7018 PPTC7 7018 AP1B1 7018 MTR 7018 ARV1 7018 RAB4A 7018 SUPV3L1 7018 FAM45A 2827 TACC2 2827 PAF1 2827 IIB fiber type SSC QTL ID RGP1 1 2795 IKBKAP 2795 TTL 3 7037 UXS1 7037 IL1R1 7037 MAP4K4 7037 NAGK 7037 PCYOX1 7037 EFEMP1 7037 MTIF2 7037 FEZ2 7037 CRIM1 7037 HDGF 4 2807 2810 PAQR6 2807 2810 RFX5 2807 2810 ANAPC7 14 7038 PPTC7 7038 AP1B1 7038 MTR 7038 ARV1 7038 RAB4A 7038 SUPV3L1 7038 FAM45A 2828 TACC2 2828 PAF1 2828 QTL, quantitative trait loci; SSC, chromosome 2. Table 6. Association of analyzed polymorphisms with carcass and meat quality traits FOXO1 Items Genotype LSM [+ or -] SE Weight of loin (kg) AA 6.07 [+ or -] 0.03 (a) AG 6.11 [+ or -] 0.04 (a) GG 5.93 [+ or -] 0.10 (b) Loin 'eye' area (cm) AA 52.6 [+ or -] 0.26 (b) AG 54.7 [+ or -] 0.54 (a) GG 53.5 [+ or -] 0.12 (ab) Lena meat percentage AA 61.4 [+ or -] 0.14 (b) AG 61.5 [+ or -] 0.23 (b) GG 62.7 [+ or -] 0.04 (a) IMF AA 1.40 [+ or -] 0.02 AG 1.36 [+ or -] 0.03 GG 1.38 [+ or -] 0.11 L* AA 54.15 [+ or -] 0.12 AG 54.03 [+ or -] 0.21 GG 54.01 [+ or -] 0.45 a* AA 16.89 [+ or -] 0.08 AG 17.02 [+ or -] 0.14 GG 17.04 [+ or -] 0.38 b* AA 3.24 [+ or -] 0.10 AG 3.41 [+ or -] 0.18 GG 3.58 [+ or -] 0.54 DEGS1 Items Genotype LSM [+ or -] SE Weight of loin (kg) TT 6.11 [+ or -] 0.04 CT 6.03 [+ or -] 0.03 CC 6.05 [+ or -] 0.03 Loin 'eye' area (cm) TT 53.3 [+ or -] 0.52 (a) CT 53.7 [+ or -] 0.46 (a) CC 52.1 [+ or -] 0.33 (b) Lena meat percentage TT 61.5 [+ or -] 0.27 CT 61.8 [+ or -] 0.18 CC 61.4 [+ or -] 0.20 IMF TT 1.36 [+ or -] 0.06 (b) CT 1.37 [+ or -] 0.03 (b) CC 1.46 [+ or -] 0.03 (a) L* TT 54.10 [+ or -] 0.22 CT 54.06 [+ or -] 0.14 CC 54.30 [+ or -] 0.20 a* TT 16.84 [+ or -] 0.14 CT 17.00 [+ or -] 0.10 CC 16.84 [+ or -] 0.14 b* TT 3.38 [+ or -] 0.17 CT 3.25 [+ or -] 0.12 CC 3.26 [+ or -] 0.19 TACC2 Items Genotype LSM [+ or -] SE Weight of loin (kg) CC 6.06 [+ or -] 0.08 CT 6.11 [+ or -] 0.03 TT 6.02 [+ or -] 0.03 Loin 'eye' area (cm) CC 52.8 [+ or -] 0.84 CT 54.0 [+ or -] 0.42 TT 52.8 [+ or -] 0.32 Lena meat percentage CC 60.6 [+ or -] 0.38 (b) CT 61.5 [+ or -] 0.19 (ab) TT 61.9 [+ or -] 0.17 (a) IMF CC 1.44 [+ or -] 0.09 (a) CT 1.39 [+ or -] 0.03 (AB) TT 1.35 [+ or -] 0.03 (B) L* CC 53.64 [+ or -] 0.51 CT 54.06 [+ or -] 0.16 TT 54.25 [+ or -] 0.13 a* CC 17.21 [+ or -] 0.06 (a) CT 17.07 [+ or -] 0.12 (ab) TT 16.80 [+ or -] 0.09 (b) b* CC 3.82 [+ or -] 0.40 CT 3.42 [+ or -] 0.14 TT 3.14 [+ or -] 0.12 TNNT2 Items Genotype LSM [+ or -] SE Weight of loin (kg) TT 6.07 [+ or -] 0.02 TC 6.11 [+ or -] 0.07 CC 5.96 [+ or -] 0.27 Loin 'eye' area (cm) TT 53.3 [+ or -] 0.25 (ab) TC 54.9 [+ or -] 0.87 (a) CC 51.7 [+ or -] 3.10 (b) Lena meat percentage TT 61.4 [+ or -] 0.13 TC 62.8 [+ or -] 0.35 CC 61.5 [+ or -] 0.78 IMF TT 1.35 [+ or -] 0.02 (C) TC 1.56 [+ or -] 0.08 (b) CC 2.00 [+ or -] 0.31 (a) L* TT 54.2 [+ or -] 0.11 TC 53.4 [+ or -] 0.30 CC 52.7 [+ or -] 0.96 a* TT 17.03 [+ or -] 0.08 TC 16.91 [+ or -] 0.20 CC 17.04 [+ or -] 0.73 b* TT 3.46 [+ or -] 0.10 (b) TC 3.23 [+ or -] 0.19 (b) CC 5.23 [+ or -] 1.37 (a) FOXO1, forkhead box O1; DEGS1, delta 4-desaturase, sphingolipid 1; TACC2, transforming acidic coiled-coil-containing protein 2; TNNT2, troponin T2; LSM, least squares mean; SE, standard error; IMF, intramuscular fat content; a*, b*, L*, meat colour parameters. (a,b) p <0.05; (A,B,C) p<0.01. Table 7. Association of selected polymorphisms with muscle microstructure traits FOXO1 Items Genotype LSM [+ or -] SE Percentage of fiber type I AA 12.3 [+ or -] 0.33 AG 12.3 [+ or -] 0.42 GG 12.2 [+ or -] 1.11 IIA AA 16.8 [+ or -] 0.45 AG 16.4 [+ or -] 0.47 GG 15.8 [+ or -] 0.84 IIB AA 71.3 [+ or -] 0.48 AG 71.3 [+ or -] 0.65 GG 71.9 [+ or -] 1.44 Diameter of fiber type I AA 54.4 [+ or -] 0.69 AG 56.2 [+ or -] 0.84 GG 53.3 [+ or -] 2.50 IIA AA 55.8 [+ or -] 0.63 (A) AG 56.9 [+ or -] 0.83 (A) GG 50.6 [+ or -] 1.78 (b) IIB AA 71.6 [+ or -] 0.82 (ab) AG 73.6 [+ or -] 1.07 (a) GG 67.0 [+ or -] 2.44 (b) DEGS1 Items Genotype LSM [+ or -] SE Percentage of fiber type I TT 12.6 [+ or -] 0.44 CT 12.5 [+ or -] 0.37 CC 12.0 [+ or -] 0.50 IIA TT 17.1 [+ or -] 0.53 CT 16.7 [+ or -] 0.54 CC 15.8 [+ or -] 0.45 IIB TT 70.3 [+ or -] 0.71 CT 71.2 [+ or -] 0.54 CC 72.1 [+ or -] 0.69 Diameter of fiber type I TT 55.2 [+ or -] 1.10 CT 55.4 [+ or -] 0.81 CC 55.1 [+ or -] 0.89 IIA TT 55.8 [+ or -] 1.00 CT 55.5 [+ or -] 0.73 CC 56.3 [+ or -] 0.78 IIB TT 71.3 [+ or -] 1.24 CT 72.2 [+ or -] 0.97 CC 71.3 [+ or -] 1.03 TACC2 Items Genotype LSM [+ or -] SE Percentage of fiber type I CC 12.4 [+ or -] 0.90 CT 12.1 [+ or -] 0.34 TT 12.7 [+ or -] 0.39 IIA CC 15.9 [+ or -] 0.97 CT 15.7 [+ or -] 0.41 TT 17.4 [+ or -] 0.50 IIB CC 67.5 [+ or -] 1.60 (b) CT 73.2 [+ or -] 0.55 (a) TT 71.6 [+ or -] 0.51 (a) Diameter of fiber type I CC 52.0 [+ or -] 1.35 (b) CT 55.4 [+ or -] 0.88 (a) TT 55.8 [+ or -] 0.76 (a) IIA CC 54.0 [+ or -] 1.73 (b) CT 56.2 [+ or -] 0.79 (a) TT 55.7 [+ or -] 0.64 (ab) IIB CC 67.5 [+ or -] 1.80 CT 73.2 [+ or -] 1.01 TT 71.6 [+ or -] 0.88 TNNT2 Items Genotype LSM [+ or -] SE Percentage of fiber type I TT 12.4 [+ or -] 0.26 TC 12.6 [+ or -] 0.98 CC 11.1 [+ or -] 2.62 IIA TT 16.3 [+ or -] 0.33 TC 16.6 [+ or -] 0.86 CC 12.4 [+ or -] 4.04 IIB TT 71.2 [+ or -] 0.38 TC 70.7 [+ or -] 1.34 CC 76.4 [+ or -] 6.60 Diameter of fiber type I TT 55.3 [+ or -] 0.55 TC 55.2 [+ or -] 1.64 CC 56.9 [+ or -] 8.01 IIA TT 56.0 [+ or -] 0.50 TC 54.3 [+ or -] 1.54 CC 55.9 [+ or -] 5.06 IIB TT 71.9 [+ or -] 0.66 TC 71.6 [+ or -] 1.97 CC 75.4 [+ or -] 5.60 FOXO1, forkhead box O1; DEGS1, delta 4-desaturase, sphingolipid 1; TACC2, transforming acidic coiled-coil-containing protein 2; TNNT2, troponin T2; LSM, least squares mean; SE, standard error. (a,b) p <0.05; (A,B) p<0.01. Table 8. The obtained significant correlation coefficients between percentage or diameter of fiber types and carcass characteristics Items Weight of Loin eye Lean meat b * loin(kg) area (cm) percentage Percentage of fiber type I ns -0.20 ** ns ns IIA ns ns ns ns IIB ns 0.12 * ns ns Diameter of fiber type I 0.12 * ns 0.15 * 0.20 ** IIA 0.21 ** ns -0.17 * ns IIB 0.15 ** 0.11 * ns 0.16 * (b)*, yellowness; ns, not significant. * p<0.05; ** p<0.01.