|
|
||||||||
MOLECULAR, CELLULAR, AND DEVELOPMENTAL BIOLOGY |


* Department of Veterinary and Biomedical Sciences, University of Minnesota, St. Paul 55108; and
Department of Food Science and Human Nutrition, Michigan State University, East Lansing 48824
1 Corresponding author: reedx054{at}umn.edu
| ABSTRACT |
|---|
|
|
|---|
Key Words: turkey cardiac muscle expressed sequence tag round heart cardiomyopathy
| INTRODUCTION |
|---|
|
|
|---|
The etiology for RH disease is still not well characterized. Environmental factors such as altitude, diet, rapid growth, and viral infection can affect disease prevalence (Czarnecki, 1984; Julian et al., 1992). A DCM condition similar to spontaneous RH can be induced in turkeys by oral administration of high doses of furazolidone (Jankus et al., 1972). There is also a genetic predisposition to RH in that flocks with a high incidence of the disorder (greater than 70%) have been selectively bred (Hunsaker, 1971; Staley et al., 1981; Pierpont et al., 1985). These selected birds were also more susceptible to the effects of furazolidone than nonselected controls (Staley et al., 1978). A genetic approach to investigating this disease is clearly warranted.
The focus of turkey genomics has been on development of an integrated genetic linkage map with comparative alignment to the chicken genome (Reed et al., 2007a). Although this is an important accomplishment, genomic research in the turkey still lags behind other agriculturally valuable species, such as the chicken (Gallus gallus). Comparative genomics provides researchers with valuable resources based on the synteny between genomes. However, regardless of the degree of homology between their genomes, many transcripts may be unique to each species (Boardman et al., 2002). In addition, genes may exhibit duplicate loci, differential expression, and alternate splice variants at varying developmental stages (Chaves et al., 2003), making species-specific sequences invaluable. Several genes including phospholamban (PLN), cardiac troponin T (TNNT2; Biesiadecki et al., 2002), and titin (Hein et al., 1994) are suggested to influence RH in turkeys.
The identification and functional analysis of gene products is maximized by the development of species-specific expressed sequence tags (EST). Expressed sequence tags represent the transcribed portion of the genome and identify only genes that are expressed in the selected tissue. Species-specific sequencing and EST sequence identification is a crucial first step in downstream expression studies including microarray or quantitative PCR analysis, as well as interspecies comparisons for in silico mapping. Previous studies in the chicken included cDNA from pre-cardiac mesoendoderm cells not mature cardiac tissue (Afrakhte and Schultheiss, 2004).
This is the first study examining the mature cardiac gene transcripts of an avian species. The purpose of our study was to identify genes naturally expressed in cardiac tissue of turkeys (unknown RH etiology) and to catalog a suite of associated single nucleotide polymorphisms (SNP). Of primary interest is the identification of turkey-specific sequences for downstream applications in studying RH disease.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Library Construction
Total RNA was extracted from the tissues using the TRI reagent (Molecular Research Center, Cincinnati, OH) following the protocols of the manufacturer. Ribonucleic acid quality was analyzed by the Agilent Bioanalyzer 2100 (Agilent, Santa Clara, CA), to ensure RNA quality. PolyA+ selection was performed on the total RNA by the PolyATract mRNA Isolation System (Promega, Madison, WI) according to the protocols of the manufacturer.
The library was constructed from 4 to 5 µg of mRNA using CloneMiner cDNA library construction kit (In-vitrogen, Carlsbad, CA) following the protocols of the manufacturer. Superscript II RT was used to generate first strand cDNA with Biotin-attB2-Oligo (dT) primer. Second strand cDNA was synthesized from the first strand cDNA using T4 DNA polymerase, creating blunt-ended, double-stranded cDNA. The adaptor aatB1 was ligated to the 5' end of the cDNA. Excess primers, adaptors, and small cDNA fragments were removed by size-fractioned column chromatography. A recombination reaction was performed to ligate the aatB-flanked cDNA molecules to the aatP-donor vector pDONR222 (Invitrogen). This ligation was transformed into ElectroMAX DH10B T1 phage-resistant cells. The resulting cDNA library titer was estimated at 2.0 x 107 cfu/mL. Template generation for sequencing was performed by rolling circle amplification at the University of Washington High-Throughput Genomics Unit (Seattle).
Sequence Analysis
Single-pass 5' clone sequences were generated from the cardiac cDNA library at the High-Throughput Genomics Unit, University of Washington (Seattle). All sequences were trimmed of pDONR222 vector and low-quality tails then deposited in GenBank (EX717061-EX719678 and ES216654-ES217501). Mitochondrial DNA (mtDNA) sequences were identified by comparison with the chicken whole mtDNA sequence (GenBank NC_001323 [GenBank] ) and were removed from the project. In addition, sequences with reads of less than 50 bp were eliminated. The remaining sequences were assembled into contigs using Sequencher software (Gene Codes Corp., Ann Arbor, MI) using 98% match and 50-bp minimum overlap parameters. All contigs and singletons were compared with the chicken whole genome sequence (WGS; build 2.1) by BLASTN search using Ensembl Genome browser. Sequences that returned overlapping positions in the chicken WGS were then manually examined for sequence overlap and further contig assembly. The resulting contigs and singletons were then compared with the GenBank National Center for Biotechnology Information nonredundant (nr) database by BLASTN. Gene ontology annotations for the cardiac sequences were retrieved by BLASTN searches of the Ensembl chicken cDNA database. For this comparison, an E-value cutoff of 0.001, bit score cutoff of 100, and positives cutoff of 80 were used.
Assembled contigs were examined for the presence of SNP and insertion-deletions (INDELS). Initially, Sequencher was used to generate a list of all polymorphisms by setting the contig consensus sequence to inclusivity to generate a reference sequence, with subsequent comparison of the contig sequences to the reference by plurality. Each SNP identified by Sequencher was then individually verified by examination of the trace files.
| RESULTS |
|---|
|
|
|---|
All TC and singletons were compared with sequences in GenBank by BLASTN searches of the nr database using default parameters. Significant E-value scores (
1.0 x 10–4) were returned for 345 of the 356 TC and 791 of the 921 singletons. Among the TC, 256 hits were to previously annotated genes, 81 to unannotated EST sequences, and 8 hits to chicken bacterial artificial chromosome clone sequences. Eleven TC represent novel sequences with no significant hits in the National Center for Biotechnology Information nr database. Among the singletons, 641 hits were to previously annotated genes, 150 to unannotated EST sequences, and 130 were novel sequences.
The majority of significant matches among TC and singletons occurred with chicken EST sequences (1,056 of 1,136 queries, 92.9%). The 4 genes most highly represented in the cardiac library sequences are
cardiac actin (125 sequences), β globulin (85 sequences),
globin (80 sequences), and myosin light chain 2 (56 sequences). Other important cardiac-specific sequences also represented in the EST collection include C protein, LIM protein, phospholamban, troponin I, and troponin T.
The assembled distinct transcript set (1,277; 356 TC and 921 singletons) was compared with the chicken WGS by BLASTN. A total of 833 distinct transcripts consisting of 2,052 EST sequences had significant matches to putative homologous sites on 30 of the assembled chicken chromosomes (Table 1
). The number of matches ranged from 121 on GGA1 to 4 on GGA16. Of the aligned transcripts, roughly half (446) occurred on macrochromosomes (GGA1 to 9 and Z) with the remainder aligned to microchromosomes. These data provide in silico predictions of the chromosomal location for polymorphisms (SNP) identified in this project (see below). The remaining sequences had no significant match at the selected cutoff E-value, did not identify a unique homologous region within the chicken sequence, or were assigned to sequence contigs not yet assigned to chromosomes.
|
The 353 TC from the second assembly were then examined for sequence polymorphisms (SNP). Among the TC, a total of 310 polymorphisms were identified, with an average of 1 per 746 bp. The majority of polymorphisms (275 of 310) were single substitutions (SNP), but 35 INDELS were also identified. As expected, transition substitutions were the most commonly observed (195 of 275), followed by transversions (80 of 275). The number of individuals included in any SNP discovery effort will directly influence the frequency at which minor alleles may be observable for a given sequence depth and estimates of minor allele frequency are more accurate for TC with greater sequence depth. Because the cDNA library was constructed from cardiac tissue of 4 individuals (8 possible alleles), the minor allele frequency cutoff at which 1 individual in the library carries at least 1 copy of the minor allele is 0.125 (1/8). Minor allele frequencies for 263 polymorphisms occurred at frequencies of greater than 0.125. Sixty-four of the polymorphisms (56 SNP and 8 INDELS) in 39 TC were recurrent in that they were observed in more than 1 sequence in a contig (Table 2
). Examples of cardiac-specific candidate genes for RH that contained recurrent SNP were phospholamban, troponin I, and troponin T. When sequences were translated, 41 of the recurrent polymorphisms occurred within coding regions, with 34 representing synonymous substitutions and 7 being nonsynonymous substitutions. Twenty of the polymorphisms occurred in untranslated regions (5' or 3').
|
(4 SNP with 3 haplotypes) and ferritin (9 SNP with 4 haplotypes) suggests the cardiac EST sequences will provide a starting point for identification of additional sequence variants. Of particular interest is TNNT2, which is a candidate gene for cardiomyopathy (Biesiadecki et al., 2002; Osterziel and Perrot, 2005). Although not all TNNT2 transcripts are of sufficient length to include all of the polymorphic sites, we detect a minimum of 3 SNP haplotypes for this gene in the cardiac library (Table 3
alanine). In addition to these SNP, a 3-nucleotide insertion (single glutamic acid residue) was observed at the end of exon 7 in roughly half of the cDNA sequences. All sequences with this insertion also had the C nucleotide at the SNP locus within exon 7. The cardiac EST sequences contained both developmental isoforms of TNNT2 denoted by the presence or absence of exon 5; embryonic isoform contains exon 5, whereas adult isoform excludes it. The observed SNP variants were not specific for either the embryonic or adult isoforms.
|
| DISCUSSION |
|---|
|
|
|---|
Our goal was to create a cardiac-specific turkey cDNA library and to obtain turkey-specific sequences of important cardiac genes. Despite the high frequency of mtDNA sequences in this nonnormalized library, examination of nearly 4,000 EST sequences identified over 1,200 putative transcripts. In addition, a large number of EST sequences expressed in cardiac tissue but with unknown homology were also identified. These included 11 (3.1%) of the 354 contigs and 130 (14.8%) of the 874 singletons. Further work is needed to determine chromosomal location and function of these transcripts.
Single nucleotide polymorphisms are the most common form of polymorphisms between individuals in the genome (Emara and Kim, 2003), and over 300 such polymorphisms were cataloged among the assembled contigs with 64 for which the minor allele was observed to be present in more than one sequence. The SNP haplotypes were easily identified in a small set of sequence contigs containing multiple SNPs. These SNP provide a new source of allelic variants useful in candidate gene studies or genetic and comparative mapping, or both. For example, Paxton et al. (1999, 2005) identified polymorphic DNA sequences (sequence-characterized amplified regions; SCAR) that differed in frequency between RH+ and RH– birds. The genomic location and position of these markers relative to putative cardiac-related genes were determined via the turkey-chicken comparative map (Reed et al., 2007a). Similarly, 25 of the transcribed sequences from the cardiac library have predicted chromosomal locations within 1 Mb of these SCAR markers. These included transcripts of 3 genes (phospholamban, myosin-binding protein c, and creatine kinase M) identified by Reed et al. (2007b) as being associated with the RH SCAR.
Sequences corresponding to several important cardiac genes were identified in the turkey including troponin T, a candidate gene for cardiomyopathy. Both the chicken and turkey normally express 2 isoforms of cardiac troponin T (TNNT2), embryonic and adult, that differ in the presence of exon 5 (Cooper and Ordahl, 1985; Biesiadecki et al. 2002; Table 3
). In addition, an abnormal splice variant that excludes exon 8 (11 amino acids) is described as being associated with RH in the turkey (Biesiadecki and Jin, 2002; Biesiadecki et al., 2004). This low-molecular weight TNNT2 variant shows altered molecular conformation and binding affinity for troponin I and tropomyosin, as well as an increase in the calcium sensitivity of myosin ATPase (Biesiadecki et al., 2002, 2004). The low-molecular weight variant is also found in normal birds but is expressed at a significantly higher rate in RH birds. The cardiac library contained transcripts of both the embryonic and adult isoforms; however, low-molecular weight exon 8 variants were not observed. The TNNT2 consensus sequence of the turkey was 96.7% similar to that of the chicken with 1 amino acid difference (there is an additional amino acid residue in some turkey sequences). Variable presence of this additional amino acid was originally described by Biesiadecki et al. (2002) and attributed to allelic variation in exon 7.
The new turkey sequences identified in this study contain at least 1 isoform of 13 of the 20 DCM candidate genes identified in humans (Osterziel and Perrot, 2005). Six of these gene sequences [
actin, troponin T (isoforms 1 and 2), troponin I, phospholamban, and dystrophin] also contain cataloged SNP. These turkey-specific sequences are currently being used to design gene expression assays to examine differences in gene expression between RH+ and RH– birds.
| ACKNOWLEDGMENTS |
|---|
Received for publication December 20, 2007. Accepted for publication February 24, 2008.
| REFERENCES |
|---|
|
|
|---|
Biesiadecki, B. J., B. D. Elder, Z. Yu, and J. Jin. 2002. Cardiac troponin T variants produced by aberrant splicing of multiple exons in animals with high instances of dilated cardiomyopathy. J. Biol. Chem. 277:50275–50285.
Biesiadecki, B. J., and J. P. Jin. 2002. Exon skipping in cardiac troponin T of turkeys with inherited dilated cardiomyopathy. J. Biol. Chem. 277:18459–18468.
Biesiadecki, B. J., K. L. Schneider, Z. Yu, S. M. Chong, and J. Jin. 2004. An R111c polymorphism in wild turkey cardiac troponin I accompanying the dilated cardiomyopathy-related abnormal splicing variant of cardiac troponin T with potentially compensatory effects. J. Biol. Chem. 279:13825–13832.
Boardman, P. E., J. Sanz-Ezquerro, I. M. Overton, D. W. Burt, E. Bosch, W. T. Fong, C. Tickle, W. R. Brown, S. A. Wilson, and S. J. Hubbard. 2002. A comprehensive collection of chicken cDNAs. Curr. Biol. 12:1965–1969.[CrossRef][Web of Science][Medline]
Chaves, L. D., B. J. Ostroski, and K. M. Reed. 2003. Myosin light chain genes in the turkey (Meleagris gallopavo). Cytogenet. Genome Res. 102:340–346.[CrossRef][Web of Science][Medline]
Cooper, T. A., and C. P. Ordahl. 1985. A single cardiac troponin T gene generates embryonic and adult isoforms via developmentally regulated alternative splicing. J. Biol. Chem. 260:11140–11148.
Czarnecki, C. M. 1984. Cardiomyopathy in turkeys. Comp. Biochem. Physiol. A 77:591–598.
Emara, M. G., and H. Kim. 2003. Genetic markers and their application in poultry breeding. Poult. Sci. 82:952–957.
Frame, D. D., R. E. Buckner, and G. L. Anderson. 1999. Causes and control of spontaneous cardiomyopathy or round heart disease in Utah turkeys. Utah State University, online extension publication AG506. http://extension.usu.edu/files/publications/publication/AG_506.pdf Accessed Apr. 23, 2008.
Genao, A., S. Keirnan, U. Schmidt, M. Carles, and J. Gwathmey. 1996. Dilated cardiomyopathy in turkeys: An animal model for the study of human heart failure. Lab. Anim. Sci. 46:399–404.[Web of Science][Medline]
Hein, S., D. Scholz, N. Fujitani, H. Rennollet, T. Brand, A. Friedl, and J. Schaper. 1994. Altered expression of titin and contractile proteins in failing human myocardium. J. Mol. Cell. Cardiol. 26:1291–1306.[CrossRef][Web of Science][Medline]
Hunsaker, W. G. 1971. Round heart disease in four commercial strains of turkeys. Poult. Sci. 50:1720–1724.[Web of Science][Medline]
Jankus, E. F., G. R. Noren, and N. A. Staley. 1972. Furazolidone-induced cardiac dilatation in turkeys. Avian Dis. 16:958–961.[CrossRef][Web of Science][Medline]
Julian, R. J., S. M. Mirsalimi, L. G. Bagley, and E. J. Squires. 1992. Effect of hypoxia and diet on spontaneous turkey cardiomyopathy (round-heart disease). Avian Dis. 36:1043–1047.[CrossRef][Web of Science][Medline]
Magwood, S. E., and D. F. Bray. 1962. Syndrome condition of turkey poults characterized by enlarged and rounded hearts. Can. J. Comp. Med. 26:268–272.[Web of Science][Medline]
Osterziel, K. J., and A. Perrot. 2005. Dilated cardiomyopathy: More genes means more phenotypes. Eur. Heart J. 26:751–754.
Paxton, C., S. Hamblin, M. E. Pierpont, R. L. Park, E. Zepeda, R. N. Thwaits, and D. L. Kooyman. 1999. Markers associated with round heart disease in turkeys identified by RAPD and AFLP technologies. Page 50 in Proc. 80th Annu. Conf. Res. Work. Anim. Dis Iowa State Univ., Ames.
Paxton, C. N., M. E. Pierpont, and D. L. Kooyman. 2005. Identification of AFLP markers associated with round heart syndrome in turkeys. Int. J. Poult. Sci. 4:133–137.
Pierpont, M. E. M., D. Judd, B. Borgwardt, G. R. Noren, N. A. Staley, and S. Einzig. 1985. Carnitine alterations in spontaneous and drug-induced turkey congestive cardiomyopathy. Pediatr. Res. 19:415–420.[CrossRef][Web of Science][Medline]
Reed, K. M., L. D. Chaves, and K. M. Mendoza. 2007a. An integrated and comparative genetic map of the turkey genome. Cytogenet. Genome Res. 119:113–126.[CrossRef][Web of Science][Medline]
Reed, K. M., K. M. Mendoza, G. R. Hu, L. R. Sullivan, M. W. Grace, L. D. Chaves, and D. L. Kooyman. 2007b. Genomic analysis of genetic markers associated with inherited cardiomyopathy (round heart disease) in the turkey (Meleagris gallopavo). Anim. Genet. 38:211–217.[CrossRef][Web of Science][Medline]
Roberson, K. D. 2005. Induction of round heart disease in turkey poults by feed texture. Int. J. Poult. Sci. 4:177–181.
Sautter, J. H., J. A. Newman, S. H. Kleven, and C. T. Larsen. 1968. Pathogenesis of the roundheart syndrome in turkeys. Avian Dis. 12:614–628.[CrossRef][Web of Science][Medline]
Staley, N. A., G. R. Noren, C. M. Bandt, and H. L. Sharp. 1978. Furazolidone-induced cardiomyopathy in turkeys. Association with a relative
1-antitrypsin deficiency. Am. J. Pathol. 91:531–544.[Abstract]
Staley, N. A., G. R. Noren, and S. Einzig. 1981. Early alterations in the function of sarcoplasmic reticulum in a naturally occurring model of congestive cardiomyopathy. Cardiovasc. Res. 15:276–281.
USDA. 1997. Agriculture Facts book. US Dept. Agric., Washington, DC.
Wu, Y., A. Tobias, K. Bell, W. Barry, M. Helmes, K. Trombitas, R. Tucker, K. Campbell, and H. Granzier. 2004. Cellular and molecular mechanisms of systolic and diastolic dysfunction in an avian model of dilated cardiomyopathy. J. Mol. Cell. Cardiol. 37:111–119.[CrossRef][Web of Science][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |