Physiol. Genomics Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Physiol. Genomics 29: 267-279, 2007. First published January 30, 2007; doi:10.1152/physiolgenomics.00178.2006 Free Article
1094-8341/07 $8.00
This Article
Free upon publication Free Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Tables
Right arrow All Versions of this Article:
29/3/267    most recent
00178.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ponsuksili, S.
Right arrow Articles by Wimmers, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ponsuksili, S.
Right arrow Articles by Wimmers, K.
Received 13 August 2006; accepted in final form 25 January 2007.
Physiological Genomics 29:267-279 (2007)
1094-8341/07 $8.00 © 2007 American Physiological Society

Pre- and postnatal hepatic gene expression profiles of two pig breeds differing in body composition: insight into pathways of metabolic regulation

Siriluck Ponsuksili1, Eduard Murani2, Christina Walz1, Manfred Schwerin1 and Klaus Wimmers2

1 Research Group Functional Genomics
2 Research Unit Molecular Biology, Research Institute for the Biology of Farm Animals, FBN, Dummerstorf, Germany


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The liver plays a central role in the regulation of the metabolic status, partitioning of nutrients, and expenditure of energy. To gain insight into hepatic metabolic pathways and key transcripts affecting traits related to body composition, liver expression profiles were compared of pigs of two breeds, the obese German Landrace (DL) and the lean Pietrain (Pi). Porcine oligonucleotide microarray were hybridized with liver cRNAs obtained at peripubertal age (180 days of age) and prenatal stages (35, 63, and 91 days postconception) that represent three developmental stages of liver, i.e., period of differentiation, period of metabolic activity, and period of glycogen accumulation. In terms of the number of genes regulated between DL and Pi, the most striking distinctions were found at peripubertal age with upregulation of key genes of lipid metabolism pathways (FASN, ACSS2, ACACA) in obese DL pigs and upregulation of genes of cell growth and/or maintenance, and protein syntheses, as well as cell proliferation pathways (PPARD, POU1F1, IGF2R), in lean Pi pigs. Moreover, time course analysis of breed-dependent expression profiles revealed breed-typical temporal regulation from prenatal stages to peripubertal age of genes assigned to biological processes involving lipid pathways and cell activity, i.e., breed differences are already initiated during early prenatal development. Information about mRNA expression levels of the two breeds differing in body composition, partitioning and utilization of nutrients and energy reveals functional candidate genes for traits related to obesity and leanness.

liver; microarray; obesity; leanness


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
MOST NUTRIENTS ABSORBED by the intestine pass through the liver, which largely contributes to the overall metabolic status of the organism. The liver plays a central role in the regulation of carbohydrate, lipid, and amino acid metabolism and thus comprise genes potentially affecting traits related to leanness and obesity. German Landrace (DL) and Pietrain (Pi) pigs differ considerably in growth rate, body composition, and nutrient utilization (6, 55) and thus may represent a valuable model to study the genetic control of these traits, with DL being more "obese" and Pi being more "lean." Trait-/breed-dependent expression profiling of the liver will underscore the genes involved in genetic control of the metabolic status. Knowledge of the changes in the gene expression profile in pig liver during prenatal development and postnatal life is limited. We hypothesize that breed differences in liver expression profiles will be present at prenatal stages. These mainly depend on genetic effects rather than environmental effects. Moreover, prenatal events have been shown to affect postnatal growth and to be targets of fetal programming and are thus of interest. Expression profiling of fetal and peripubertal liver of pig breeds differing in utilization and partitioning of calories will point to metabolic pathways, which affect these physiological properties, and will reveal relevant candidate genes of growth and performance traits. Ultrastructural studies of porcine fetal livers show that development can be divided into three periods: a period of differentiation [18–40 days postconception (dpc)], a period of metabolic activity (40–80 dpc), and a period of glycogen accumulation (80–113 dpc) (8). These three main periods of liver development plus peripubertal stage were selected to compare the gene expression profile of DL and Pi.

Microarray analyses have been applied in the pig, for example, to unravel differential expression in muscle tissue due to fiber distribution, developmental stage, and breed (4, 10, 49, 50, 51, 67) and in the analysis of immune response and host-pathogen interaction (1, 33). Analysis of hepatic expression patterns in the pig has focused on diet-induced liver injuries (15). Using an application-specific cDNA microarray we have previously shown differential expression of the liver in discordant sib pairs differing in leanness- and obesity-related traits (38). The aim of this study was to obtain insight into pathways of hepatic metabolic regulation at the transcriptome level by comparing liver expression profiles from prenatal stages to peripubertal age of pigs representing breeds of divergent body composition by microarray analysis.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Animals and tissue collection.
To address the prenatal stages of 35, 63, and 91 dpc, fetuses were obtained from five sows per breed, DL or Pi, at each stage. Sows in their second or third parity were obtained from two large herd-book farms employing accurate on-farm herd record-keeping systems. Average performance of animals derived from these farms is given in Table 1. Sows were mated to boars of the same breed, DL or Pi, respectively. Sows were fed a standard gestation diet containing 12.6% crude protein, which was given restrictively throughout gestation. Sows were slaughtered at 35, 63, or 91 days of pregnancy. Uteri were removed immediately after bleeding. Fetuses were already dead when obtained from the uteri by cutting the umbilical cord. We used one male and one female fetus from each litter to prepare liver samples. We selected 20 peripubertal pigs (10 DL and 10 Pi with five castrated male and five female each; approximate age of 180 days) from the same herd. These pigs had no common sire or dam. Pigs were given ad libitum access to feed containing 16% crude protein. All animals were slaughtered according to European standards of handling including preslaughter fasting and welfare issues. Tissue samples were taken from lobus sinister hepaticus. All tissue samples were flash frozen in liquid nitrogen and stored at –80°C until use for RNA isolation. The experiments described in this study were performed in accordance with all appropriate regulations regarding care and use of research animals and were approved by the authors' institutional animal care and use committee.


View this table:
[in this window]
[in a new window]

  		Table 1. Means of traits related to leanness and obesity of the DL and the Pi herds from which animals examined were taken

 
RNA and target preparation for hybridization.
For isolation of total RNA liver samples were first ground in a mortar, then mixed, and homogenized with 1 ml of TRI Reagent (Sigma, Taufkirchen, Germany) using syringes and needles according to the manufacturer's protocol. After DNase treatment RNA was quantified and integrity was checked by electrophoresis of 1 µg of RNA on 1% agarose gels containing formaldehyde-stained with ethidium bromide. mRNA was isolated from total RNA using magnetic oligo(dT)-beads (Dynabeads; Dynal, Oslo, Norway). Different mRNA pools per stage and breed were made for subsequent chip hybridizations. Pools of 1 µg of mRNA of each group of animals were labeled with Cy3 or Cy5 using the CyScribe Post-Labeling Kit (GE Healthcare, Freiburg, Germany), which involves the use of oligo(dT) primers and random nanomers for synthesis of amino allyl-labeled first-strand cDNA and coupling of cDNA with CyDye NHS ester. Cy3/Cy5-labeled cDNAs were purified used CyScribe GFX purify kit (GE Healthcare). Cy3- and Cy5-labeled cDNAs were hybridized to the set of 13,297 pig-specific 70-mer oligonucleotides at 42°C for 14–18 h. To reduce the background of the slide, the Pronto Universal Hybridization kit was used (Corning Life Science, Schiphol-Rijk, The Netherlands). Hybridized slides were washed once with 1x SSC, 0.2% (wt/vol) SDS for 10 min at 42°C and twice with 0.1x SSC, 0.2% (wt/vol) SDS. Finally, slides were dipped into distilled water and air-dried.

Data analysis.
Microarrays were scanned with the 428 Array Scanner (Affymetrix, Santa Clara, CA). Spots were quantified, and quality filtering was applied using ImaGene5 software version 5.6 (BioDiscovery, Los Angeles, CA). Each spot had to pass a number of quality criteria, including spot consistency, minimum intensity levels, and minimum signal-to-background ratio. Arabidosis gene oligonucleotides, which were on the array as negative controls, were successfully discarded by this procedure. The subtracted signal median of genes that passed filtering was analyzed further with The Institute for Genomic Research (TIGR) Microarray Data Analysis System (40), which allows one to perform normalization steps: per spot, intensity dependent (lowess fit).

Filtered and normalized expression data were analyzed with TIGR Multiexperiment Viewer (MEV) (40). The t-test was conducted for each stage, and the set of P values for each stage was converted to a set of q values using the algorithm proposed by Storey and Tibshirani (45) (QVALUE). To compare all stages between DL and Pi, variance analysis (ANOVA, P < 0.05) was used, and these subsequent analyses were compared by hierarchical clustering.

All the microarray information has been submitted to the National Center for Biotechnology Information Gene Expression Omnibus web site (www.ncbi.nlm.nih.gov/geo/) (experiment series number: GSE5399; accession numbers: GSM123432–123434, GSM123442–123450).

Pathway identification by Expression Analysis Systemic Explorer.
Gene lists (GenBank accession numbers of human homologs) from microarray results were submitted to the Expression Analysis Systematic Explorer (EASE, accessible via: http://david.abcc.ncifcrf.gov/), which subsequently obtains corresponding "GeneID" (Entrez Gene; http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene) for further analysis. EASE takes into account the frequencies of genes belonging to particular Gene Ontology (GO) (http://www.geneontology.org/index.shtml) terms among the genes found to be regulated and among all genes that were addressed in an experiment. EASE performs a statistical analysis to detect overrepresented functional gene categories in the data set compared with all genes on the arrays. GO terms are reported with corresponding EASE scores, i.e., a conservative statistical test that gives the upper bound of the distribution of Jackknife Fisher exact probabilities and favors robust categories. Functional gene categories were considered significantly overrepresented at P < 0.05. A more detailed analysis of the genes' association with physiological pathways was performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG, http://www.genome.jp/kegg/pathway.html) and the PathwayAssist program (V3.0) (Stratagene, Amsterdam, The Netherlands). Each identified process was confirmed through PubMed/Medline.

Quantitative real-time RT-PCR.
Differential expression data of selected genes obtained from porcine oligonucleotide chips were validated by real-time PCR carried out on a LightCycler instrument (Roche Diagnostics, Mannheim, Germany). A two-step procedure was performed. First, RNAs obtained from liver of 10 animals per breed per stage were reverse transcribed using random hexamers and oligo(dT) primers and Superscript III reverse transcriptase (Invitrogen, Karlsruhe, Germany). Then, gene-specific PCR were done with reaction mixtures consisting of cDNA, 5 µM upstream and downstream primers (Table 2), and LightCycler DNA Master SYBR Green I (1x) (Roche). Templates were amplified by 45 cycles of 95°C for 15 s denaturation and 60°C for 10 s annealing and 72°C for 15 s for extension preceded by initial denaturation of 95°C for 10 min as universal thermal cycling parameters. On the basis of the analyses of melting curves of the PCR products, a high-temperature fluorescence acquisition point was determined and included in the amplification cycle program. For all assays a standard curve was generated by amplifying serial dilutions of specific PCR products. Fluorescence signals, which were recorded on-line during amplification, were subsequently analyzed using the "second derivative maximum" method of the LightCycler Data Analysis software. Normalization of variation in RT-PCR efficiency and initial RNA input was performed by using RPL32 gene as internal standard. In total, 80 individual liver mRNA samples of animals of the four stages of the two breeds were analyzed in duplicate. Quantitative real-time RT-PCR (qRT-PCR) data from each stage between DL and Pi were analyzed by t-test, and differences were considered significant at P < 0.05 (SAS version 8.02; SAS Institute, Cary, NC).


View this table:
[in this window]
[in a new window]

  		Table 2. Genes and primers used in quantitative real-time PCR for quantification of transcripts in liver at three prenatal stages and peripubertal age in DL and Pi

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
The animals used were representative of the two breeds that showed markedly different performance (Table 1). Comparative expression profiling by competitive hybridization of the microarrays revealed differential expression of genes between DL and Pi pigs in liver tissue at prenatal stages (35, 63, and 91 dpc) and at peripubertal age.

Different expression patterns and biological pathways identified by EASE between DL and Pi in each fetal stage and at peripubertal age.
The hepatic gene expression patterns of DL and Pi breeds were compared at each of the four stages of development. Three and four biological and technical replicate microarray hybridizations were performed for each stage, respectively (Table 3). Between 7,182 and 8,431 genes expressed were displayed after filtering with ImaGene software, and between 1,068 and 2,205 genes were found differentially expressed between the breeds at each stage after statistical analysis (t-test, P < 0.05) by MEV with corresponding false discovery rates varying between 0.05 and 0.13 as estimated by QVALUE (Table 3). Statistically significantly overrepresented GO terms associated with genes upregulated and downregulated between the pig breeds DL and Pi at three prenatal stages and one peripubertal stage were found with the Expression Analysis Systematic Explorer (EASE score, P < 0.05) (Table 4) (20).


View this table:
[in this window]
[in a new window]

  		Table 3. Numerical summary of gene expression profiles and changes in porcine liver between DL and Pi breeds at different prenatal stages and peripubertal age

 

View this table:
[in this window]
[in a new window]

  		Table 4. GO categories of genes differentially expressed between DL and Pi in liver at three prenatal and peripubertal stages as revealed by t-tests of each stage with more than twofold change and subsequent cluster analysis

 
At 35 dpc, the expression level of 310 transcripts varied by more than twofold, of which 210 were downregulated and 100 were upregulated in DL compared with Pi. The list of differentially regulated genes at 35 dpc is shown in Supplemental Table S1 (the online version of this article contains supplemental material). To gain a more mechanistic understanding of the process, the EASE score was used to identify GO functional categories, which were significantly overrepresented in DL or Pi with more than twofold different expression. GO terms at 35 dpc representing genes higher expressed in Pi than DL were related to transmission of nerval signals and components of membranes (Table 4). Transcripts with higher abundance in DL than Pi were genes assigned to categories related to physiological and defense mechanisms and nucleotide binding (Table 4).

At 63 dpc, 167 genes were found with more than twofold difference in expression (Supplemental Table S2). The 119 genes upregulated in Pi compared with DL belong to pathways associated with kinase activity or phosphotransferase activity, while 48 genes upregulated in DL belong to categories of early endosome and heat shock protein activity; however, EASE score was not significant as shown in Table 4.

In total 387 regulated genes were found at 91 dpc with only 25 gene being upregulated in Pi (Supplemental Table S3). Genes of the GO biological processes related to cell growth and/or maintenance, intracellular and cell communication and signaling, and cellular process were found higher in Pi, whereas genes involved in response to stimulus, heat shock protein activity, transferase activity, and energy pathways were upregulated in DL fetuses at 91 dpc (Table 4).

The highest number of regulated genes was found at the peripubertal stage. For DL vs. Pi, the 380 downregulated genes (Supplemental Table S4) belong to the biological processes of cell growth and/or maintenance, cell proliferation, cell communication, cell cycle, cell-cell signaling, transcription, protein binding, binding, and DNA binding (Table 4), whereas 174 upregulated genes (Supplemental Table S4) were assigned to the biological processes of lipid metabolism or biosynthesis, endosome and membrane fraction, as well as cell fraction (Table 4). The transcripts associated to cell proliferation and lipid metabolism pathways were subject of common regulators as revealed by network analysis of upregulated genes in obese DL compared with lean Pi pigs (Figs. 1 and 2).


Figure 1
View larger version (21K):
[in this window]
[in a new window]

  		Fig. 1. Network analysis of genes upregulated (blue ovals) in obese German Landrace (DL) pigs to identify the common regulator molecules performed using PathwayAssist. Genes (light blue): ACACA (acetyl-coenzyme A carboxylase alpha), ACSS2 (acetyl-coenzyme A synthetase 2), FASN (fatty acid synthetase), PKLR (pyruvate kinase), CYP21A2 (cytochrome P450, family 21, subfamily A, polypeptide 2), NROB1 (nuclear receptor subfamily 0, group B, member 1). Functional classes (orange): hormone activity and cAMP-dependent protein kinase. Transcription factors (violet): SREBF1 (sterol regulatory element binding transcription factor 1). Hormones, red: INS (insulin), GCG (glucagons), LEP (leptin). Green arrow, protein modification; blue arrow, expression; gray arrow (with dotted line), regulation.

 

Figure 2
View larger version (34K):
[in this window]
[in a new window]

  		Fig. 2. Network analysis of genes upregulated (blue ovals) in lean Pietrain (Pi) pigs to identify the common regulator molecules performed using PathwayAssist. Genes (light blue): PPARD (peroxisome proliferator-activated receptor delta), POU1F1 (POU domain, class 1, transcription factor 1), BIRC5 (baculoviral IAP repeat-containing 5), HMGB1 (high-mobility group box 1), IGF2R (insulin-like growth factor 2 receptor). Functional classes (orange): growth factor activity, JAK (Janus tyrosine kinase), PKC (protein kinase C beta 1), RPS6BK (ribosomal protein S6 kinase), MAP kinase. Transcription factors (violet): CREB1 (cAMP responsive element binding protein 1); proteins (red): INS (insulin), IFNG (interferon gamma), TNF (tumor necrosis factor), IGF1 (insulin-like growth factor 1), TGFB1 (transforming growth factor, beta 1), GH1 (growth hormone 1).

 
Time course analysis of breed-dependent expression profiles from fetal to peripubertal stage and related biological processes.
In total 13 array hybridizations were used (Table 3). The gene expression patterns of DL and Pi were analyzed throughout fetal to peripubertal stages by ANOVA (P < 0.05). These data were clustered to reveal prominent groups of genes with similar breed-dependent changes in expression patterns using a k-means algorithm (Fig. 3). To get more insight into the functional and structural components and impact of the genes represented in each cluster, the EASE score was used to identify GO functional categories, which are significantly overrepresented in each of the defined k-means clusters (Table 5). From this analysis, eight clusters can be distinguished. Genes of all clusters are presented in Supplemental Table S5.


Figure 3
View larger version (24K):
[in this window]
[in a new window]

  		Fig. 3. Clusters derived from time course analysis of gene expression profiles of DL and Pi throughout fetal to peripubertal stages obtained by ANOVA (P < 0.05). Data of comparisons between DL and Pi at all stages were clustered to reveal prominent groups of genes with similar changes in expression patterns by a k-means algorithm. Cluster number and the number of genes are denoted at top left of each cluster. For each cluster, the log2(DL/Pi) ratio is on the y-axis. The 13 microarrays analyzed [3 at 35 days postconception (dpc); 3 at 63 dpc; 4 at 91 dpc; 3 at 180 days of age] are represented along the x-axis. Data points represent means of the log2(DL/Pi) ratio of the number of genes as denoted for each cluster; error indicators show SD.

 

View this table:
[in this window]
[in a new window]

  		Table 5. GO categories of genes differentially expressed between DL and Pi in liver along three prenatal and postnatal developmental stages as revealed by ANOVA and subsequent cluster analysis

 
As shown in Fig. 3, clusters 1, 2, 3, and 7 represent genes that have the same tendency of higher gene expression in the obese DL pigs, whereas the clusters 4, 5, 6, and 8 comprise genes showing the tendency of higher expression in the Pi pigs. Cluster 1 (172 genes) and cluster 3 (80 genes) cover genes that were slightly higher expressed in DL fetuses than Pi fetuses and that showed even higher breed differences in peripubertal pigs. The GO overrepresentation analysis of these clusters revealed genes associated mainly with lipid biosynthesis, lipid metabolism, metabolism, biosynthesis, macromolecule biosynthesis, innate immune response, and catalytic activity (Table 5). In particular clusters 1 and 3 cover 15 and 9 genes, respectively, of the GO functional categories of lipid metabolism. As revealed by functional annotation the genes to metabolic pathways defined by KEGG, out of these 24 loci, three genes encode enzymes of fatty acid metabolism pathway (acetyl-coenzyme A acetyltransferase 2, acyl-coenzyme A dehydrogenase, dodecenoyl-coenzyme A delta isomerase), and three genes encode enzymes of steroid biosynthesis pathway (mevalonate decarboxylase, 3-hydroxy-3-methylglutaryl-coenzyme A reductase, 7-dehydrocholesterol reductase), while the other genes belong to distinct pathways of lipid metabolism. Cluster 2 contains 108 genes that are more highly expressed in DL than Pi at 91 dpc. This cluster is made up of genes associated mainly with nucleotide and RNA metabolism, as well as defense/immunity protein activity. Cluster 7 genes (71) show slightly decreasing differences of expression between breeds from fetal to peripubertal stages. GO overrepresentation analysis of this downward-tending cluster 7 highlighted genes encoding and controlling cell components, processes of cell cycle, and perception of external stimulus being more active in DL fetuses than in Pi fetuses, but not at the peripubertal age. The tendency of Pi breeds of having higher expression at all stages compared with DL was found in clusters 4, 5, 6 and 8, which represent 146, 95, 129, and 84 genes, respectively. Clusters 5 and 6 showed the same tendency of slightly lower expression in fetal stages and dramatically lower expression in peripubertal DL compared with Pi. The overrepresented biological processes in this cluster primarily involve cell growth and/or maintenance, cell proliferation, protein transport, and protein metabolism, as well as protein synthesis (Table 5). Cluster 4 had the opposite trend to cluster 2 with higher gene expression at 91 dpc in Pi. Many of the genes belong to the categories covering plasma membrane, protein modification, protein kinase activity, and phosphotransferase activity (Table 5). Cluster 8 displayed higher expression in Pi throughout all stages, thus the opposite of cluster 3. The most significant GO categories in this group are cellular processes, cell growth and/or maintenance, as well as plasma membrane (Table 5).

Comparison of qRT-PCR and microarray data.
qRT-PCR was used to confirm differential expression as indicated by microarray expression patterns. All of the eight genes selected for this analysis showed more than twofold different expression between breeds at peripubertal stage and were examined along the four developmental stages. Four upregulated genes [acetyl-coenzyme A synthetase 2 (ACSS2); fatty acid synthase (FASN); inhibitor of growth family, member 4 (ING4); and hematopoietic cell signal transducer (HCST)] and four downregulated genes [Janus kinase 3 (JAK3); eukaryotic translation initiation factor 4 gamma, 2 (EIF4G2); methionine aminopeptidase 2 (METAP2); and SIN3 homolog A, transcription regulator (SIN3A)] were selected to compare DL vs. Pi. The majority of these selected genes belong either to protein metabolism pathways (JAK3, EIF4G2, and METAP2), which were upregulated in peripubertal Pi, or to lipid biosyntheses pathways (FASN and ACSS2), which were upregulated in peripubertal DL. The primer sequences of the eight selected genes and the housekeeping control gene [ribosomal protein L32 (RPL32)] are shown in Table 2. In total 80 individual liver mRNA samples were analyzed in duplicate. All the expression patterns of the eight transcripts throughout three fetal stages and one peripubertal stage are shown in Fig. 4. ACSS2 transcripts showed higher abundance in DL than in Pi at all stages with the highest difference at the peripubertal stage. While the transcript abundance increased in peripubertal DL pigs compared with that of DL fetal pigs, transcript abundance decreased in the peripubertal Pi pigs. FASN tended to be higher expressed at fetal stages in DL, with a significant difference at 91 dpc and peripubertal stage. Transcript abundance was highest at 63 dpc in both breeds. ING4, METAP2, and EIF4G2 showed steadily decreasing transcript abundance along with development, with, mainly at early stages, significantly higher expression in DL than in Pi. Also JAK3 showed a steady decrease of expression alongside development but tended to be more highly expressed in Pi than in DL. HCST transcripts are most abundant at 63 dpc in both breeds. At 35 dpc HCST transcript abundance is higher in Pi than in DL, but at peripubertal stage it is higher in DL than in Pi. SIN3A is strongly upregulated in the peripubertal stage compared with fetal stages. Peripubertal DL had lower transcript abundance than peripubertal Pi. For each gene at each stage of development, the log2-fold changes of qRT-PCR and microarrays are presented in Table 6 together with Pearson correlations of qRT-PCR and microarrays. Most genes selected with more than twofold differences in expression at peripubertal age and 91 dpc according to microarray analyses showed corresponding values by qRT-PCR with high correlation. Coefficients of correlations are smaller and/or less reliable for the stages 35 dpc and 63 dpc due to a number of genes selected not being detected or scored after quality filtering at early prenatal samples by microarray analysis. The average correlation between microarray and qRT-PCR expression data over all developmental stages was 0.82 (P < 0.0001).


Figure 4
View larger version (26K):
[in this window]
[in a new window]

  		Fig. 4. Relative transcript levels of 4 genes downregulated and 4 genes upregulated in liver of peripubertal DL compared with peripubertal Pi. Bars represent means of data obtained from 10 individual samples per breed per stage; error indicators show SD. P values are given for comparisons between breeds within stage.

 

View this table:
[in this window]
[in a new window]

  		Table 6. Comparison of log2 (DL/Pi) ratios obtained by qRT-PCR and microarray analyses of 8 genes at 3 prenatal stages and peripubertal age and coefficients of correlation (and P values) between qRT-PCR and microarrays results at each of the 4 stages of development and all stages

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Prenatal development is an important predisposing factor for perinatal development and postnatal growth (37, 53). The development of a complex organ such as the liver relies on precise temporal and spatial gene expression patterns during ontogenesis. The unique peripubertal phenotype is a result of a cascade of transcriptional events that finally trigger gene expression in a liver-specific fashion and determine the phenotype and function of the liver (18). While a certain set of genes is constantly expressed to maintain the organ structure, a varying number of genes are regulated according to the metabolic demand of the surrounding organism. Since prenatal development affects postnatal growth, a good understanding of the genetic control of prenatal events will facilitate further progress in selection for growth and body composition traits. The gain in knowledge of genotype-adapted and -dependent nutritional requirements obtained from pigs of divergent breeds can provide information that is also relevant in nutritional genomics of other species, including human.

So far, research to evaluate changes in gene expression during the early stages of fetal pig development has focused on whole fetuses or muscle tissue but not on the liver (49–51, 54, 57, 61–63). Competitive hybridizations were carried out of the 13K-long oligonucleotide microarray, Qiagen-NRSP8, to determine biological pathways that are differentially expressed during liver ontogenesis between two breeds of pigs differing in traits related to growth, obesity, and leanness. The porcine 13K-long oligonucleotide microarray, Qiagen-NRSP8, has been shown to enable reliable expression profiling of different porcine tissues including liver (68). RNA pools were used for hybridisations that were derived from of 4–10 fetal or peripubertal liver samples each. Pooling was done to minimize technical and statistical issues resulting from individual variation in gene expression or variation that might arise from differences between sexes and families within breed, but to highlight breed typical differences only (22, 36). The average number of 7,000–8,500 genes expressed in each stage of liver is in agreement with the former study of Zhao et al. (68). Expression patterns of liver derived from DL and Pi at 35, 63, and 91 dpc and peripubertal age were examined. These time points were chosen to cover the three developmental stages of liver of differentiation (18–40 dpc), metabolic activity (40–80 dpc), and glycogen accumulation (80–113 dpc) that were defined based on histological studies of porcine liver during prenatal development (7, 8).

For validation of oligonucleotide chip data, we performed qRT-PCR with the same RNA samples that were used for the microarray experiments. In total, eight target genes were tested by qRT-PCR. RPL32 was used as a housekeeping control. RPL32 showed no regulation between DL and Pi along the development in our study and has been used successfully as a housekeeping control in other studies as well (12, 13, 67). Other housekeeping genes that have been proposed and used manifold such as ß-actin or GAPDH were also tested and found to be regulated in pigs with divergent body composition within or between breeds in this study as well as other studies (38, 39). In this study, the average fold change correlation between microarray and qRT-PCR was 0.82 (P < 0.0001), which is in line with results reported by Zhao and colleagues (68) using the same porcine long-oligonucleotide microarray. Recently, a large-scale study of the consistency of real-time PCR and gene expression measurements with two commercial long-oligonucleotide microarrays concluded that microarrays are invaluable discovery tools with acceptable reliability for genome-wide gene expression screening, though validation of putative changes in gene expression remains advisable (59).

Genes and biological pathways identified by EASE dominant in DL compared with Pi in fetal and peripubertal stages.
Genes that were more highly expressed in obese DL pigs than in lean Pi at the fetal stages of 35, 63, and 91 dpc encode elements of defense/immunity, protein activity, response to external stimulus, response to biotic stimulus, as well as heat shock protein activity pathways, covering indicators of cellular stress (28). Such genes have been found to be higher expressed in obese men (17). Diet-induced obesity in mice led to increased hepatic expression of genes categorized into the groups of defense, stress, and inflammation responses (23, 42). Evidence has been reported that abdominal fat tissue is metabolically and immunologically active and modulates response to dietary and other external stimuli of other tissues via the release of adipocyte-derived mediators, so-called adipocytokines (52). Interestingly, differences between live weight at age of slaughter and carcass weight (25% of live weight in DL and 21% of live weight in Pi) indicate that DL has a higher proportion of abdominal fat tissue. One might speculate that this higher amount of abdominal fat correlates with higher expression of genes involved in response to biotic and abiotic stressors. Moreover, it has been reported that porcine breed and age are important factors influencing the response to various stressors or infectious challenges (47). The present study indicates that the genetic control of properties related to metabolic status, nutrient utilization, and partitioning engage genes of defense/immunity, protein activity, and response to external stimulus. Existing breed differences of these properties are even established at early fetal development.

Peripubertal pigs of the breeds DL and Pi differ in their hepatic expression profile with significant EASE scores for the categories of lipid metabolism, lipid biosynthesis, and biosynthesis being more highly expressed in obese DL pigs than in Pi.

We found 14 genes of lipid metabolism pathways to be significantly more highly expressed in DL than in Pi peripubertal pigs, including transcripts that encode key enzymes of lipid metabolism such as acetyl-coenzyme A carboxylase alpha (ACACA); ACSS2; FASN; pyruvate kinase (PKLR); cytochrome P450, family 21, subfamily A, polypeptide 2 (CYP21A2); and nuclear receptor subfamily 0, group B, member 1 (NROB1). These key molecules were subjected to pathway analysis that revealed a network of common regulation of genes involved in anabolic pathways being upregulated in DL (Fig. 1). Leptin, insulin, and glucagons, as well as the transcription factor sterol regulatory element binding transcription factor 1, represent common regulatory molecules of the key enzymes of lipid metabolism (ACACA, ACSS2, FASN, PKLR, CYP21A2, NROB1).

ACACA is a major regulatory enzyme of fatty acid biosynthesis that catalyzes the carboxylation of acetyl-CoA to malonyl-CoA, which is the rate-limiting step in fatty acid synthesis. The main function of FASN is to catalyze the synthesis of long-chain saturated fatty acids from acetyl-CoA and malonyl-CoA. ACSS2 encodes a cytosolic enzyme that catalyzes the activation of acetate, i.e., the production of acetyl-CoA from acetate, for use in lipid synthesis and energy generation.

ACSS2 and FASN were selected for confirmation of differential expression by qRT-PCR throughout the developmental stages addressed here. The tendency for higher expression levels of both FASN and ACSS2 in DL than Pi was confirmed for the peripubertal stages but was found at earlier developmental stages as well. For ACSS2 most prominent breed differences were seen at peripubertal age. FASN expression was unregulated from 35 dpc to 63 dpc and downregulated from 63 dpc to peripubertal age; thus it was highest at 63 dpc. The prenatal regulation of expression of FASN in the liver and also of other genes encoding enzymes of the fatty acid metabolism as found here and in other studies (31) indicates that there are marked changes and suggests an important role of fatty acid metabolism during gestation. Indeed, deletion of FASN or ACACA in mice resulted in embryonic lethality, indicating that the de novo fatty acid synthesis is essential for embryonic development (30). Differences in gene expression or enzyme activity in lipid metabolism pathways in liver were also observed in mice or obese rats (39, 19), as well as in mice after long-term high-fat diet feeding (23, 24).

PKLR is an ubiquitously expressed enzyme that catalyzes the conversion of phosphoenolpyruvate to pyruvate with the generation of ATP. PKLR is associated with increased risk of Type 2 diabetes (56). CYP21A2 encodes a member of the cytochrome P450 superfamily of enzymes. This protein hydroxylates steroids at the 21 position and is required for the synthesis of steroid hormones including cortisol and aldosterone. NROB1 encodes a DAX1 protein that have been found to function primarily in steroidogenesis (34). Plasma levels of steroid hormones were also found to be associated with obesity (35). Interestingly, in addition to genes directly involved in fatty acid synthesis and genes affecting steroid hormone synthesis, a gene controlling cholesterol synthesis, isopentenyl diphosphate delta isomerase, was found to be more highly expressed in DL than in Pi. Differences in cholesterol biosynthesis were shown to be associated with obesity, resulting in a higher level of cholesterol deposition in adipose cells and increased adiposity (3, 32, 46).

Genes and biological pathways identified by EASE dominant in Pi compared with DL in both fetal and peripubertal stages.
We found 109 transcripts in the category of cell growth and/or maintenance pathways and 42 transcripts in cell proliferation pathways to be more abundant in Pi than in DL during peripuberty. Among these, peroxisome proliferator-activated receptor delta (PPARD); POU domain, class 1, transcription factor 1 (POU1F1); baculoviral inhibitor of apoptosis (IAP) repeat-containing 5 (BIRC5); high-mobility group box 1 (HMGB1); and insulin-like growth factor 2 receptor (IGF2R) are key molecules. These key molecules of cell proliferation pathways were used to display a network of interactions as shown in Fig. 2.

PPARD belongs to the peroxisome proliferator-activated receptor (PPAR) family of nuclear hormone receptors. Three different PPAR isoforms (PPARA, PPARG, and PPARD) have been identified. Activation of PPARD has been shown to promote lipid oxidation and energy uncoupling (60). Transgenic mice with PPARD hyperexpression have greatly reduced serum triglycerides and free fatty acids as well as improved insulin sensitivity; in contrast, PPARD knockout mice cannot upregulate fat burning and are metabolically less active (16, 25). We showed breed-dependent variation in PPARD expression, which may contribute to the genetically determined differences between the lean Pi and obese DL with Pi exhibiting a more active energy metabolism, thus limiting energy availability for fat deposition.

POU1F1 is a member of the POU family of transcription factors that regulate genes encoding growth hormone, prolactin, and thyroid-stimulating hormone beta subunit (27, 64). Polymorphisms in this gene were shown to be associated with growth and carcass traits in meat-type and fat-type porcine breeds (43, 44). An increase of POU1F1 transcript in lean Pi indicates that these effects might depend on differential regulation of POU1F1 expression.

Baculoviral IAP repeat-containing 5 (survivin, BIRC5) is a member of the IAP gene family. BIRC5 is expressed in the G2/M phase of the cell cycle in a cycle-regulated manner (26). At the beginning of mitosis, BIRC5 associates with microtubules of the mitotic spindle in a specific and saturable reaction that is regulated by microtubule dynamics. The upregulation of BIRC5 in the peripubertal lean Pi pig may indicate a higher dynamic of cell growth and proliferation in Pi compared with obese DL.

High mobility group proteins are small DNA-binding proteins playing an important role in transcriptional regulation (9). HMGB1 is the most intensively investigated member of this group of protein and plays an important role in inflammation, cell migration, differentiation, and tumorigenesis (2, 41, 66). HMGB1 affects cell growth and apoptosis via intracellular signaling pathway and is activated via key cell-signaling pathways such as MAP kinase and nuclear factor NF-{kappa}B (48).

The IGF2R gene is maternally imprinted (5), and its overexpression increases fetal overgrowth in humans, mice, and sheep (29, 65), as well as overproduction in tumors (14).

Pathway analysis revealed that among others NF-{kappa}B is a common regulator of BIRC5, HMGB1, and PPARD. MAP kinase regulates PPARD as well as HMGB1 (Fig. 2). Upregulation of HMGB1 is also activated of JAK/STAT pathway in the liver (58). Both IGFR2 and PPARD are positively regulated by insulin and thus functionally linked.

In this study we demonstrated that the genetic differences between obese and lean pigs are associated with upregulation of lipid biosynthesis pathways in obese pigs on the one hand and upregulation of cell growth and/or maintenance, as well as cell proliferation pathways, in lean pigs on the other hand. Thus this study reveals a number of functional candidate genes to be studied to provide further genetic and functional evidence for their impact on body composition. For some genes association studies have already indicated such effects. However, for a number of genes it remains to be elucidated which of these genes within these categories are causative for the phenotypic differences or are themselves affected by other hierarchically higher factors. Moreover, the study shows that phenotypic breed differences in obesity and leanness in the pig are not due to upregulation of lipolysis pathways as in lean and obesity-resistant knockout mouse (11).

Time course analysis of expression profiles and their breed-dependent biological processes.
Comparison of transcription profiles from fetal stages to peripubertal age yielded interesting insights into molecules involved in the developmental pattern of genetically predisposed obese and lean breeds. It highlighted classes of genes that showed common breed-typical expression throughout embryogenesis and postnatal growth through peripubertal age. Out of the eight clusters obtained, two groups of genes with the tendency of up- or downregulation in one breed compared with the other can be defined. The genes clustered according to their timely expression along development that made up the two groups of clusters display, when assigned to GO categories, the same picture of higher lipid metabolism and biosynthesis in DL and higher cellular activity in Pi as revealed by EASE of each single developmental stage.

However, the functional analysis of clustered genes reveals a more striking picture of breed differences along development, with clusters 3 and 8 as well as 1 and 5 being most conclusive: Clusters 3 and 8 comprise genes that are differentially expressed between DL and Pi throughout pre- and postnatal development. Clusters 1 and 5 cover genes that are mainly differentially expressed at peripubertal age. However, both pairs of clusters mark among other categories of genes loci involved in lipid metabolism on the one hand and cell growth and/or maintenance on the other hand as being either upregulated in DL or Pi, respectively. Thus, the analysis demonstrated that the transcriptional differences between obese and lean pigs involving lipid pathways and cell activity are already initiated during early embryogenesis.

This is the first microarray analysis that demonstrates 1) that lipid/sterol synthetic pathways and also responses to stimuli are more pronounced in obese pigs but cell growth/cell maintenance, cell proliferation, as well as protein synthesis, are more pronounced in lean pigs and 2) that this divergence between the breeds is already initiated during early embryogenesis.

Application of the 13K-long porcine oligonucleotide microarray to compare gene expression of prenatal and postnatal liver samples of two divergent breeds indicated that Pi tend to exhibit higher expression of genes involved in changes and development of tissues and cells while DL tend to show higher transcript abundance of genes involved in intracellular energy and metabolite conversion, especially of lipids. These trends are visible even at early stages of development. Many candidate genes with potential roles in meat quality and carcass traits have been identified. Genome scans have successfully been used to map quantitative trait loci (QTL) for growth, body composition, and meat quality traits [see Pig QTL database, http://www.animalgenome.org/QTLdb/] (21). Mirroring the data provided by expression studies onto the QTL maps will reveal the positional functional candidate genes that are the most promising for further research.


   FOOTNOTES
 
Address for reprint requests and other correspondence: K. Wimmers, FBN-Dummerstorf, Wilhelm-Stahl-Allee 2, 18196 Dummerstorf, Germany (e-mail: wimmers{at}fbn-dummerstorf.de).

Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Afonso CL, Piccone ME, Zaffuto KM, Neilan J, Kutish GF, Lu Z, Balinsky CA, Gibb TR, Bean TJ, Zsak L, Rock DL. African swine fever virus multigene family 360 and 530 genes affect host interferon response. J Virol 78: 1858–1864, 2004.[Abstract/Free Full Text]
  2. Andersson U, Erlandsson-Harris H, Yang H, Tracey KJ. HMGB1 as a DNA-binding cytokine. J Leukoc Biol 72: 1084–1091, 2002.[Abstract/Free Full Text]
  3. Angel A, Bray GA. Synthesis of fatty acids and cholesterol by liver, adipose tissue and intestinal mucosa from obese and control patients. Eur J Clin Invest 9: 355–362, 1979.[ISI][Medline]
  4. Bai Q, McGillivray C, da Costa N, Dornan S, Evans G, Stear MJ, Chang KC. Development of a porcine skeletal muscle cDNA microarray: analysis of differential transcript expression in phenotypically distinct muscles. BMC Genomics 4: 8, 2003.[CrossRef][Medline]
  5. Barlow DP, Stoger R, Herrmann BG, Saito K, Schweifer N. The mouse insulin-like growth factor type-2 receptor is imprinted and closely linked to the Tme locus. Nature 349: 84–87, 1991.[CrossRef][Medline]
  6. Baumung R, Lercher G, Willam A, Sölkner J. Feed intake behaviour of different pig breeds during performance testing on station. Arch Tierz Dummerstorf 49: 77–88, 2006.
  7. Bielanska-Osuchowska Z, Krzynowek-Wojciechowska J. Morphometric investigations of the pig developing liver during the prenatal period. Pol Arch Weter 30: 7–16, 1990.[Medline]
  8. Bielanska-Osuchowska Z. Ultrastructural and stereological studies of hepatocytes in prenatal development of swine. Folia Morphol (Warsz) 55: 1–19, 1996.[Medline]
  9. Bustin M, Lehn DA, Landsman D. Structural features of the HMG chromosomal proteins and their genes. Biochim Biophys Acta 1049: 231–243, 1990.[Medline]
  10. Cagnazzo M, Te Pas MFW, Priem J, de Wit AAC, Pool MH, Davoli R, Russo VV. Comparison of prenatal muscle tissue expression profiles of two pig breeds differing in muscle characteristics. J Anim Sci 84: 1–10, 2006.[Abstract/Free Full Text]
  11. Castro-Chavez F, Yechoor VK, Saha PK, Martinez-Botas J, Wooten EC, Sharma S, O'Connell P, Taegtmeyer H, Chan L. Coordinated upregulation of oxidative pathways and downregulation of lipid biosynthesis underlie obesity resistance in perilipin knockout mice: a microarray gene expression profile. Diabetes 52: 2666–2674, 2003.[Abstract/Free Full Text]
  12. Dawson HD, Beshah E, Nishi S, Solano-Aguilar G, Morimoto M, Zhao A, Madden KB, Ledbetter TK, Dubey JP, Shea-Donohue T, Lunney JK, Urban JF Jr. Localized multigene expression patterns support an evolving Th1/Th2-like paradigm in response to infections with Toxoplasma gondii and Ascaris suum. Infect Immun 73: 1116–1128, 2005.[Abstract/Free Full Text]
  13. Dawson HD, Royaee AR, Nishi S, Kuhar D, Schnitzlein WM, Zuckermann F, Urban J Jr, Lunney JK. Identification of key immune mediators regulating T helper 1 responses in swine. Vet Immunol Immunopathol 100: 105–111, 2004.[CrossRef][ISI][Medline]
  14. De Souza AT, Hankins GR, Washington MK, Orton TC, Jirtle RL. M6P/IGF2R gene is mutated in human hepatocellular carcinomas with loss of heterozygosity. Nat Genet 11: 447–449, 1995.[CrossRef][ISI][Medline]
  15. Esfandiari F, Villanueva JA, Wong DH, French SW, Halsted CH. Chronic ethanol feeding and folate deficiency activate hepatic endoplasmic reticulum stress pathway in micropigs. Am J Physiol Gastrointest Liver Physiol 289: G54–G63, 2005.[Abstract/Free Full Text]
  16. Evans RM, Barish GD, Wang YX. PPARs and the complex journey to obesity. Nat Med 10: 355–361, 2004.[CrossRef][ISI][Medline]
  17. Gabrielsson BG, Johansson JM, Lonn M, Jernas M, Olbers T, Peltonen M, Larsson I, Lonn L, Sjostrom L, Carlsson B, Carlsson LM. High expression of complement components in omental adipose tissue in obese men. Obes Res 11: 699–708, 2003.[ISI][Medline]
  18. Gualdi R, Bossard P, Zheng M, Hamada Y, Coleman JR, Zaret KS. Hepatic specification of the gut endoderm in vitro: cell signaling and transcriptional control. Genes Dev 10: 1670–1682, 1996.[Abstract/Free Full Text]
  19. Hastings IM, Hill WG. 1: Analysis of lines of mice selected for fat content. 2: Correlated responses in the activities of enzymes involved in lipogenesis. Genet Res 55: 55–61, 1990.[ISI][Medline]
  20. Hosack DA, Dennis G Jr, Sherman BT, Lane HC, Lempicki RA. Identifying biological themes within lists of genes with EASE. Genome Bio 4: R70, 2003.[CrossRef]
  21. Hu ZL, Dracheva S, Jang W, Maglott D, Bastiaansen J, Rothschild MF, Reecy JM. A QTL resource and comparison tool for pigs: PigQTLDB. Mamm Genome 16: 792–800, 2005.[CrossRef][ISI][Medline]
  22. Kendziorski CM, Zhang Y, Lan H, Attie AD. The efficiency of pooling mRNA in microarray experiments. Biostatistics 4: 465–477, 2003.[Abstract]
  23. Kim S, Sohn I, Ahn JI, Lee KH, Lee YS, Lee YS. Hepatic gene expression profiles in a long-term high-fat diet-induced obesity mouse model. Gene 340: 99–109, 2004.[CrossRef][ISI][Medline]
  24. Kim S, Sohn I, Lee YS, Lee YS. Hepatic gene expression profiles are altered by genistein supplementation in mice with diet-induced obesity. J Nutr 135: 33–41, 2005.[Abstract/Free Full Text]
  25. Lee CH, Olson P, Hevener A, Mehl I, Chong LW, Olefsky JM, Gonzalez FJ, Ham J, Kang H, Peters JM, Evans RM. PPARdelta regulates glucose metabolism and insulin sensitivity. Proc Natl Acad Sci USA 103: 3444–3449, 2006.[Abstract/Free Full Text]
  26. Li F, Ambrosini G, Chu EY, Plescia J, Tognin S, Marchisio PC, Altieri DC. Control of apoptosis and mitotic spindle checkpoint by survivin. Nature 396: 580–584, 1998.[CrossRef][Medline]
  27. Li S, Crenshaw EB, Rawson EJ, Simmons DM, Swanson LW, Rosenfeld MG. Dwarf locus mutants lacking three pituitary cell types result from mutations in the POU-domain gene Pit-1. Nature 347: 528–533, 1990.[CrossRef][Medline]
  28. Liu Y, Steinacker JM. Changes in skeletal muscle heat shock proteins: pathological significance. Front Biosci 6: D12–D25, 2001.[ISI][Medline]
  29. Ludwig T, Eggenschwiler J, Fisher P, D'Ercole AJ, Davenport ML, Efstratiadis A. Mouse mutants lacking the type 2 IGF receptor (IGF2R) are rescued from perinatal lethality in Igf2 and Igf1r null backgrounds. Dev Biol 177: 517–535, 1996.[CrossRef][ISI][Medline]
  30. Mao J, Demayo FJ, Li H, Abu-Elheiga L, Gu Z, Shaikenov TE, Kordari P, Chirala SS, Heird WC, Wakil SJ. Liver-specific deletion of acetyl-CoA carboxylase 1 reduces hepatic triglyceride accumulation without affecting glucose homeostasis. Proc Natl Acad Sci USA 103: 8552–8557, 2006.[Abstract/Free Full Text]
  31. McNeil CJ, Finch AM, Page KR, Clarke SD, Ashworth CJ, McArdle HJ. The effect of fetal pig size and stage of gestation on tissue fatty acid metabolism and profile. Reproduction 29: 757–763, 2005.
  32. Miettinen TA, Tilvis RS. Cholesterol synthesis and storage in adipose tissue. Int J Obes 5: 613–618, 1981.[ISI][Medline]
  33. Moser RJ, Reverter A, Kerr CA, Beh KJ, Lehnert SA. A mixed-model approach for the analysis of cDNA microarray gene expression data from extreme-performing pigs after infection with Actinobacillus pleuropneumoniae. J Anim Sci 82: 1261–1271, 2004.[Abstract/Free Full Text]
  34. Niakan KK, McCabe ER. DAX1 origin, function, and novel role. Mol Genet Metab 86: 70–83, 2005.[CrossRef][ISI][Medline]
  35. Pasquali R. Obesity and androgens: facts and perspectives. Fertil Steril 85: 1319–1340, 2006.[CrossRef][ISI][Medline]
  36. Peng X, Wood CL, Blalock EM, Chen KC, Landfield PW, Stromberg AJ. Statistical implications of pooling RNA samples for microarray experiments. BMC Bioinformatics 4: 26, 2003.[CrossRef][Medline]
  37. Pond WG, Mersmann HJ. Comparative response of lean or genetically obese swine and their progeny to severe feed restriction during gestation. J Nutr 118: 1223–1231, 1988.[Abstract/Free Full Text]
  38. Ponsuksili S, Murani E, Schellander K, Schwerin M, Wimmers K. Identification of functional candidate genes for body composition by expression analyses and evidencing impact by association analysis and mapping. Biochim Biophys Acta 1730: 31–40, 2005.[Medline]
  39. Rolland V, Dugail I, Le Liepvre X, Lavau M. Evidence of increased glyceraldehyde-3-phosphate dehydrogenase and fatty acid synthetase promoter activities in transiently transfected adipocytes from genetically obese rats. J Biol Chem 270: 1102–1106, 1995.[Abstract/Free Full Text]
  40. Saeed AI, Sharov V, White J, Li J, Liang W, Bhagabati N, Braisted J, Klapa M, Currier T, Thiagarajan M, Sturn A, Snuffin M, Rezantsev A, Popov D, Ryltsov A, Kostukovich E, Borisovsky I, Liu Z, Vinsavich A, Trush V, Quackenbush J. TM4: a free, open-source system for microarray data management and analysis. Biotechniques 34: 374–378, 2003.[ISI][Medline]
  41. Schlueter C, Weber H, Meyer B, Rogalla P, Roser K, Hauke S, Bullerdiek J. Angiogenetic signaling through hypoxia: HMGB1: an angiogenetic switch molecule. Am J Pathol 166: 1259–1263, 2005.[Abstract/Free Full Text]
  42. Schmid GM, Converset V, Walter N, Sennitt MV, Leung KY, Byers H, Ward M, Hochstrasser DF, Cawthorne MA, Sanchez JC. Effect of high-fat diet on the expression of proteins in muscle, adipose tissues, and liver of C57BL/6 mice. Proteomics 4: 2270–2282, 2004.[CrossRef][ISI][Medline]
  43. Song C, Gao B, Teng Y, Wang X, Wang Z, Li Q, Mi H, Jing R, Mao J. MspI polymorphisms in the 3rd intron of the swine POU1F1 gene and their associations with growth performance. J Appl Genet 46: 285–289, 2005.[Medline]
  44. Stancekova K, Vasicek D, Peskovicova D, Bulla J, Kubek A. Effect of genetic variability of the porcine pituitary-specific transcription factor (PIT-1) on carcas traits in pigs. Anim Genet 30: 313–315, 1999.[CrossRef][ISI][Medline]
  45. Storey JD, Tibshirani R. Statistical significance for genomewide studies. Proc Natl Acad Sci USA 100: 9440–9445, 2003.[Abstract/Free Full Text]
  46. Stylianou IM, Clinton M, Keightley PD, Pritchard C, Tymowska-Lalanne Z, Bunger L, Horvat S. Microarray gene expression analysis of the Fob3b obesity QTL identifies positional candidate gene Sqle and perturbed cholesterol and glycolysis pathways. Physiol Genomics 20: 224–232, 2005.[Abstract/Free Full Text]
  47. Sutherland MA, Rodriguez-Zas SL, Ellis M, Salak-Johnson JL. Breed and age affect baseline immune traits, cortisol and performance in growing pigs. J Anim Sci 83: 2087–2095, 2005.[Abstract/Free Full Text]
  48. Taguchi A, Blood DC, del Toro G, Canet A, Lee DC, Qu W, Tanji N, Lu Y, Lalla E, Fu C, Hofmann MA, Kislinger T, Ingram M, Lu A, Tanaka H, Hori O, Ogawa S, Stern DM, Schmidt AM. Blockade of RAGE-amphoterin signalling suppresses tumour growth and metastases. Nature 405: 354–360, 2000.[CrossRef][Medline]
  49. Te Pas MF, de Wit AAC, Priem J, Cagnazzo M, Davoli R, Russo V, Pool MH. Transcriptome expression profiles in prenatal pigs in relation to myogenesis. J Muscle Res Cell Motil 26: 157–165, 2005.[CrossRef][ISI][Medline]
  50. Te Pas MFW, Cagnazzo M, de Wit AAC, Priem J, Pool M, Davoli R. Muscle transcriptomes of Duroc and Pietrain pig breeds during prenatal formation of skeletal muscle tissue using microarray technology. Arch Tierz Dummerstorf 48: 141–147, 2005.
  51. Te Pas MFW, Pool MH, Hulsegge I, Janss LLG. Analysis of the differential transcriptome expression profiles during prenatal muscle tissue development in pigs. Arch Tierz Dummerstorf 49: 110–115, 2006.
  52. Tilg H, Moschen AR. Adipocytokines: mediators linking adipose tissue, inflammation and immunity. Nat Rev Immunol 6: 772–783, 2006.[CrossRef][ISI][Medline]
  53. Van der Lende T, Knol EF, Leenhouwers JI. Prenatal development as a predisposing factor for perinatal losses in pigs. Reprod Suppl 58: 247–261, 2001.[Medline]
  54. Vaughan TJ, James PS, Pascall JC, Brown KD. Expression of the genes for TGF alpha, EGF and the EGF receptor during early pig development. Development 116: 663–669, 1992.[Abstract]
  55. Volk B, Biedermann G, Kuhn M, Jatsch C. Influence of different genetic origins on fattening and carcass performance, on quality of meet and fat as well as fatty acid pattern of phospholipids from pigs. Arch Tierz Dummerstorf 47: 455–462, 2004.
  56. Wang H, Chu W, Das SK, Ren Q, Hasstedt SJ, Elbein SC. Liver pyruvate kinase polymorphisms are associated with type 2 diabetes in northern European Caucasians. Diabetes 51: 2861–2865, 2002.[Abstract/Free Full Text]
  57. Wang K, McCarter R, Wright J, Beverly J, Ramirez-Mitchell R. Regulation of skeletal muscle stiffness and elasticity by titin isoforms: a test of the segmental extension model of resting tension. Proc Natl Acad Sci USA 88: 7101–7105, 1991.[Abstract/Free Full Text]
  58. Wang SB, Yao YM, Dong N, Yu Y, Tao GC. Role of Janus kinase/signal transducer and activator of transcription pathway in mediating mRNA expression of high mobility group box1 protein in the liver in septic rats. Hongguo Wei Zhong Bing Ji Jiu Yi Xue 15: 147–149, 2003.
  59. Wang Y, Barbacioru C, Hyland F, Xiao W, Hunkapiller KL, Blake J, Chan F, Gonzalez C, Zhang L, Samaha RR. Large scale real-time PCR validation on gene expression measurements from two commercial long-oligonucleotide microarrays. BMC Genomics 7: 59, 2006.[CrossRef][Medline]
  60. Wang YX, Lee CH, Tiep S, Yu RT, Ham J, Kang H, Evans RM. Peroxisome-proliferator-activated receptor delta activates fat metabolism to prevent obesity. Cell 113: 159–170, 2003.[CrossRef][ISI][Medline]
  61. Wesolowski SR, Raney NE, Ernst CW. Developmental changes in the fetal pig transcriptome. Physiol Genomics 16: 268–274, 2004.[Abstract/Free Full Text]
  62. Wilson ME, Sonstegard TS, Smith TPL, Fahrenkrug SC, Ford SP. Differential expression during elongation in the preimplantation pig embryo. Genesis 26: 9–14, 2000.[CrossRef][ISI][Medline]
  63. Wimmers K, Murani E, Schellander K, Ponsuksili S. Combining QTL- and expression-analysis: identification of functional positional candidate genes for meat quality and carcass traits. Arch Tierz Dummerstorf 48: 23–31, 2005.
  64. Wu W, Cogan JD, Pfaffle RW, Dasen JS, Frisch O'Connell SM H, Flynn SE, Brown MR, Mullis PE, Parks JS, Phillips JA, Rosenfeld MG. Mutations in PROP1 cause familial combined pituitary hormone deficiency. Nat Genet 18: 147–149, 1998.[CrossRef][ISI][Medline]
  65. Young LE, Fernandes K, McEvoy TG, Butterwith SC, Gutierrez CG, Carolan C, Broadbent PJ, Robinson JJ, Wilmut I, Sinclair KD. Epigenetic change in IGF2R is associated with fetal overgrowth after sheep embryo culture. Nat Genet 27: 153–154, 2001.[CrossRef][ISI][Medline]
  66. Zetterstrom CK, Bergman T, Rynnel-Dagoo B, Harris HE, Soder O, Andersson U, Boman HB. High mobility group box chromosomal protein 1 (HMGB1) is an antibacterial factor produced by the human adenoid. Pediatr Res 52: 148–154, 2002.[CrossRef][ISI][Medline]
  67. Zhao SH, Nettleton D, Liu W, Fitzsimmons C, Ernst CW, Raney NE, Tuggle CK. Complementary DNA macroarray analyses of differential gene expression in porcine fetal and postnatal muscle. J Anim Sci 81: 2179–2188, 2003.[Abstract/Free Full Text]
  68. Zhao SH, Recknor J, Lunney JK, Nettleton D, Kuhar D, Orley S, Tuggle CK. Validation of a first-generation long-oligonucleotide microarray for transcriptional profiling in the pig. Genomics 86: 618–625, 2005.[CrossRef][ISI][Medline]



This article has been cited by other articles:


Home page
 Physiol. GenomicsHome page
L. Ordovas, R. Roy, S. Pampin, P. Zaragoza, R. Osta, J. C. Rodriguez-Rey, and C. Rodellar
The g.763G>C SNP of the bovine FASN gene affects its promoter activity via Sp-mediated regulation: implications for the bovine lactating mammary gland
Physiol Genomics, July 9, 2008; 34(2): 144 - 148.
[Abstract] [Full Text] [PDF]

This Article
Free upon publication Free Article
Right arrow Abstract Freely available