| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ANIMAL GROWTH, PHYSIOLOGY, AND REPRODUCTION |
* Department of Animal Science and Technology, National Taiwan University, Taipei 106, Taiwan; and
and
Institute of Molecular Medicine, National Cheng Kung University Medical College, Tainan 701, Taiwan
 |
Abstract |
|---|
 Top Abstract INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
Key Words: adipose tissue • expressed sequence tag • pig
|
INTRODUCTION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
; Havel, 2002). Porcine adipose tissue has abundant mRNA for adiponectin mRNA and its receptors (Ding et al., 2004; Wang et al., 2004) and produces leptin (Ramsay et al., 1998). These findings indicate that the adipose tissue in pigs also acts as an endocrine tissue and expresses genes involved in regulating metabolism and physiological functions in other tissues.
The expressed sequence tag (EST) technique is an effective approach to study functional expression of genes in various cells and tissues (Adams et al., 1992). To understand the function of adipose tissue, an analysis of abundantly expressed genes in the tissue is needed. Mikawa et al. (2004) characterized 298 EST clones from pig adipose tissue. Today, the largest EST library related to porcine adipose tissue in the National Center for Biotechnology Information (NCBI) dbEST is a library (dbEST library ID.1316) with 2,155 sequences, which is not full-length enriched. The full-length cDNA sequences provide more reliable evidence for determining the existence, structure, and function of a gene (The RIKEN Genome Exploration Research Group Phase II Team and the FANTOM Consortium, 2001; Mammalian Gene Collection Program Team, 2002).
Therefore, we conducted this experiment to clone full-length-enriched cDNA sequences of genes expressed in porcine adipose tissue to test the hypothesis that in addition to housekeeping genes, the genes involved in metabolism are abundantly expressed in adipose tissue, and to study the tissue distribution of several of these abundantly expressed genes.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
For construction of the full-length cDNA enriched library, 3 male Lee-Sung pigs (18-mo-old) were killed by electrocution combined with exsanguination. Subcutaneous adipose tissue was removed from the dorsal/ lumbar region for RNA extraction. Lee-Sung pigs, described by Lee et al. (1983) and Mason and Porter (2002), are a breed selected for reduced BW; they were derived by crossing Lanyu sows (small ears and size, and mature BW < 50 kg) with Landrace boars in Taiwan. For tissue distribution studies, four 2-mo-old Lee-Sung pigs (2 females and 2 males) were used. Heart, LM, liver, spleen, kidney, lung, testis, and subcutaneous adipose tissue were dissected, frozen quickly in liquid N, and stored at –70°C until RNA extraction.
Extraction of RNA
Total RNA was extracted by the guanidinium-phenol:chloroform extraction method (Chomczynski and Sacchi, 1987), with modifications described by Hsu and Ding (2003). The quality of the RNA was monitored by examination of the 18S and 28S ribosomal RNA bands after electrophoresis. The RNA was quantified by spectrophotometry at 260 nm and stored at –70°C. Poly(A) RNA from adipose tissue was extracted using a micro poly(A) mRNA purification kit (Ambion, Cambridgeshire, UK), according to the manufacturer’s instructions.
Construction of a Full-Length-Enriched cDNA Library
The SMART cDNA Library Construction Kit (Clontech, Palo Alto, CA) was used to construct a porcine full-length-enriched cDNA library. Adipose tissue poly (A) RNA (4 µg obtained from a pool of equal amounts of RNA from each of 3 pigs) was reverse transcribed at 42°C with a MMLV RNase H– point mutant reverse transcriptase (Clontech). Approximately 3.5 µg of cDNA of the RT product was used to construct the library. The enzyme terminal transferase activity of MMLV reverse transcription adds additional deoxycytidines (dC) to the 3' end of the cDNA. Addition of the modified SMART oligonucleotide containing deoxyguanidines (dG) at the 3' end allows it to anneal to the RT products, thereby switching templates and continuing replication on to the end of the SMART oligonucleotide sequence. Using primer pairs in the SMART cDNA Library Construction Kit, the full-length cDNA RT products were amplified by PCR.
The PCR-amplified cDNA library was subjected to SfiI restriction enzyme digestion. Digested products were extracted with phenol:chloroform and precipitated with ethanol (95%). The cDNA products were then separated according to size using a Chroma Spin-400 column (Clontech) following the manufacturer’s recommendations. Fractions were collected and molecular weight was monitored by agarose gel electrophoresis. Fractions greater than 500 bp were combined (fractions 4 to 7) and used for cDNA library construction.
The cDNA was then ligated to restriction enzyme-digested, pDNR-LIB vectors (Clontech). After the ligation, plasmids were electroporated into EPI400 electroporation-competent cells (Epicentre, Madison, WI). Transformants were cultured in salt-optimized carbon medium (2% tryptone, 0.5% yeast extract, 8.5 mM NaCl, 20 mM glucose) at 37°C for 1 h. The titer of the library was then determined using an LB (Luria-Bertani Medium with 1% tryptone, 0.5% yeast extract, and 0.5% NaCl) agar plate supplemented with 50 ug of ampicillin/ uL. Colonies were picked and screened to eliminate self-ligated colonies. Colonies (n = 96) were picked randomly to determine the average insert size of the cDNA library. In the current experiment, 2,880 colonies were sequenced.
cDNA Sequencing
Plasmid DNA purification followed an alkaline lysis procedure developed by Birnboim and Doly (1979), with modification to fit a high-throughput plasmid DNA preparation using 96-well PVDF filter plates (Corning, NY). Sequences were determined using a modification of the Sanger et al. (1977) method with fluorescent dideoxy termination in an automated ABI 3730 DNA Sequencer (Applied Biosystems, Foster City, CA).
Cluster Analysis and Annotation
The porcine cDNA sequences were validated by removal of poor-quality bases (QV < 20, equivalent to 0.01 of sequencing error) using the trim option in the Phred program (Ewing et al., 1998); acceptable clones had at least 100 bp after trimming. The vector sequences were removed, and sequences with long poly (A) RNA tails were rejected. Sequences were clustered and assembled into contigs using TGI clustering tools based on Megablast (Zhang et al., 2000) and CAP3 (Huang and Madan, 1999). The comparison quality was set greater than 95% identity over at least a 40-bp fragment. Sequences were then searched against the nonredundant nucleotide database in NCBI and the Gene Indices database in TIGR using Blastn. The determination of putative ATG (translation start codon) was performed to estimate the percentage of full-length cDNA sequences.
Gene Ontology Analysis
To understand the functional classification of genes in porcine adipose tissue, the accession number of known gene identities for those EST sequences in the adipose cDNA library was used to functionally sort the genes according to the descriptive terms of gene ontology (Ashburner et al., 2000). Frequency of each functional category was then summarized and reported in a pie chart format.
Northern Analysis
Total RNA (20 µg of each sample) from various tissues was electrophoresed and transferred to nylon membranes. The membrane was prehybridized at 42°C in UltraHyb (Ambion) for 1 h, and then the denatured cDNA probe (95°C for 5 min) was added at a concentration of 1 pM to hybridize with the targeted gene transcripts overnight at 42°C. The porcine 18S sequence was previously described (Wang et al., 2004), and other probe sequences are indicated in Table 1. Primers used to synthesize the probes for 18s, decorin, adipocyte (A–) fatty acid binding protein (FABP), and lactate dehydrogenase B subunit (LDH-B) are also in Table 1. Probes for the remaining 5 genes were synthesized using parts of the M13 vector sequence (Table 1). Hybridization results were quantified by phosphor-image analysis as previously described (Hsu et al., 2004; Liu et al., 2005). The densitometric value for an individual transcript in a sample lane was normalized to the densitometric value for the 18S ribosomal RNA in the same lane.
|
|
RESULTS AND DISCUSSION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
. Because the clones in this library were picked randomly and sequenced, 85% of the expressed genes found in this library represent those abundantly expressed mRNA species (200 to 1,000 copies per cell; Ewens and Grant, 2001). The average insert size was approximately 800 bp for our cDNA library. This average size was between that of 617 bp reported by Whitworth et al. (2004) for an early embryo library and that of 1,500 bp reported by Uenishi et al. (2004) for a multitissue cDNA library. When the known gene fragments of the current EST collection were compared with sequences in Genbank by Blast analysis, the full-length cDNA content in this library was estimated to be 15.6% (Table 3). In Mikawa et al. (2004) study, only 298 EST clones from porcine subcutaneous adipose tissue cDNA were characterized. Uenishi et al. (2004) have compiled a large assembly of pig EST from several tissues, including liver, lung, and spleen, but not adipose tissue. Recently, Whitworth et al. (2004) analyzed 2,489 porcine gene fragments expressed during embryogenesis, expanding our understanding of the gene expression profile during early porcine embryonic development. The total EST library related to porcine adipose tissue in the NCBI dbEST so far is about 2,518 sequences, which is not full-length enriched. It is known that the full-length cDNA sequences are more reliable evidence for determining the existence, structure, and function of a gene [The RIKEN Genome Exploration Research Group Phase II Team and the FANTOM Consortium, 2001; Mammalian Gene Collection (MGC) Program Team, 2002]. In the current study, this full-length cDNA enriched library added 1,527 unique gene fragments to the existing 2,518 sequences in the NCBI dbEST that is useful for further gene function study.
|
|
). Under the biological processes classification, genes other than mitochondrial genes were distributed into the following categories: cellular process = 41%, physiological process = 41%, regulation of biological process = 11%, and development 5% (Figure 1B). The largest 4 fractions totaled to more than 98% of the sequences (Figure 1B). Analysis according to cellular components (Figure 1C), distributed genes into the following categories: 41% cell, 32% organelle, 20% protein complex, and 3% extracellular region. The functional analysis indicated that many of the genes are involved in binding (1 of the top 10 abundantly expressed genes is A-FABP and several other proteins that bind fatty acids are on the abundantly expressed gene list) and metabolism. A serial analysis of gene expression has demonstrated that about 32% of the genes expressed in rodent adipose tissue are involved in metabolism (Bolduc et al., 2004), supporting the concept that adipose tissue is not an inactive tissue but is metabolically dynamic.
|
and Bolduc et al. (2004), the adipose tissue cDNA library contained many mitochondrial genes [10% in the current study compared with 30% in the study of Mikawa et al. (2004)]. The extensive expression of mitochondrial genes in white adipose tissue is expected because adipose tissue is dependent on energy metabolism to support anabolic and catabolic lipid metabolism. Mitochondrial genes involved in energy metabolism were highly expressed in porcine adipose tissue, similar to the observation in the rodent adipose tissue (Bolduc et al., 2004). Mitochondrial function with continual synthesis of ATP must be maintained to allow adipocyte metabolic homeostasis. For example, disruption of ATP synthesis in transgenic mice by expressing the mitochondrial uncoupling protein 1 in white adipose tissue results in a decrease in genetic or diet-induced obesity (Kopecky et al., 2001).
Another group of genes abundantly expressed in porcine adipose tissue are genes for ribosomal proteins. The data strongly suggest that there are active protein synthetic activities in porcine adipose tissue. This observation is similar to the results generated by serial analysis of gene expression in rodent adipose tissue (Bolduc et al., 2004).
Finally, a group of FABP and stearoyl coenzyme A desaturase (SCD; AY487830) were highly expressed in porcine adipose tissue and represented a great proportion of this adipose tissue cDNA library (Table 2
). The FABP are important for transporting fatty acids, whereas the SCD enzyme is important because it adds a double bond to C16 and C18 fatty acids to produce the MUFA. The SCD transcript was highly expressed in porcine adipose tissue, and to a much lesser extent, in the liver (Smith et al., 1999; Wang et al., 2004). It also appears early in porcine adipocyte differentiation (Wang et al., 2004).
Porcine Full-Length cDNA
We found there were a large number of full-length cDNA sequences in this library (Table 3). Eleven full-length highly expressed genes were submitted to Gen-Bank (the accession numbers with NM are known porcine genes, and the ones with DQ are newly discovered sequences in the current study), and their accession numbers are indicated in Table 4. The AA sequences of the 11 selected porcine genes were deduced from the cDNA sequences. The degree of homology between the nucleotide and AA sequences of the porcine genes and those from the human and mouse were high (Table 4).
|
). Many of these genes were found to be present in porcine adipose tissue in high concentrations.
|
). In rats, expression of LDH-B mRNA in the soleus muscle is low, but LDH-B represents >75% of the LDH mRNA and protein in the plantaris and extensor digitorum longus muscles (Rossignol et al., 2003). Lactate dehydrogenase is a tetrameric protein catalyzing the conversion of pyruvate to lactate. The LDH tetramer is composed of various combinations of the subunits, varying from total LDH-A to total LDH-B (Markert et al., 1975). In rat adipocytes, lactate production by LDH is associated with lipid accumulation capability (King and DiGirolamo, 1998). Whereas lactate accounts for 34% of total glucose metabolism in porcine adipocytes (Heckler and Carey, 1997), the metabolism of lactate in bovine adipocytes is also important (Smith and Prior, 1986).
The A-FABP (also called FABP4 or aP2) mRNA was highly expressed in adipose tissue, marginally expressed in heart, but not in other tissues (Figure 2B). The A-FABP belongs to a family of homologous proteins present in various tissues. These relatively small proteins (approximately 15 kDa) bind a variety of long-chain fatty acids and presumably prevent the toxicity of nonesterified fatty acids, as well as provide a mechanism for intracellular fatty acid transport. The A-FABP accounts for approximately 1% of the cytoplasmic protein in the adipose tissue of humans (Baxa et al., 1989) and mice (Bolduc et al., 2004). The A-FABP mRNA is highly expressed in developing porcine adipocytes (Ding et al., 1999).
The decorin mRNA was highly expressed in most porcine tissues but had low abundance in the liver (Figure 2C). Decorin is a proteoglycan molecule highly expressed in connective tissues (Scholzen et al., 1994). It binds collagen fibrils and is involved in the organization of the extracellular matrix. Decorin also can cause phosphorylation of the epidermal growth factor receptor with activation of the mitogen-activated protein kinase pathway to regulate the cell cycle and suppress growth. Porcine adipocytes are embedded in a net-like structure of connective tissue, and in addition, individual adipocytes are surrounded by small connective tissue fibrils (Mersmann et al., 1975). The extensive amount of connective tissue present in porcine adipose tissue supports the observation of expression of decorin mRNA.
Heart (H)-FABP (or FABP3) belongs to same family as A-FABP. The 15-kDa H-FABP also binds many types of long-chain fatty acids for intracellular transport and to alleviate toxicity. In rat tissues, H-FABP mRNA is highly expressed in heart, skeletal muscle, and testis with low levels in kidney; there is little expression in liver, spleen, lung, or 3T3-L1 adipocytes (Heuckeroth et al., 1987). The H-FABP mRNA was highly expressed in porcine adipose tissues and in heart, kidney, and liver (Figure 2D).
Glutathione S-transferase 3 (MGST3) is one of the glutathione transfer enzymes involved in detoxification of xenobiotics, drug metabolism, and reduction of the oxidative stress caused by lipid peroxidation. The MGST3 mRNA was highly expressed in most porcine tissues (Figure 2E). Tissue distribution of human MGST3 mRNA (Jakobsson et al., 1997) does not show the relatively high expression in liver, kidney, or lung we observed from pig tissues. A relatively high expression was observed for skeletal muscle, whereas we did not find this; adipose tissue was not included in the survey of tissues in Jakobsson et al. (1997).
Nucleoside diphosphate (NDP) kinase activity, i.e., the conversion of nucleoside diphosphates to triphosphates, was the initial activity observed for the protein nm23. The nm23-H2 represents one of the subunits of the hexameric enzyme composed of several different types of subunits (Gilles et al., 1991; Postel, 1998; Masse et al., 2002). More recently, nm23 has been implicated in regulation of cell growth (Postel, 1998). Tissue distribution of NDP kinase/nm23-H2 in porcine tissues indicated a high degree of expression in kidney and lung; subcutaneous and abdominal adipose tissue exhibited relatively high expression (Figure 2F). Mouse nm23-H2 is highly expressed in kidney, but also in liver and heart (Masse et al., 2002), in contrast to our result with pig tissues. The protein might be expected to have wide distribution because of its important NDP kinase activity.
The S100 calcium-binding protein A6 (calcyclin) mRNA was ubiquitously distributed in all porcine tissues surveyed (Figure 2G). Expression of the mRNA was greatest in lung and least in skeletal muscle. In rats, calcyclin mRNA has high expression in lung and kidney with moderate expression in muscle and low expression in liver and testis (Murphy et al., 1988). The S100 proteins are a family of at least 16 different Ca-binding proteins (Zimmer et al., 1995). The calcyclin promoter lacks any regions determining cell specificity (Lesniak et al., 2000), suggesting ubiquitous tissue distribution, as observed for pig tissues (Figure 2G).
The epidermal (E) -FABP (also called FABP5) belongs to the family of approximately 15 kDa fatty acid-binding proteins found in various tissues. The E-FABP is highly expressed in epidermis (Siegenthaler et al., 1994). It is highly expressed, not only in human skin but also in heart and adipose tissue, but not in liver and kidney. The porcine E-FABP mRNA was highly expressed in skeletal muscle, heart, and kidney (Figure 2H). This member of the FABP family was not highly expressed in adipose tissue in contrast to A-FABP (Figure 2B) and H-FABP (Figure 2D). The expression in porcine liver and kidney sharply contrasts with the human data.
In the current study, the adipose tissue library contained 16% full-length sequences with more than 240 individual genes, including full-length sequences for 40 unknown genes. This EST library almost doubled the number of EST from porcine adipose tissue from the current quantity of 2,283 EST reported in the TIGR Porcine Gene Index. These additional sequences will help to advance the study of gene structure, identification, and function in pigs. The tissue distribution of the 8 selected genes confirms that the abundantly expressed genes in the EST library constructed in the current study are actually highly expressed in porcine adipose tissue. The abundant expression of 3 fatty acid binding proteins suggests active fatty acid metabolism in the adipose tissue. Although the expression pattern of these genes may be affected by sex and age, the older castrated pigs in the current study with fully developed adipose tissue should actively express genes associated with adipocyte and fat accretion. To avoid the contribution of genes from other cell types, a library for pure adipocytes would have been preferable. However, the isolation procedure for adipocytes makes it impossible to harvest adipocytes in normal physiological states. In fact, our unpublished data indicate that mitochondrial sequences represent more than 60% of the total expressed genes in a cDNA library generated from freshly isolated porcine adipocytes. Such libraries may not be more valuable for understanding adipocyte biology compared with the library reported in the current study. Moreover, the successful establishment of a porcine full-length enriched cDNA library provides new molecular tools to study the involvement of the functions of these genes in pigs. Functional expression of these genes, especially the unknown genes, can be used to decipher the functional roles exhibited by the adipocyte under various perturbations via endocrine, environmental, genetic, nutritional, pharmacological, or physiological manipulations.
|
Footnotes |
|---|
2 Visiting scientist, National Taiwan University.
3 Corresponding author: sding{at}ntu.edu.tw
Received for publication December 19, 2005. Accepted for publication May 3, 2006.
|
LITERATURE CITED |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. Gondret, N. Guitton, C. Guillerm-Regost, and I. Louveau Regional differences in porcine adipocytes isolated from skeletal muscle and adipose tissues as identified by a proteomic approach J Anim Sci, September 1, 2008; 86(9): 2115 - 2125. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. H. Yu, S. C. Wu, W. T. K. Cheng, H. J. Mersmann, and S. T. Ding Ectopic expression of porcine peroxisome proliferator-activated receptor {delta} regulates adipogenesis in mouse myoblasts J Anim Sci, January 1, 2008; 86(1): 64 - 72. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. J. Hausman, C. R. Barb, and R. G. Dean Patterns of gene expression in pig adipose tissue: Transforming growth factors, interferons, interleukins, and apolipoproteins J Anim Sci, October 1, 2007; 85(10): 2445 - 2456. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Li, H. N. Zerby, and K. Lee Heart fatty acid binding protein is upregulated during porcine adipocyte development J Anim Sci, July 1, 2007; 85(7): 1651 - 1659. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
