Adiponectin, an adipokine secreted from adipose tissue (AT), exerts beneficial pleiotropic effects on obesity-related metabolic diseases. We have analyzed the adiponectin gene (ACDC) and its expression in two genetically different breeds of pigs, lean type, large white (LW) and fat type, Casertana (CE). DNA, RNA, and protein extracts from 10 LW and 10 CE pigs were analyzed by sequence analysis, enzyme-linked immunosorbent assay (ELISA), fast protein liquid chromatography, and northern and western blotting. Sequence analysis revealed an identity of 100% between the ACDC gene from the two breeds, but the expression of the adiponectin protein was higher in LW than in CE pigs. We identified sexual dimorphism of adiponectin in both breeds, namely a balanced distribution of the low isoforms (∼50 kDa), whereas the middle isoforms (∼75–150 kDa) were increased in sows. In conclusion, in this study, we demonstrate that adiponectin is produced and secreted differently in the two breeds of pig, namely adiponectin is more abundant in LW than in CE. Moreover, the visceral AT of LW expresses more adiponectin than the subcutaneous AT. This relationship is absent in CE. These observations provided the first evidence that adiponectin expression is correlated with the “fat” phenotype in pig.
Besides being an energy storage organ, adipose tissue (AT) is an important endocrine organ involved in the regulation of insulin action, inflammation, hemostasis, and other physiological processes (1,2). In fact, AT synthesizes a series of cytokines that affect the functioning and the structural integrity of other tissues (2). Among these adipocytokines, adiponectin has attracted attention because it exerts beneficial pleiotropic effects on a cluster of obesity-related metabolic and cardiovascular disorders (3,4). The hormone adiponectin is secreted at high levels by AT and accounts for 0.01% of the total serum proteins. It improves hepatic insulin sensitivity, increases fuel oxidation, and decreases vascular inflammation (4). The human adipocyte C1q and collagen-domain-containing gene (ACDC) maps to the 3q27 region, which harbors various quantitative trait loci for the metabolic syndrome and a locus for type 2 diabetes (5,6). ACDC contains three exons that encode a 244-amino acid protein, adiponectin. It circulates at very high levels in the blood as different molecular weight isoforms produced by multimerization of the 30-kDa monomer (7,8). The basic unit of adiponectin is a trimer that, by disulfide bonds, forms the more complex high-molecular weight isoforms of adiponectin. A sexual dimorphism in terms of adiponectin total levels and of its higher complexes was observed in both humans and rodents (9,10,11). On the other hand, experimental and clinical data indicate that the oligomeric complex distribution of adiponectin is critical for the antidiabetic and antiatherogenic activity of this hormone (12). Molecular investigations yielded strong evidences that genetic variants in the human ACDC gene may be considered susceptibility factors for insulin resistance, type 2 diabetes, obesity, and severe obesity (3,4,13). Adiponectin has been found to be an insulin-sensitizer that exerts antidiabetic, antiatherogenic, anti-inflammatory and cardioprotective activities in various animal models (14,15,16,17). Pig, among the large animals, is a good model for the study of human disorders, such as obesity, diabetes, and cardiovascular diseases because they have a similar pathological response to high-caloric intake as humans (18). The porcine adiponectin has been recently cloned and its expression is partially evaluated (19,20,21).
In this study, we characterized adiponectin expression in two genetically different breeds of pigs, the large white (LW), a cosmopolitan lean type and the Casertana (CE), an endangered breed native of Campania (Southern Italy) chiefly raised half-wild. CE and LW exhibit an opposite genetic behavior with respect to the energy metabolism. Studies with CE pigs may shed light on the physiological process of fat deposition because these animals are prone to adipogenesis and have a strong aptitude for fat deposition. In fact, CE pigs have a higher percentage of body fat and produce more than double backfat thickness as LW pigs (22,23).
Methods and Procedures
This study involved 10 LW and 10 CE pigs. The male pigs were castrated. The diet comprised concentrated commercial feed administered “semi ad libitum” in four daily meals; water was provided using automatic feeding troughs. The LW pigs were slaughtered at 11 months of age and the CE animals at over 12 months. Animals were not fed the night before they were slaughtered. Jugular blood samples were collected during slaughtering. Subcutaneous AT samples between the 3rd and 4th lumbar vertebrae were freshly collected and immediately frozen in liquid nitrogen then stored at −80 °C to extract RNA and proteins. Visceral perirenal AT samples near the last rib were also collected.
Two milliliters of blood were immediately centrifuged at room temperature. Plasma glucose, total cholesterol, high-density lipoprotein-cholesterol, and triglycerides were determined using standard enzymatic methods (Hitachi Modular, Roche, Europe) (24). Serum insulin was measured using chemiluminescence method (Immunolite 2000; Medical System, Genova, Italy).
ACDC sequence analysis
Genomic DNA, from each animal, was extracted using a standard salting out/ethanol precipitation protocol from peripheral blood leukocytes. The ACDC exonic regions (GeneBank accession number NM004797) were amplified by PCR using 5′-CGAGAAGCCTGGAGCACTAC-3′ and 5′-GCTTTGTTCCTTCCCTGTGA-3′ for exon 1 and 5′-CTGTAAGTCAAGGAGGCTGT-3′ and 5′-TCAGTCTCCTAATGACACTG-3′ for exon 2. The reaction was carried out as described previously (25). The AT cDNAs were amplified by reverse transcription-PCR with GeneAmp RNA PCR Kit (Applied Biosystems, Mouza, Italy) using 5′-GCTCAGGATGCTGTTGTTGG-3′ and 5′-TGGTGGAGGCTCTGAGTTGG-3′. Sequence analysis of the PCR products was performed on both strands with an automated procedure using the 3100 Genetic analyzer (Applied Biosystems, Mouza, Italy).
Northern blotting analysis
Total RNA was prepared from visceral and subcutaneous ATs using TRIzol reagent (GIBCO BRL Life Technologies, Roskilde, Denmark) (26). Ten microgram of total RNA was electrophoresed and transferred to nylon membrane (Hybond-N; Amersham Biosciences, Milan, Italy). The membranes were processed in rapid hybridization buffer (Amersham Biosciences, Milan, Italy). An EcoRII cDNA fragment (from nucleotide c.71 to c.128) was used as probe. Porcine α-actin served as an internal control. The probes were labeled with [α-32P]2′-deoxycytidine-5′-triphosphate (3,000 Ci/mmol) (Amersham Biosciences, Milan, Italy) using Ready-To-Go kit (Amersham Biosciences). This analysis was performed in duplicate.
Production of polyclonal antibody and ELISA
A human adiponectin amino acid region (H2N-ETTTQGPGVLLPLPKG-COOH) was used to immunize two rabbits; adiponectin antiserum was precipitated with ammonium sulfate and the antibodies were purified on a protein A-Sepharose column (Custom antibody; PRIMM, Milan, Italy). The porcine serum adiponectin concentrations were measured by enzyme-linked immunosorbent assay (ELISA) method. ELISA plates were coated with serial dilutions of serum (100 μl) in carbonate-bicarbonate buffer (0.1 mol/l, pH 9.6) in order to determine the absorbances in linear range. Furthermore, in order to control for the possible background of the technique, the antibody step was removed and checked that the absorbances achieved were <0.15 (ref. 27). A human recombinant adiponectin was used as the standard (Phoenix Pharmaceuticals, Burlingame, CA). The experiments were performed in triplicate on each animal.
Validation of antibody assays by electrophoresis fractionation and in situ digestion
The adiponectin protein was analyzed using sodium dodecyl sulfate-polyacrylamide gel electrophoresis and the corresponding Colloidal Coomassie stained band was excised and completely destained in 50 mmol/l ammonium bicarbonate pH 8.0, in 50% acetonitrile. The gel piece was resuspended in 50 mmol/l ammonium bicarbonate, pH 8.0, containing with 100 ng of trypsin and incubated for 2 h at 4 °C and overnight at 37 °C. The supernatant containing the peptide mixtures was removed and directly analyzed by LC-MS/MS using an LC/MSD Trap XCT Ultra equipped with an 1100 HPLC system and a Chip Cube (Agilent Technologies, Palo Alto, CA). After loading, the peptide mixture (8 μl in 0.2% formic acid) was concentrated and washed at 4 μl/min in a 40-nl enrichment column with 0.1% formic acid as eluent. The sample was then fractionated on a C18 reverse-phase capillary column (75 μm × 43 mm) onto the reported above CHIP device at a flow rate of 200 nl/min, with a linear gradient of eluent B (0.2% formic acid in 95%) in eluent A (0.2% formic acid in 2% acetonitrile) from 7 to 60% in 50 min. Peptide analysis was measured using data-dependent acquisition of one MS scan (mass range from 400 to 2,000 m/z) followed by MS/MS scans of the three most abundant ions in each MS scan. Raw data from nanoLC-MS/MS analyses were introduced into the MASCOT software (Matrix Science, Boston, MA) to search for a nonredundant protein database.
Western blot analysis
Porcine serum was purified using Aurum Affi-GEL Blue columns (Bio-Rad, Hercules, CA). The proteins were quantified by Bradford's method (Bio-Rad, Hercules, CA). Total proteins were prepared using TRIzol reagent (GIBCO BRL Life Technologies) (26).
Ten microgram of purified serum were lysed in Laemmli buffer with and without dithiothreitol 10 mmol/l (28). The gel was exposed to high-performance autoradiography film (Amersham Biosciences, Milan, Italy), digitalized with a scanner (1,200 dpi) and analyzed by densitometry with the Jasc Paint Shop Pro 7.00 software. All experiments were performed in triplicate for all animals.
Gel filtration analysis
The distribution of adiponectin oligomers in porcine sera was analyzed by gel filtration chromatography on a Superdex 75 10/300 GL column connected to a fast protein liquid chromatography system (Amersham Pharmacia Biotech, Milan, Italy). Two hundred microliter of serum samples, prepared as above reported, were fractionated at 0.5 ml/min using a 100 mmol/l Tris-HCl, pH 7.8 elution buffer containing 100 mmol/l KCl. Fractions (500 μl) were collected and the concentrations of the adiponectin isoforms were measured using western blot.
Data are expressed as means ± s.e. We used the nonparametric Mann-Whitney test to compare the means of the various groups. All statistical analyses were performed using SPSS (v.13.0) for Windows software (SPSS Headquarters, Chicago, IL).
ACDC gene, cDNA, and adiponectin mRNA in ATs of LW and CE pig breeds
Sequence analysis was performed on genomic and cDNA from both LW and CE animals: no nucleotide changes were detected between the two pig breeds (Figure 1a). Next, we measured the mRNA adiponectin expression in the visceral and in the subcutaneous ATs of the two animals from the two breeds of pigs using northern blot analysis, and detected a predominant transcript of ∼3.0 kb. Adiponectin mRNA levels differed greatly between the two breeds. Moreover, the expression of adiponectin mRNA was greater in LW ATs than in CE ATs (Figure 1b, lanes 1 and 2 vs. lanes 3 and 4). Furthermore, in LW ATs, adiponectin mRNA was higher in visceral tissue than in subcutaneous tissue (Figure 1b, lanes 1 and 2), whereas in CE, the amounts of mRNA were slightly lower in visceral tissue than in subcutaneous tissue (Figure 1b, lanes 3 and 4). Additional, less-abundant transcripts of ∼1.7, 1.3, and 0.9 kb were detected in visceral and subcutaneous tissues from LW (Figure 1b, lanes 1 and 2). The three smaller transcripts were absent from the ATs of CE pigs. Adjustment of each mRNA level from ATs with α-actin produced similar results (Figure 1c).
Serum levels of total adiponectin and of its isoforms
Table 1 shows the metabolic features of the animals analyzed. We found no association between breed and fasting plasma insulin and glucose concentrations. We measured serum concentrations of adiponectin using ELISA test in the two groups of animals: LW representing the “lean” phenotype and CE representing the “fat” phenotype. As shown in Table 1, the serum level of adiponectin was higher (P < 0.01) in LW (9.79 ± 0.8 g/ml) than in CE (7.6 ± 1.2 g/ml). Moreover, total adiponectin levels were higher in sows (9.2 ± 1.4 g/ml) than in the castrated male group (8.0 ± 1.5 g/ml) (P = ns). No correlation was found between serum adiponectin and glucose or insulin concentrations (data not shown).
Table 1. Phenotypes and morphological features of LW and CE pigs
The ELISA results were confirmed by western blot analysis after heat denaturation. As shown in Figure 2, the total amounts of adiponectin were higher (P > 0.02) in LW (384,660.7 ± 5,189.4 pixels) than in CE (256,771.7 ± 8,027.1 pixels). This analysis also revealed that the higher molecular structures (MW > 250 kDa) of adiponectin were slightly more abundant in LW than in CE (54,057.0 ± 63.7 pixels vs. 27,618.7 ± 574.8 pixels, respectively) (Figure 2a). A fast protein liquid chromatography analysis, followed by western blot, confirmed the differences of adiponectin complexes between LW and CE pigs (Figure 2b).
Sexual dimorphism of pig adiponectin
Figure 3 shows the analysis of serum samples from males and sows of the two pig breeds using western blot analysis after denaturation with dithiothreitol. The total adiponectin and the isoforms pattern differed between the genders in both breeds. In fact, LW and CE sows had more total adiponectin and a more complex adiponectin pattern (containing additional major multimers of ∼135 and ∼60 kDa) than the castrated males of the same breed. There was also a quantitative difference between the two breeds: LW males had more abundant total adiponectin than CE males (88.381 pixels vs. 66.772 pixels, respectively). A similar trend was observed in sows (113.206 pixels vs. 106.302 pixels, respectively). Most of the adiponectin in male pigs consisted in the smaller, dimeric and trimeric structures (Figure 3, lanes 1 and 3), whereas adiponectin complexes were more evenly distributed in sows. In addition, LW sows had two adiponectin isoforms that were not present in CE sows (Figure 3, lanes 2 vs. 4).
Adiponectin expression in ATs of LW and CE pigs
Figure 4 shows the western blot analysis performed on the visceral and subcutaneous ATs from LW and CE pigs. Total amounts of adiponectin (i.e., the sum of different protein bands) in visceral AT of LW were slightly higher than in the CE counterpart tissue (342,618 vs. 307,320 pixels); an inverse ratio was observed in subcutaneous AT (120,534 vs. 323,483 pixels). In the ATs of both breeds, adiponectin was essentially present in low-molecular weight isoforms (∼75 and ∼50 kDa); lesser amounts of higher molecular weight structures (≥150 kDa) were present. The pattern of adiponectin isoforms differed in the visceral AT of the two breeds: the visceral AT of LW contained high amounts of the ∼50-kDa isoform and low amounts of the ∼75-kDa isoform (Figure 4, lane 1). Differently, in the visceral AT of the CE, the proportion of the ∼50 kDa and the ∼75-kDa isoforms was reversed (Figure 4, lane 2). In the subcutaneous AT, adiponectin was found essentially as the ∼50-kDa isoform albeit in larger amounts in CE vs. LW (Figure 4, lanes 3 and 4). A low amount of aggregates (≥150 kDa) was present in the ATs of both breeds. Furthermore, in LW, ∼50- and ∼75-kDa levels were higher in visceral than in subcutaneous AT (Figure 4, lanes 1 and 3); in CE, ∼75-kDa isoform levels were higher in the visceral than in subcutaneous AT, whereas ∼50-kDa isoform levels were more abundant in the subcutaneous than in the visceral AT (Figure 4, lanes 2 and 4).
Obesity-related diseases represent an important health problem worldwide (29). One of the causes of these diseases may be the accumulation of visceral fat that abnormally expresses a variety of adipocytokines, such as adiponectin. Adiponectin is a key hormone because it shows pleiotropic biological functions and exhibits protective properties against obesity-related metabolic disorders. Moreover, studies on various animal models have demonstrated that adiponectin is an insulin-sensitizer with potent antidiabetic, antiatherogenic, anti-inflammatory, and cardioprotective activities (3,4). Therefore, adiponectin is a likely target for treatment and/or prevention of these diseases, in particular among obese subjects with visceral fat accumulation.
In this study, we have investigated adiponectin expression in the pig because this animal's response to obesity is similar to that of humans. In addition, pig has several advantages: homogeneous feeding regime, and the absence of confounding factors typical of humans, such as smoking or alcohol drinking (18,30). We compared adiponectin expression in two different genetic breeds of pig, LW and CE which have an opposite genetic behavior in energy metabolism. In addition, CE was dismissed in the past century, and has not been intensively and rationally selected for lean production. Hence, CE retains the feral energy storage genes (30).
To look for differences in the expression of adiponectin between the two breeds, we first analyzed and compared the ACDC gene founding an identity of 100%. Therefore, ACDC nucleotide alterations and/or polymorphisms, which in humans have been regarded as susceptibility factors to insulin resistance and type 2 diabetes, did not found in the pig (3,4,13). Instead, expression studies demonstrated that adiponectin is differently produced and secreted in the two breeds of pig with the lean type (LW) displaying higher levels than the fat type (CE). Our study also demonstrates a qualitative difference in mRNA between the two breeds of pig. In fact, three additional transcripts (∼1.7, ∼1.3, and ∼0.9 kb) were detected only in the LW. Previous studies of pig demonstrated the presence of these three smaller transcripts only in the over-cervical spine dorsal AT, whereas in vitro cultured adipocytes expressed only ∼3.0-kb mRNA (31). Three similar transcripts have been identified in rat ATs (32) but not in mouse (5). The smaller, less-abundant transcripts have not yet been characterized but their specific presence in cell types, tissues, or breeds indicates that they may be the result of alternative splicings that reflect cell/tissue/breed specificity.
The LW and CE pigs also differed in adiponectin secretion and in the expression pattern of its isoforms, with the lean type (LW) displaying larger amounts of adiponectin and in particular of its high-molecular weight isoforms than the fat type (CE). Lower amounts of adiponectin and in particular of the higher molecular weight isoforms have been associated with type 2 diabetes, metabolic syndrome and obesity (33). Weight loss or treatment with the insulin-sensitizing drug rosiglitazone preferentially elevates the high-molecular weight form of adiponectin (34), and the rare G84R and G90S mutations, which are associated with insulin resistance and type 2 diabetes, determine a reduction of the higher molecular isoforms (6,34).
We also found a sexual dimorphism in pig regarding total serum adiponectin levels and the adiponectin isoform pattern. In fact, ∼50-kDa forms had a more balanced distribution whereas the middle order structures (∼75–150 kDa) were increased in sows with respect to castrated males. A similar dimorphism has been observed in humans and in rodents (9,10,11).
Our study also demonstrated differences in the expression pattern of adiponectin between the two breeds at level of fat depots. In fact, in LW, the visceral tissue expressed more adiponectin than subcutaneous tissue, whereas in CE, the subcutaneous tissue produced more adiponectin than visceral tissue. Higher amounts of adiponectin mRNA have been found in visceral tissue vs. subcutaneous tissue in different breeds of pig (20) and in lean rats (35). These results indicated that visceral fat plays a key function in the production and secretion of adiponectin. Studies demonstrated that AT depot differences play an important role in relation to health risk in humans (1). However, data on adiponectin expression in relation to AT depot differences are conflicting. In lean and obese white women, ACDC expression is lower in visceral AT than in subcutaneous AT (36). In contrast, adiponectin mRNA levels in nondiabetic Asian women did not differ between the two ATs (37). In nondiabetic controls, subcutaneous tissue expresses greater concentrations of adiponectin than visceral fat; in the diabetic state subcutaneous tissue expresses less adiponectin irrespective of adiposity (38). In our study, we focused adiponectin expression in the CE pigs, which are characterized by a slower growth and higher backfat thickness than LW pigs (22).
The functional mechanisms governing the pleiotropic functions of adiponectin remain poorly understood. Further studies should focus on elucidating the molecular and structural basis of adiponectin. In this study, we found that the adiponectin expression differs between two genetically different breeds of pigs both at message and protein levels with the lean type (LW) displaying higher levels than the fat type (CE). We also show that adiponectin expression in LW is higher in the visceral fat than in the subcutaneous AT, and that this relationship is lost in the fat CE pig. In conclusion, we provide the first evidence of a correlation of adiponectin level and the obese phenotype in pigs.
This work was supported by grants from MIUR (PRIN 2004), from University of Molise, and CEINGE Biotecnologie Avanzate. We dearly thank Dr Cardillo for helpful discussions. We are grateful to Jean A. Gilder for revising and editing the text.