The full-length cDNA of porcine PNAS-4 gene consists of 4,059 bp that contains an open reading frame (ORF) of 582-bp encoding a protein of 194 amino acids with a calculated molecular mass of 21.4 kDa and an isolectric point (pI) of 4.807 (Figure 1). Two polyadenylation signals (AATAAA) were observed at positions 821 to 826 and 2307 to 2312 upstream of the poly (A) stretch. The Brd-Box (AGCTTTA) and GY-Box (GTCTTCC) were found in the 3' untranslated region (UTR) (Figure 1). The mRNA sequence of this gene has been submitted to GenBank [GenBank: DQ435075].

Comparison of the porcine amino acid sequence with the proteins of seven species reported in GenBank shows that the porcine putative protein possesses a high level of identity (94–98%) with mammals and fowls, and it shares 74% identity with Xenopus laevis throughout the protein sequence. The porcine protein contained several conserved motifs such as C-terminal microbody targeting signal (CMTS), N-myristoylation and a Protein kinase C phosphorylation site (Figure 2). In addition, it showed that the porcine PNAS-4 encoded a non-secretory protein without signal peptide and transmembrane regions.

The relative expression of the porcine PNAS-4 transcript in various pig tissues showed highest expression in skeletal muscle (in decreasing order: gastrocnemius, semitendinosus, longissimus dorsi and biceps femoris) and lymph node. Lower expression levels were determined in the liver, large intestine, backfat, heart, small intestine, stomach, spleen, lung, kidney and brain (Figure 3).

Figure 4 showed that the expression level of porcine PNAS-4 seemed to decrease gradually during embryonic development of Landrace and Tongcheng pig breeds, which was significant in Landrace pigs between 33 days post conception (dpc) and 90 dpc. This was consistent with the results of semi-quantitative RT-PCR 7. Curiously, after birth, the amount of porcine PNAS-4 mRNA in skeletal muscles increased drastically (P < 0.01). However, the mRNA expression on the day of weaning (28-day neonate) was lower than the second day after birth and adult (P < 0.05). After maturity, the mRNA expression of porcine PNAS-4 reached a peak level (P < 0.01).

The CGI-146-GFP fusion proteins localized in the Golgi complex (Figure 5), which was identified by the complementation tests using the homo-functional P4-pEGFP-N3 construct. Green fluorescence was detected in the cytoplasm of the control cells transfected with pEGFP vector alone (E-H, Figure 5).

The somatic cell hybrid (SCH) panel revealed that the porcine PNAS-4 gene was assigned to SSC10 (q11–q16) (probability of regional localization: 0.9998, error risk < 0.1%). Two-point analysis of INRA-University of Minnesota porcine radiation hybrid (IMpRH) panel revealed that the porcine PNAS-4 was closely linked to SW497 (LOD = 13.68, 28 cRs) with 22% retention frequencies. This result was consistent with the SCH panel. The corresponding conserved syntenic region on human chromosome was the end of Homo sapiens chromosome 1 long arm (HSC1q)11.

Comparative sequencing revealed 9 and 5 polymorphisms within two PCR amplified fragments [GenBank: DQ406743 and DQ435075] which were parts of PNAS-4 genomic DNA. Because of linkage disequilibrium among several polymorphisms, two single nucleotide polymorphisms (SNPs) were selected for further study. The T1400C SNP within intron 2 was detectable after digestion with Sty I, resulting in a 762-bp PCR amplicon produced allele T (762 bp) and allele C (549 bp and 213 bp). The other SNP, A1813C within the 3'UTR, located more than 20 kb away from the T1400C SNP in genome, was harbored in a 789-bp PCR amplicon and digested by Msp I, resulting in allele A (789 bp) and allele C (606 bp and 183 bp).

Analysis of polymorphism revealed that allele C for both polymorphisms (T1400C and A1813C) was prevalent in four Chinese indigenous pig breeds whereas the allele T (T1400C) and the allele A (A1813C) respectively had higher frequencies in Landrace and Large White (Table 1). The two SNPs could be linked in Landrace, Large White pigs and a pair of three breed-crosses (data not shown). Therefore, we selected one of the two SNPs for association analysis. Three significant associations were observed between the A1813C SNP genotypes and three fat-related traits (Table 2). For the three considered traits, homozygous CC pigs of our experimental population showed higher values than those of the pigs with other genotypes (P < 0.05).

Full-length cDNA sequence analysis revealed high identity of the gene at both the mRNA and the amino acid level with mammalian and avian species. This may indicate both structural and functional conservation of the gene. The Brd-Box and GY-Box within the 3'UTR suggests that the PNAS-4 expression was regulated negatively at the post-transcriptional level, both temporal and spatial 131415.

PNAS-4 mRNA showed a widespread expression with the highest levels in muscle tissue, and could be detected at all times during the skeletal muscle development. In more detail, the expression of PNAS-4 tends to decrease gradually from prenatal day 33 to day 90, an important period for development of muscle fibers and determination of muscle fiber numbers 11617. The muscle fiber number appears to increase as a result of the decrease of PNAS-4 mRNA expression level during prenatal muscle development, which was consistent with that the PNAS-4 is involved in apoptosis8. However, the expression profiles of Tongcheng pigs and Landrace pigs were not complete similar in early embryonic development, and this kind of difference may be related to breed difference 5. The PNAS-4 mRNA expression decreased faster in early Landrace pig development, and this result may be correlated with the high number of primary fibers found in Landrace compared with Tongcheng pigs and other indigenous Chinese breeds 18192021. In other words, there might be a positive correlation between the decrease speed of PNAS-4 mRNA expression level in embryonic stage and porcine muscle fiber number. Muscle fiber number is an important determinant of postnatal growth such that pigs with a high fiber number tend to grow faster than pigs with a lower fiber number 22. This is also supported by the carcass weight of Landrace compared to Tongcheng pigs18. After birth, the expression of PNAS-4 mRNA in muscle tissue was increased, which could be attributed to an increase in length and girth of the muscle fibres and/or muscle function 1. Consistent with this, the highest expression of the gene was recorded in skeletal muscle of mature pigs. The protein was found to localize in the Golgi complex. It was thus deduced that PNAS-4 protein acted more possibly as a regulator than as a structural protein, which was generally in agreement with the opinion that PNAS-4 may be a pro-apoptotic gene 8.

Porcine PNAS-4 gene mapped to q11–16 of SSC10, closely linked to marker SW497. It showed that several quantitative trait loci (QTL) related to growth and carcass traits, such as average daily gain (ADG), Backfat Thickness (BFT), Carcass Weight (CCWT), Ham Weight (HAMWT), are located on this chromosome region in the Pig QTL Database release 232425. Consistent with this, significant associations between A1813C SNP genotypes of PNAS-4 and three fat-related traits were found in our experimental pig populations. Moreover, the three fat traits values of pigs with homozygous CC were higher than that of the pigs with other genotypes (P < 0.05). Our findings have also revealed that allele C of SNP A1813C was prevalent in four Chinese indigenous breeds whereas allele A had higher frequency in Landrace and Large White. Interestingly, western commercial pig breeds (Duroc, Large White and Landrace) are well known for their high growth rate and lean meat percentage 26. These correlated results suggest that the pigs with allele C in SNP A1813C will have higher fat content. In order to confirm it, further investigations in the use of the gene for commercial breeding are recommended.

BLAST in the pig expressed sequence tags (ESTs) database using the human PNAS-4 mRNA sequence [GenBank: NM 016076] as probe, revealed three porcine ESTs [GenBank: CN156918, CV877591 and AJ665351]. To obtain the full-length cDNA, primer set GAPF/GAPR (Table 3) was used to amplify the unknown nucleotide sequence between ESTs CV877591 and AJ665351, and the primer 5GSP (Table 3) in combination with the Universal Primer A Mix (UPM) supplied in the SMART™ RACE Kit (Clontech Inc, Palo Alto, CA, USA) used to amplify the 5'-end of the gene. Rapid amplification of cDNA ends (RACE) PCR was carried out in a PTC-200 thermal cycler (Bio-Rad, Richmond, CA, USA) with touchdown PCR conditions: 3 min at 95°C; 5 cycles of 94°C for 20 s, 72°C for 3 min; 5 cycles of 94°C for 20 s, 70°C for 30 s; 72°C for 3 min; 32 cycles of 94°C for 20 s, 68°C for 30 s; 72°C for 3 min; 5 min at 72°C. The PCR products were purified by agarose gel electrophoresis and sequenced. Functional elements analysis of the putative protein and UTR were performed using the PROSITE program27, SignalP 3.0 server28 and UTRScan29, respectively.

The intracellular distribution of porcine PNAS-4 protein was studied by fluorescence and confocal analysis of pig kidney epithelial cells (PK15) cells transiently transfected with P4-pEGFP-C1 construct. The ORF, encoding porcine PNAS-4 protein, was amplified by PCR using the primers (P4GFPF/P4GFPR, Table 3) designed from its cDNA sequence and subcloned into the Xho I-Hind III site of the pEGFP-C1 vector (BD Biosciences Clontech, CA, USA).

PK15 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (GIBCO-BRL, Grand Island, NY), 4 mM glutamine, 100 U/ml penicillin, and 0.1 mg/ml streptomycin under humidified air containing 5% CO₂ at 37°C and seeded onto cover slips in 6-well plates. The PK15 cells were plated at 5 × 10⁵/ml and transfected with P4-pEGFP-C1/N3 constructs (1 μg of plasmid DNA) using Lipofectamine™ 2000 reagent according to the manufacturer's instructions until they were grown to 80–90% confluence.

Twenty-four hours after transfection, the cells were incubated at 37°C for 30 min in growth medium containing 200 nM MitoTracker Red CM-H₂Xros (Molecular Probes, Eugene, Oregon, USA) for mitochondrial labeling. Then the cells were fixed with pre-warmed growth medium containing 3.7% formaldehyde for 15 min at 37°C. After the final washing steps and incubation with 10 μM Hoechst33342 for 10 min, the slides were mounted, sealed and analyzed by Olympus FluoView™ FV1000 Confocal Microscope. The software FV1000 Viewer was used to generate individual fluorescent pictures as well as overlay pictures that demonstrated the relative distribution of the fusion protein.

The French SCH panel composed of 27 pig/rodent hybrids 31 was used for the regional assignment of PNAS-4 gene on the SSC. To more precisely map genes, the IMpRH panel consisting of 118 hybrid clones obtained by fusing irradiated pig cells with the recipient hamster cell was employed 32. The PCR was performed in a total volume of 10 μL 1 × PCR buffer (TaKaRa), containing 25 ng panel DNA, 0.3 μM each primer (MapF, MapR, Table 3), 75 μM each dNTP, 1.5 mM MgCl₂, and 1.0 U Taq DNA polymerase (TaKaRa). The PCR cycling conditions were: Initial denaturation at 95°C for 3 min, followed by 32 cycles (20 s at 94°C, 30 s at 65°C and 30 s at 72°C). Statistical analysis of SCH PCR results were performed using the program 33available for regional assignment of genetic markers. The results of radiation hybrid PCR products were analysed with the IMpRH mapping tool 34.

Four Chinese indigenous pig breeds (Xiang (N = 42), Bamaxiang(N = 45), Wuzhishan (N = 43) and Tongcheng (N = 44) pigs) and two western white breeds (Large White (N = 215) and Landrace (N = 15)) were used to analyze genetic variability. Association analyses were performed in our experimental populations that contained three pure breeds, Tongcheng (N = 44), Landrace (N = 15) and Large White^a (N = 19), and a pair of three-breed cross, Landrace × (Large White × Tongcheng) (N = 44) and Large White × (Landrace × Tongcheng) (N = 35). Growth traits (average daily gain), carcass and meat quality traits (live weight at slaughter, carcass length, dressing percent, loin muscle area, meat color, marbling, drip loss, 0 h loin muscle PH and shear force) were studied. Fat traits studied were percentage of leaf fat, percentage of leaf and caul fat, and backfat thickness between 6th and 7th ribs.

Two pairs of primers (SNPF1/SNPR1 located in intron 2 and SNPF2/SNPR2 located in 3'UTR) were designed using the submitted sequences [GenBank: DQ406743 and DQ435075]. The PCR products were cloned into the T-easy vector and sequenced to identify the polymorphisms in the porcine PNAS-4 gene.

Associations between growth traits, carcass traits and meat quality traits on the one hand and genotype on the other hand were studied using a two-step analysis method. Briefly, in the first step, a preliminary general linear model without the genotype information was used to eliminate system effects, including sex, combination and batch, using the SPSS program with the following model:

Y_ijkl = μ + C_i + B_j + S_k + (CB)_ij +(BS)_jk + e_ijkl

where Y_ijkl is phenotypic value of the target trait, μ is the population mean, C_i is the combination (including breed) effect, B_j is the batch effect, S_k is the sex effect and e_ijkl is random error effect for each observation. In addition, the interaction effects between combination and batch (CB)_ij, and between batch and sex (BS)_jk, were also considered. In the second step, the resulting standardized residual values were used to partition the effects of the genotypes using the one-way analysis of variance (ANOVA) model. Simultaneously, significant difference tests between different genotypes were performed with Bonferroni t tests 35.