1. ^*Department of Animal Sciences, Washington State University, Pullman, Washington
2. ^†Department of Animal Science, University of Ljubljana, Domzale, Slovenia
3. ^‡School of Molecular Biosciences, Washington State University, Pullman, Washington

Correspondence: Zhihua Jiang Department of Animal Sciences, East Wilson Road, Washington State University, Pullman, WA 99164-6351. E-mail: jiangz@wsu.edu

^**The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received 31 March 2007; Accepted 4 May 2007.

Abstract

Obesity and type 2 diabetes constitute leading public health problems worldwide. Studies have shown that insulin resistance affiliated with these conditions is associated with skeletal muscle lipid accumulation, while the latter is associated with mitochondrial dysfunctions. However, the initiation and regulation of mitochondrial biogenesis rely heavily on 1000 nuclear-encoded mitochondrial regulatory proteins. In this study, we targeted the ubiquinol-cytochrome c reductase core protein I gene, a nuclear-encoded component of mitochondrial complex III, for its association with subcutaneous fat depth (SFD) and skeletal muscle lipid accumulation (SMLA) using cattle as a model. Four promoter polymorphisms were identified and genotyped on 250 Wagyu Limousin F₂ progeny. Statistical analysis revealed that two completely linked polymorphic sites, g.13487C>T and g.13709G>C (r² = 1), were significantly associated with both SFD (p < 0.01) and SMLA (p < 0.0001). The difference between TTCC and CCGG haplotypes was 0.178 cm for SFD and 0.624 scores for SMLA. Interestingly, the former haplotype produced higher promoter activities than the latter by 43% to 49% in three cell lines (p < 0.05). In addition to Rett syndrome and breast/ovarian cancer observed in other studies, we report evidence for the first time, to our knowledge, that overexpression of ubiquinol-cytochrome c reductase core protein I might affect mitochondrial morphology and/or physiology and lead to development of obesity and related conditions.

Both obesity and type 2 diabetes (T2D)¹ are global public health problems, and their prevalence will increase dramatically over the coming decades. The rise in obesity has been matched by a rise in diabetes in all ethnic groups in the United States (1). The Third National Health and Nutrition Examination Survey (2) found that approximately two thirds of adult men and women in the United States diagnosed with T2D have a BMI of 27 kg/m². It is well known that a core characteristic of patients with obesity and T2D is an increase in insulin resistance, while many studies have indicated that intramyocellular accumulation of triglycerides is a major contributor to insulin resistance (3). Interestingly, mitochondrial dysfunction may predispose an individual to intramyocellular lipid accumulation. However, because of the limited protein-coding capacity of mitochondria, the initiation and regulation of mitochondrial biogenesis rely heavily on 1000 nuclear-encoded mitochondrial regulatory proteins (4). The majority of mitochondrial proteins are nuclear encoded, synthesized in the cytosol, and are post-translationally imported into mitochondria. Therefore, most inherited mitochondrial diseases are reported to be caused by mutations in nuclear-encoded mitochondrial genes.

Among a large number of reactions occurring in mitochondria, probably the most impressive of these is oxidative phosphorylation, in which five multisubunit complexes cooperate to generate most of the cell's energy. Among them, the ubiquinol-cytochrome c reductase complex or complex III is an oligomeric enzyme that catalyzes transfer of electrons from coenzyme QH2 to ferricytochrome c with the coupled translocation of protons across the mitochondrial inner membrane (5). The bovine heart mitochondrial complex III has been well characterized and is composed of 11 subunits, including 10 nuclear-encoded subunits and 1 mitochondrial-encoded subunit (6). In this study, we determined genomic organization of the bovine ubiquinol-cytochrome c reductase core protein I (UQCRC1), a nuclear-encoded component, and developed four genetic markers in its promoter region. Statistical analysis using a general linear model (GLM) and quantitative transmission-disequilibrium test (QTDT) revealed that promoter polymorphisms are significantly associated with both subcutaneous fat depth (SFD) and skeletal muscle lipid accumulation (SMLA) in Wagyu Limousin F₂ cross cattle. The result implies that genetic polymorphisms in UQCRC1 gene might explain some cases of obesity in humans.

The cDNA sequence of the bovine UQCRC1gene was cloned 16 years ago (7). Alignment between the cDNA (NM 174629) and its genomic DNA contig (AAFC03053028) retrieved from the bovine whole genome shotgun sequence indicated that, like its human ortholog, the bovine gene consists of 13 exons (Figure 1A). Screening of genetic polymorphisms on six Wagyu Limousin F₁ bulls detected four single nucleotide polymorphisms (SNPs) in the promoter region: AAFC03053028.1:g.13487C>T, g.13671T>C, g.13709G>C, and g.13725A>G, respectively. The minor alleles among these four SNPs are T, T, C, and A, with a frequency of 0.299, 0.297, 0.299, and 0.079, respectively. Sequencing on 250 F₂ progeny indicated that all four SNPs fall into Hardy-Weinberg equilibrium (p > 0.05).

Figure 1:.

Genomic organization of the bovine UQCRC1 gene and haplotype analysis of its promoter polymorphisms. (A) The bovine gene contains 13 exons and 12 introns. The size of each exon is as follows: exon 1, 84 bp (including 5' untranslated region); exon 2, 141 bp; exon 3, 87 bp; exon 4, 130 bp; exon 5, 199 bp; exon 6, 80 bp; exon 7, 116 bp; exon 8, 144 bp; exon 9, 161 bp; exon 10, 86 bp; exon 11, 89 bp; exon 12, 76 bp; exon 13, 191 bp (including 3' untranslated region). The size of each intron is as follows: intron 1, 251 bp; intron 2, 1708 bp; intron 3, 1057 bp; intron 4, 183 bp; intron 5, 530 bp; intron 6, 1950 bp; intron 7, 211 bp; intron 8, 170 bp; intron 9, 92 bp; intron 10, 294 bp; intron 11, 304 bp; intron 12, 434 bp. (B) Pairwise linkage disequilibrium relationship for four mutations is shown based on r² measurements in percentage (e.g., r² = 100% between g.13487C>T and g.13709G>C).

Full figure and legend (76K)

Initial sequencing of the promoter region on six Wagyu Limousin F₁ bulls indicated that both g.13487C>T and g.13709G>C form two haplotypes: CG and TC. The HAPLOVIEW analysis on genotype data of all F₂ progeny further confirmed the no historical recombination status between these two SNPs with a r² value of 1 (Figure 1B) (8). Two markers, AAFC03053028.1:g.13671T>C and g.13725A>G, are still segregating in the population. In particular, the linkage was hardly detected between g.13725A>G and three other SNPs because of logarithm of the odds scores of <3.0. A total of four haplotypes among these four SNPs were identified in the population using the HAPLOVIEW program, including CCGG, TCCG, CTGG, and CCGA, with a frequency of 0.325, 0.299, 0.297, and 0.079, respectively.

Because both SNPs, g.13487C>T and g.13709G>C, have no historical recombination events in the population, three tagging SNPs, g.13487C>T, g.13671T>C, and g.13725A>G, were used in the association analysis. Overall, the F₂ population had an average SFD of 1.001 0.457 (standard deviation) cm. Both GLM analysis and a QTDT revealed that only SNP g.13487C>T was significantly associated with SFD in the population (p = 0.0040 for GLM analysis and p = 0.0022 for QTDT; Table 1). The CC animals had 0.178 cm of SFD less than the TT animals and 0.170 cm less than the CT heterozygotes. Both TT and CT animals contained equal amounts of SFD, indicating that the allele T has a dominant effect over the allele C (Table 1).

Table 1 - Associations of UQCRC1 promoter polymorphisms with SFD and SMLA in a Wagyu Limousin reference population.

Full table

Overall, all F₂ progeny had an average marbling score of 5.916, which is a subjective, visual appraisal of the fat on a meat cut surface. The polymorphic site g.13487C>T showed an extremely significant association with the trait (p < 0.0001 for both GLM and QTDT; Table 1). Animals with the CC genotypes had marbling scores that were 0.624 and 0.559 lower than animals with TT and CT genotypes, respectively. Again, allele T is dominant to allele C, but by increasing the fat deposition in muscle. Interestingly, GLM analysis indicated marker 13671T>C approaching significance, but the QTDT further confirmed that this SNP was also significantly associated with SMLA (Table 1). The difference in marbling scores was 0.428 between TT and CC homozygotes, which also approaches the significance level (p = 0.0813). No significant association was observed between g.13725A>G and SMLA in the population (Table 1). In humans, the fat stored in muscle is classified into intramyocellular (IMCL) and extramyocellular (EMCL) lipid content. By definition, the muscle lipid accumulation measured by marbling score in this study would mostly represent the EMCL content, because the IMCL cannot be observed by unaided eyes. However, most methods for quantifying IMCL content, such as computed tomography, magnetic resonance imaging, magnetic resonance spectroscopy, and biochemical analysis cannot truly separate IMCL and EMCL (9). Furthermore, both IMCL and EMCL contents in human subjects are highly correlated with each other (r = 0.68) (10). Both measurements are also highly correlated with percent total body fat (r = 0.69 for IMCL and r = 0.66 for EMCL), BMI (r = 0.67 for IMCL and r = 0.68 for EMCL), visceral fat (r = 0.73 for IMCL and r = 0.86 for EMCL), and insulin-to-glucose ratio (r = 0.72 for IMCL and r = 0.68 for EMCL) (10). Nevertheless, how both IMCL and EMCL contribute to development of obesity in humans needs to be further determined.

As indicated above, the polymorphic site g.13487C>T was significantly associated with both SFD and SMLA in the population (Table 1), and it had no historical recombination with g.13709G>C (Figure 1B). Therefore, only two haplotype constructs, T-C-C-G and C-C-G-G—which are different at both g.13487C>T and g.13709G>C sites but the same at g.13671T>C and g.13725A>G sites—were used to study how these associated mutations affect promoter activities of the bovine UQCRC1 gene in three cell lines. Overall, the former construct produced higher promoter activities than the latter construct by 47% in H1299 cells (p = 0.0073), 49% in HCT116 cells (p = 0.0197), and 43% in Cos7 cells (p = 0.0185; Figure 2 A–C). In H1299 cells, the average firefly luciferase activity was 28,133 2782 for the T-C-C-G haplotype and 19,146 1903 for the C-C-G-G haplotype (Figure 2A). In HCT116 cells, the former haplotype resulted in an average of 10,713 3046 relative luciferase units, whereas the latter haplotype yielded an average of 7183 1250 relative luciferase units (Figure 2B). Cos7 cells had the lowest promoter activities, but the difference between the two haplotypes was still significant (1627 208 for the T-C-C-G construct and 1140 167 for the C-C-G-G construct; Figure 2C).

Figure 2:.

Affects of haplotypes on promoter activity in the H1299 cells (A), HCT116 cells (B), and Cos7 cells (C).

Full figure and legend (99K)

As indicated above, UQCRC1 encodes a subunit of mitochondrial respiratory complex III, which operates through a Q-cycle mechanism that couples electron transfer to generation of the proton gradient that drives adenosine triphosphate synthesis. Recent studies have shown that overexpression of UQCRC1 might affect mitochondrial morphology and/or physiology and, thus, cause mitochondrial dysfunction and diseases. In the Mecp2-null mouse, an animal model for Rett syndrome, Kriaucionis et al. (11) found that UQCRC1 was significantly up-regulated in early- and late-symptomatic brains. UQCRC1 overexpression correlated positively with symptom severity and with a significant increase in mitochondrial respiratory capacity and a reduction in respiratory efficiency. In humans, UQCRC1 was highly expressed in breast (74% ) and ovarian tumors (34% ) (12). In this study, we observed that the haplotype that produced the higher promoter activity was also associated with an increase of both SFD and SMLA in the cattle model. Compared with the C-G haplotype at g.13487C>T and g.13709G>C sites, the T-C haplotype yielded 43% to 49% more promoter activity (Figure 2). As well, the animals with the T-C haplotype had a marbling score that was 0.624 higher and 0.178 cm more SFD than the animals with the C-G haplotype (Table 1). Therefore, our study, for the first time, to our knowledge, provides evidence for a possible involvement of UQCRC1/complex III in the regulation of energy metabolism and balance.

Our previous study confirmed a conserved segment of 12 Mb from CLEC3B (C-type lectin domain family 3, member B) to ERC2 (ELKS/RAB6-interacting/CAST family member 2) between human chromosome 3p22.3-p14.3 and bovine chromosome 22q24 (13). Just recently, Harder et al. (14) found this region harbors quantitative trait loci for the persistency of fat yield and the persistency of milk energy yield in dairy cattle using 16 paternal half-sib families with a total of 872 bulls. The quantitative trait loci surround the lactotransferrin gene, whereas the UQCRC1 is located 2 Mb apart. Therefore, this could be another case to support the involvement of UQCRC1 gene in fat deposition and energy production. In addition to the UQCRC1 gene presented here, we found two other nuclear-encoded mitochondrial genes, mitochondrial transcription factor A and fatty acid binding protein 4, were associated with both beef marbling score and SFD in the same population of cattle (15, 16). The current human obesity gene map also displayed 48 genes that are nuclear-encoded mitochondrial genes (17). Overall, mutations in nuclear mitochondrial genes have been shown to lead to oxidative stress, neurodegenerative diseases, and metabolic disorders. Therefore, how nuclear-encoded mitochondrial genes relate to obesity and its related conditions need to be further addressed by the obesity research community.