The heavier (mean ± SE BMI 29.9 ± 1.0 kg/m2) and lighter (24.1 ± 0.9 kg/m2) co-twins of the discordant pairs had a mean difference in body weight of 16.9 ± 1.8 kg (p < 0.001) (Table 1). The heavier co-twins had more SAT, VAT and liver fat, larger adipocytes and were more insulin resistant (p < 0.05 all, Table 1). No significant differences in fat or metabolic measures were found between the concordant co-twins, despite a small difference in BMI (light 28.2 ± 1.9 vs heavy 30.4 ± 1.8 kg/m2, p = 0.043, Table 1).
Over 2000 genes were differentially expressed in adipocytes within the discordant pairs, revealing mitochondrial associations in pathway analyses
We first analysed the genome-wide differences in adipocytes of the lighter vs heavier discordant co-twins. A total of 2538 transcripts (figshare data S2 ) were differentially expressed within the twin pairs (nominal p value < 0.05). The top ten up- and downregulated genes are presented according to p value in Table 2 and fold change (FC) in Table 3, and figshare data S2 .
The top ten pathways of the significantly differentially expressed genes between the co-twins in adipocytes from IPA analyses (all p < 0.001) were subjected to calculation of the mean centroid values, representing the global transcriptional activity of the pathway. Based on the mean centroids, the significantly downregulated pathways in adipocytes of the heavier co-twins were OXPHOS, glutaryl-CoA degradation, mTOR signalling and branched-chain amino acid (BCAA) catabolism (valine and isoleucine degradation) and the significantly upregulated pathways were glucocorticoid receptor and IL-8 signalling (p < 0.05 all, Wilcoxon’s signed rank test, Table 4 and figshare data S3 ). Mitochondria-related pathways of OXPHOS and valine, isoleucine and glutaryl-CoA degradation correlated negatively with many adiposity and insulin resistance measures and hs-CRP, and positively with Matsuda index and adiponectin (Table 5). For IL-8 receptor signalling and glucocorticoid receptor signalling, the correlations were opposite to those seen for the above mentioned mitochondrial pathways.
Upstream regulator analyses and reduced levels of a major mitochondrial regulator PGC-1α in adipocytes
To detect transcripts regulating the observed genes and pathways, we performed an IPA upstream regulator analysis for the significantly differentially expressed genes in adipocytes. The top three significant regulators according to IPA were SREBF1, CEBPA and MYC (Table 6 and figshare S4 ). However, in the subsequent Wilcoxon’s test of these transcripts, none were significantly different between the co-twins.
When using the differentially expressed genes combined with the MitoCarta gene list (figshare data S5 ) to detect specifically mitochondria-related regulators, SREBF1, MYC and PGC-1α (Table 6 and figshare data S6 ) emerged as the top three regulators. PGC-1α was the only transcription factor significantly reduced in the heavier compared with the lighter co-twins (3.9 ± 0.3 vs 4.3 ± 0.3 Affymetrix units, respectively, p = 0.0157, Fig. 1a). This result was confirmed by qRT-PCR (heavier 2.6 ± 0.6 vs lighter 4.1 ± 1.0, p = 0.0414, Fig. 1b).
Consistently up- and downregulated transcripts in adipocytes revealed mitochondria-related transcriptional downregulation and immune-related upregulation
Next, we focused on the transcripts that were not only significantly different between the co-twins but also consistently up- or downregulated in the heavier co-twins, in at least 12 out of 14 discordant twin pairs. Adipocytes had 454 consistently upregulated and 538 consistently downregulated transcripts in the heavier co-twins (figshare data S7 ). The downregulated genes revealed mitochondria-related pathways in IPA (OXPHOS, fatty acid β-oxidation, AMP-activated protein kinase signalling, glutaryl-CoA degradation, tricarboxylic acid [TCA] cycle; p < 0.001 all), while the upregulated pathways displayed an immune-related pattern (IL-10 signalling, granulocyte adhesion, IL-8 signalling, recognition of bacteria and viruses, high mobility group box 1 (HMGB1) signalling and cytokine production; p < 0.001 all, figshare data S8 ).
Reduction of mtDNA transcript levels (qRT-PCR) and mitochondrial ribosomal protein subunit gene expression in the heavier co-twins’ adipocytes
The above results strongly suggest that genes targeting mitochondria are downregulated in the heavier co-twins’ adipocytes. Because Affymetrix only captures nuclear genes, we next analysed the expression of genes encoded by mtDNA by qRT-PCR. Here, the mitochondrial rRNAs (12S, MT-RNR1, p = 0.0022 and 16S, MT-RNR2, p = 0.0186) and mRNAs (MT-COX1, CIV subunit, p = 0.0047, MT-ND5, CI subunit, p = 0.0186, and MT-CYTB, CIII subunit, p = 0.0047) were significantly downregulated in the heavier as compared with the lighter co-twins (Fig. 2).
The mitochondrial rRNAs and mRNAs are translated into proteins by mitochondrial ribosomes. The mean centroids of both small (−0.24 ± 0.18 vs 0.23 ± 0.18 arbitrary Affymetrix units [AU], p = 0.0092) and large (−0.17 ± 0.18 vs 0.18 ± 0.17 AU, p = 0.0157) mitochondrial ribosomal protein subunit transcripts were downregulated in the heavier compared with the lighter co-twins’ adipocytes (Fig. 3 and figshare data S1 ).
Reduction of OXPHOS protein levels in adipocytes of obese vs lean individuals
The mitochondrial OXPHOS protein CI subunit NDUFB8 (p = 0.0176) and CIII-core 2 subunit (CIII, p = 0.0455) were significantly reduced in the obese compared with the lean individuals, whereas COX2 (CIV subunit, p = 0.253), ATP5A (CV subunit, p = 0.254) and the nuclear-encoded CII subunit SDHB (p = 0.361) were not (Fig. 4). When the OXPHOS protein signals were normalised against the mitochondrial protein porin, complex CI subunit NDUFB8 (p = 0.0106) was significantly reduced in the obese compared with the lean individuals, suggesting a possible decrease in OXPHOS CI levels per mitochondria. CIV levels decreased non-significantly (p = 0.088) and CII (p = 0.988), CIII (p = 0.199) and CV (p = 0.775) levels were unchanged (Fig. 4).
Adipocyte volume and number
Whether adipocyte size matters in reduced mitochondrial biogenesis or the metabolic complications in obesity is debatable, so we analysed the associations of adipocyte size and number with the above gene expression results. Adipocyte volume was significantly larger in the heavier than in the lighter co-twins, while the number of adipocytes did not differ between the co-twins (Table 1). Adipocyte volume correlated negatively with mitochondrial glutaryl-CoA degradation and mTOR signalling and positively with inflammatory IL-8 signalling for adipocytes (Table 5). However, in a multiple regression analysis adjusting for body fat mass (kg) and sex, adipocyte volume did not yield independent associations with the gene expression pathways (data not shown).
Adipose tissue transcriptomic analyses mimicked those of the adipocytes
To compare the gene expression results from adipocytes with those from adipose tissue, we performed the same transcriptional analyses of the 14 discordant co-twins in adipose tissue. A total of 2135 transcripts (figshare data S2 ) were differentially expressed between the co-twins (nominal p value < 0.05). As in adipocytes, the top ten upregulated genes in the heavier co-twins were associated with immune reactions and the top ten downregulated transcripts (figshare data S2 ) were associated with mitochondria, fatty acid synthesis and β-oxidation. Analyses according to the largest FC between the co-twins revealed chemokines and inflammatory genes among the top upregulated list, as in adipocytes. Downregulated transcripts included mostly the same genes as in adipocytes.
The pathway analyses in adipose tissue mimicked those of adipocytes, with OXPHOS, valine and lysine degradation, glutaryl-CoA degradation, acetate conversion to acetyl-CoA, fatty acid β-oxidation, triacylglycerol synthesis and ketone body production and breakdown among the top ten pathways; the mean centroids of these were all downregulated in the heavier co-twins (p < 0.05 all, Wilcoxon’s signed rank test, ESM Table 2 and figshare data S3 ). In general, significant negative correlations were found for the mean centroids of the top ten pathways and adiposity, including adipocyte size, leptin, insulin resistance and hs-CRP and positive correlations with Matsuda index and adiponectin (ESM Table 3). Triacylglycerol synthesis correlated only with VAT and liver fat.
The top three upstream regulators in adipose tissue were BACH1, CEBPA and ERG, the mean centroids of which did not differ between the co-twins. When combining the gene list with MitoCarta (figshare data S5 ) to analyse the mitochondrial upstream regulators in adipose tissue, SREBF1, PGC-1α and EIF2A (figshare data S6 ) emerged as the top three regulators. PGC-1α (3.8 ± 0.3 vs 4.9 ± 0.3 AU, p = 0.0019 ESM Fig. 1a) and EIF2A (11.2 ± 0.04 vs 11.4 ± 0.1 AU, p = 0.0029) were downregulated in the heavier co-twins.
In consistency analyses we found 221 upregulated and 437 downregulated transcripts (figshare data S7 ) in the heavier co-twins’ adipose tissue. The top significantly downregulated pathways included, for example, cell cycle control, DNA modifications, mTOR signalling and cholesterol biosynthesis, and upregulated pathways mostly immune- and lipid-related signalling (p < 0.001 all, IPA, figshare data S8 ).
In our previous study with 26 BMI-discordant pairs , we reported reduced expression of mitochondrial ribosomal protein subunit transcripts in the heavier co-twins’ adipose tissue. In the current study, we repeated these ribosomal protein subunit expression level analyses in adipose tissue in the 14 discordant pairs. Both MRPS (heavier twins −0.24 ± 0.19 vs lighter twins 0.11 ± 0.14 AU, p < 0.01) and MRPL (heavier twins −0.21 ± 0.2 vs lighter twins 0.01 ± 0.15 AU, p < 0.05, figshare data S1 ) were downregulated in the heavier co-twins (ESM Fig. 1b), as in adipocytes.