Genetic Markers of Brown Adipose Tissue Identity and In Vitro Brown Adipose Tissue Activity in Humans

Study approval

The study was reviewed and approved by the ethics committee of Maastricht University Medical Center (METC 10-3-012, NL31367.068.10). Informed consent was obtained before surgery, and participants did not receive a stipend.

Human biopsies and primary adipocyte cultures

Paired human BAT and subcutaneous WAT biopsies were collected during deep neck surgery in patients with normal thyroid function. Stromal vascular fractions were obtained from adipose tissue biopsies by using collagenase digestion. Isolated cells were grown to confluence before differentiation was initiated as previously described (16).

Mitochondrial respiration experiments

The stromal vascular fractions derived from adipose tissue biopsies were plated in XF96 cell culture microplates (Agilent Technologies, Santa Clara, California). After cells were fully differentiated, oxygen consumption and mitochondrial function were measured by using the Seahorse XF96 extracellular flux analyzer from Agilent Technologies. Cells were incubated for 1 hour at 37°C in unbuffered XF assay medium supplemented with GlutaMAX (2 mM) (Thermo Fisher Scientific, Waltham, Massachusetts), sodium pyruvate (1 mM), and glucose (25 mM). The basal oxygen consumption rate was measured, followed by an injection of 1 µM oligomycin and then an injection of 1 µM norepinephrine (NE). Oligomycin, propranolol, and NE were purchased from Sigma-Aldrich (St. Louis, Missouri). Mitochondrial uncoupling was examined as mitochondrial respiration after the inhibition of ATP synthase with oligomycin (which was set to 100%), thus reflecting mitochondrial uncoupling because of proton leak. Data were plotted as a percentage calculated from the fourth measurement after the injection with oligomycin.

RNA isolation and quantitative polymerase chain reaction analysis

RNA was isolated from human adipose tissue biopsied material and cultured adipocytes by using TRIzol extraction (Thermo Fisher Scientific) followed by purification by using the RNeasy kit from Qiagen (Hilden, Germany) according to the manufacturer's instructions. Complementary DNA was synthesized by using the High-Capacity RNA-to-cDNA Kit from Applied Biosystems (Foster City, California) according to the manufacture's instructions. Quantitative polymerase chain reactions were performed on a CFX384 Touch Real-Time PCR Detection System from BioRad Laboratories (Hercules, California). Gene expression was calculated by using the 2^-ΔΔCt method. Relative gene expression was normalized to TATA box-binding protein. The following SYBR green primers were used to analyze gene expression: adiponectin (ADIPQ) (fwd: accaggaaaccacgactcaa, rev: accaataagacctggatctcctttc), leptin (fwd: gctgtgcccatccaaaaagtcc, rev cccaggaatgaagtccaaaccg), and zinc finger protein of cerebellum 1 (ZIC1) (fwd: acatgaaggtccacgaatcct, rev: cttgtggtcgggttgtctgt). Primers detecting mitochondrial tumor suppressor 1 (MTUS1) (Hs00368183_m1) and potassium channel subfamily k member 3 (KCNK3) (Hs00605529_m1) were from Applied Biosystems. Primers detecting adipocyte protein 2 (AP2) (13), UCP1 (17), epithelial v-like antigen (EVA1) (12), and TATA box-binding protein (12) have been described previously.

Statistics

For respiration experiments, differences were analyzed by using a two-way analysis of variance (ANOVA) in GraphPad Prism (GraphPad Software Inc., San Diego, California), and the correlation analyses (Spearman) were also performed in GraphPad Prism. P < 0.05 was considered significantly different.

NE-stimulated mitochondrial uncoupling is specific for differentiated adipocytes derived from human BAT and is dependent on β-adrenergic receptor

Paired BAT and WAT biopsies were collected from 16 individuals with a wide range in BMI and age undergoing deep neck surgery (Table 1) and were used for culturing preadipocytes from the stromal vascular fraction. In vitro BAT activity was measured as mitochondrial uncoupling followed by stimulation with natural β-adrenergic receptor agonist NE. In accordance with the characteristic capacity of BAT, acute stimulation with NE in differentiated adipocytes derived from human BAT significantly increased mitochondrial uncoupling to 256 ± 3.3%, an effect that was largely absent in differentiated adipocytes derived from human subcutaneous WAT coming from the same region (125 ± 9.0%; Figure 1A), which is in agreement with our previously published results (16). To demonstrate that this mitochondrial uncoupling is directly due to β-adrenergic stimulation, we used β-adrenergic receptor antagonist propranolol. Indeed, NE-stimulated mitochondrial uncoupling was completely abolished by propranolol in adipocytes derived from human BAT (Figure 1A).

NE-stimulated mitochondrial uncoupling in differentiated adipocytes derived from human BAT is related to age and BMI

Differentiated adipocytes derived from human BAT in the current study originated from volunteers with a wide range in BMI and age. Although age and BMI are related to each other, this allowed us to investigate whether the intrinsic NE-stimulated mitochondrial uncoupling activity of adipocytes derived from human BAT or WAT was related to the BMI or age of the donor. NE-stimulated mitochondrial uncoupling in differentiated adipocytes derived from human BAT displayed a negative correlation to age (ρ = −0.69; P = 0.004) and BMI (ρ = −0.58; P = 0.019), while no such correlation was established in differentiated adipocytes derived from human WAT (Figure 1B-1E). These findings are consistent with previous reports demonstrating that 2-deoxy-2-(¹⁸F)fluoroglucose glucose uptake in the human neck region is negatively associated with age and BMI (3, 5, 8).

ZIC1 messenger RNA negatively correlates with in vitro BAT activity in differentiated adipocytes derived from human BAT

The exact nature of human BAT is currently under debate. Thus far, studies devoted to genetic markers of beige and brown adipocytes in human BAT biopsies have included ZIC1 (13, 18), EVA1 (12, 18), KCNK3 (19), and MTUS1 (19) with the aim to distinguish human brown and beige from white adipocytes. Next, we investigated how messenger RNA (mRNA) expression of these genes relates to in vitro BAT activity measured in in vitro differentiated adipocytes. UCP1 mRNA was positively associated with intrinsic mitochondrial uncoupling activity in differentiated adipocytes derived from human BAT (ρ = 0.77; P = 0.002) (Figure 2A). Interestingly, ZIC1 mRNA showed a negative correlation to in vitro BAT activity (ρ = −0.59; P = 0.026) (Figure 2B). mRNA expression of EVA1, MTUS1, and KCNK3 did not correlate with in vitro BAT activity (Figure 2C-2E). In order to monitor the effects of differentiation in these experiments, we examined ADIPQ mRNA expression and correlated that to in vitro BAT activity. ADIPQ mRNA did not show a correlation with in vitro BAT activity (Figure 2F), indicating that in vitro BAT activity was not dependent on the differentiation state of the adipocytes because ADIPQ is an important indicator of adipocyte differentiation (20, 21).

No correlations to ADIPQ or AP2 were detected for UCP1 mRNA (Figure 3A-3C). Also, no correlations to ADIPQ or AP2 were detected for ZIC1 mRNA expression (Figure 3B-3D). These results further indicate that differentiation state does not seem to be a driver for UCP1 or ZIC1 expression in adipocytes derived from human BAT.

Expression of genetic markers in human BAT biopsies from the neck

Next, the previously discussed markers were analyzed in paired BAT and WAT biopsies of adult human subjects undergoing deep neck surgery. As expected, UCP1 expression was higher in BAT biopsies compared with WAT biopsies (Figure 4A), and leptin expression was higher in WAT biopsies compared with BAT biopsies (Figure 4A). In addition, expression of MTUS1, EVA1, and KCNK3 was higher in human BAT biopsies compared with WAT biopsies (Figure 4A). Although ZIC1 mRNA was increased in BAT biopsies compared with WAT biopsies, this increase did not reach statistical significance (Figure 4A).

Correlation analysis of genetic markers in human BAT biopsies from the neck to age and BMI

Because our BAT biopsies were derived from volunteers with a wide range of age and BMI, we next determined if these markers were associated with increased age and BMI. Interestingly, UCP1 and KCNK3 displayed a negative correlation to age (ρ = −0.67; P = 0.006 and ρ = −0.56; P = 0.031), while only UCP1 displayed a negative correlation to BMI (ρ = −0.53; P = 0.044) (Figure 4B-4C). In contrast, ZIC1 and EVA1 showed a positive correlation to age (ρ = 0.70; P = 0.004 and ρ = 0.70; P = 0.004) and BMI (ρ = 0.77; P = 0.001 and ρ = 0.64; P = 0.010) (Figure 4B-4C). No correlations to age or BMI were detected for MTUS1 (Figure 4B-4C).