Direct evidence of brown adipocytes in different fat depots in children

Abstract

Recent studies suggested the persistence of brown adipocytes in adult humans, as opposed to being exclusively present in infancy. In this study, we investigated the presence of brown-like adipocytes in adipose tissue (AT) samples of children and adolescents aged 0 to 18 years and evaluated the association with age, location, and obesity. For this, we analysed AT samples from 131 children and 23 adults by histological, immunohistochemical and expression analyses. We detected brown-like and UCP1 positive adipocytes in 10.3% of 87 lean children (aged 0.3 to 10.7 years) and in one overweight infant, whereas we did not find brown adipocytes in obese children or adults. In our samples, the brown-like adipocytes were interspersed within white AT of perirenal, visceral and also subcutaneous depots. Samples with brown-like adipocytes showed an increased expression of UCP1 (>200fold), PRDM16 (2.8fold), PGC1α and CIDEA while other brown/beige selective markers, such as PAT2, P2RX5, ZIC1, LHX8, TMEM26, HOXC9 and TBX1 were not significantly different between UCP1 positive and negative samples. We identified a positive correlation between UCP1 and PRDM16 within UCP1 positive samples, but not with any other brown/beige marker. In addition, we observed significantly increased PRDM16 and PAT2 expression in subcutaneous and visceral AT samples with high UCP1 expression in adults. Our data indicate that brown-like adipocytes are present well beyond infancy in subcutaneous depots of non-obese children. The presence was not restricted to typical perirenal locations, but they were also interspersed within WAT of visceral and subcutaneous depots.

Fig 1. Immunohistochemical evidence of brown-like adipocytes in adipose tissue samples.

A: Immunohistochemical staining of AT samples showed multilocular adipocytes, which stained positive for UCP1 (black arrows), surrounded by UCP1-negative white unilocular fat cells (white arrows) in subcutaneous (top), perirenal (middle) and visceral (bottom) AT samples of children. B: Two perirenal unilocular samples also exhibited positive UCP1 staining (grey arrows). C: Representative images of UCP1-negative subcutaneous (top, left) and perirenal (top, right) AT samples of lean children and subcutaneous samples of lean adults (bottom). Nuclei were counterstained with Mayer’s hematoxylin. Scale bars represent 50 μm in each panel.

Fig 2. Expression of brown, beige and white adipocyte markers in adipose tissue samples of children and adults.

A: UCP1 mRNA expression was significantly increased in UCP1^{histological+} samples compared to lean and obese children as well as adult UCP1^{histological-} samples with the exception of a single outlier from an UCP1^{histological-} subcutaneous sample (indicated as black box). B: PRDM16 was also increased in UCP1^{histological+} compared to all other UCP1^{histological-} samples. C, D, E and F: mRNA expression of PAT2, P2RX5, ZIC1 and LHX8 were not different in UCP1^{histological+} samples, although, one UCP1^{histological+} and one UCP1^{histological-} sample presented exceedingly high PAT2 expression values, marked with black squares. G,H and I: mRNA expression of the beige adipocyte markers TMEM26 and HOXC9 was not different in UCP1^{histological+} samples, although, two UCP1^{histological+} samples (from different donors) presented exceedingly high HOXC9 values, marked with black squares. These samples showed only islets of UCP1 positive adipocytes. TBX1 mRNA expression was slightly increased in UCP1^{histological-} samples of lean children and obese adults compared to UCP1^{histological+} samples. J and K: The WAT marker LEP was significantly decreased in UCP1^{histological+} samples compared to UCP1^{histological-} samples in contrast to ASC1, another white specific marker. L: ADIPOQ, a general marker for AT was not different between UCP1^{histological+} and UCP1^{histological-} samples. Target gene expression was normalized to the mean of the three housekeeping genes: ACTB, TBP and HPRT1. Statistical significance was assessed by one-way ANOVA and post-hoc Dunnett’s test and was compared to UCP1^{histological+} samples and marked with *P<0.05; **P< 0.01; ***P<0.001.

Fig 3. Expression analyses of brown and beige markers in adipose tissue samples of adults with low compared to high UCP1 mRNA expression.

We compared the upper 10% with the lowest 10% of UCP1 mRNA expression in paired subcutaneous and visceral AT samples of adults (n = 48) (A). Subcutaneous and visceral AT samples of adults with a high UCP1 expression showed significantly increased mRNA levels of PRDM16 (B) and PAT2 (C) in both subcutaneous and visceral depots. The classical brown specific marker ZIC1 was not altered between subcutaneous/visceral samples with a low or high UCP1 expression (D). Furthermore, samples with a high UCP1 expression appear to have slightly higher TMEM26 expression in subcutaneous and significantly increased mRNA levels in visceral AT samples (E). In contrast, TBX1 levels were significant increased in subcutaneous depots and slightly in visceral depots (F). Expression levels were normalized with the reference gene RPL27. Statistical significance was assessed by Student t-test and was marked with *P<0.05; **P< 0.01; ***P<0.001.

Fig 4. Correlation analysis in UCP1^{histological+} samples.

In statistical analyses which were restricted UCP1^{histological+} samples, we observed a negative association between UCP1 mRNA and age (A) and a positive association between UCP1 mRNA and PRDM16 (B). Pearson correlation coefficient R and p-value are shown in each scatterplot. Significant p-values (p<0.05) are indicated in bold. Black circles, subcutaneous depots; white circles, perirenal depots; and black squares, visceral depots.