The primary focus of this study was to determine whether or not obese youth with or without type 2 diabetes show the fasting plasma metabolic signatures of elevated amino acid and medium- to short-chain AcylCN species reported in adults. Our findings demonstrate that the general trend, in both the OB and the diabetic youth, is for lower plasma short- and medium-chain AcylCN with unchanged long-chain AcylCN, in addition to lower plasma concentrations of most amino acids. Thus, both the OB and the diabetic youth showed no evidence of defects in fatty acid or amino acid metabolism compared with their NW peers, despite being insulin resistant.
Obese adults with or without type 2 diabetes are characterized by dysregulated fatty acid and amino acid metabolism. Recent investigations have applied comprehensive metabolomic profiling to gain a broader understanding of the metabolic differences between lean, obese, and diabetic adults. We and others recently demonstrated elevations in AcylCN and amino acid concentrations in obese compared with lean (3
), in adults with diabetes compared with nondiabetic individuals (6
), and in lean insulin-resistant compared with lean insulin-sensitive adults (5
). On the basis of these observations of altered fatty acid and amino acid metabolism in adults, we hypothesized that adolescents with type 2 diabetes would have higher concentrations of specific AcylCN species and of specific plasma amino acids when compared with their OB and NW peers. In contrast to adults, our results show that adolescents with type 2 diabetes have lower concentrations of specific fatty acid– and amino acid–derived AcylCN species along with lower plasma amino acid concentrations. The lower concentrations of fatty acid–derived AcylCN species observed in our diabetic adolescents are unlikely to be attributable to greater fatty acid reesterification because fasting FFAs and the percentage of FFAs reesterified were comparable among all three groups (data not shown). Also, the concentrations of long-chain AcylCN species (C18:2-CN to C14:0-CN) were comparable among groups, suggestive of similar inputs into β-oxidation. Additionally, the ratio between freeCN and C16-CN, which is increased when carnitine palmitoyltransferase-1 function is blocked, did not differ among groups (NW, 761.0 ± 48.9; OB, 820.0 ± 64.2; and type 2 diabetes, 735.4 ± 91.5; P
= 0.69). Yet, the presence of lower concentrations of the later β-oxidation intermediates (C10:1-CN to C2-CN) in the diabetic subjects is consistent with enhanced rather than reduced utilization of these intermediates. These observations from metabolomics are consistent with the in vivo indirect calorimetry findings of increased FOX and Vo2
in youth with type 2 diabetes compared with NW, whether expressed in absolute terms or controlling for fat-free mass (data not shown). Additionally, the lower amino acid concentration with the concomitant decreases in amino acid–associated AcylCN species (C3-CN, C4-CN, and C5-CN) derived from the catabolism of BCAA also suggests enhanced utilization through β-oxidation. On the other hand, the lower plasma BCAA may be due to increased gluconeogenic drive typically seen in diabetes, with the BCAA being channeled to fuel gluconeogenesis through conversion to glutamate, pyruvate, and alanine. Additionally, increased fat mass and higher rates of lipolysis in obesity and diabetes contribute increased glycerol which fuels gluconeogenesis (24
) and may contribute to the lower BCAA concentrations being utilized as gluconeogenic substrates. The observed lower alanine concentrations in type 2 diabetic youth is in favor of increased gluconeogenic drive. Lastly, a reduction in amino acids might also reflect decreased catabolism or increased anabolism in growing adolescents. We had shown that proteolysis and protein oxidation are lower during puberty compared with prepuberty in normal weight youth (25
). Even though data are lacking in obese youth, it is possible that obesity-associated hyperinsulinemia may further augment this together with their higher dietary intakes. On the other hand, there are data suggesting that BCAA uptake by adipose tissue is reduced in an adult transgenic moderately obese mouse model (26
). It remains to be determined if such is the case in human obesity and whether or not there are differences in adipose tissue BCAA uptake between growing adolescents versus adults. This may be yet another area where adolescents and adults may differ.
Although further studies are required to clarify the underlying mechanism(s) for these metabolic differences in youth, mitochondrial adaptation and metabolic plasticity early in the course of obesity in youth may be a likely explanation. Similar adaptive mitochondrial responses were identified in a serial study of Zucker diabetic fatty rats where the animals initially upregulated their oxidative capacity in the face of progressive insulin resistance (27
). Yet, the upregulated oxidative capacity was unable to stop the diabetic progression, and with age, the mitochondrial adaptive response was lost. Consistent with our data, a recent investigation in young adults (mean age 22 years) demonstrated comparable long-chain AcylCN species but lower medium-chain AcylCN species in obese compared with lean individuals (28
). These results support the hypothesis that adolescents and young adults with obesity and type 2 diabetes have not yet developed the mitochondrial defects that are documented in older adults (29
), and they may even have enhanced mitochondrial activity as an adaptation. In support of the latter are previous reports of increased FOX in obese children, proposed to be a metabolic defense against further weight gain (31
Thus, with the present findings of enhanced rates of β-oxidation in OB and diabetic youth, we propose that their mitochondrial function is not impaired. However, over time and with continued progressive obesity from childhood to adulthood, chronic exposure to excessive β-oxidation results in mitochondrial overload (33
) and oxidative stress, culminating in a reduced overall oxidative capacity, similar to the changes found in diet-induced insulin-resistant mice (34
). Our diabetic youth compared with NW were unable to suppress their FOX during hyperinsulinemic conditions (30 vs. 70%), resulting in continued exposure and demand for excessive β-oxidation. Longitudinal studies of metabolomics from youth to adulthood are needed to test our hypothesis. Moreover, it remains to be determined if the progression to abnormal mitochondrial function could be prevented or aborted. We further propose, based on the inverse relationship between IS and FOX (r
= −0.502, P
< 0.001), that the increased rates of FOX in youth with obesity and diabetes may play a role in their insulin resistance through the Randle cycle of the competition between fatty acid and glucose oxidation (35
). We previously demonstrated that the Randle cycle is operative in pubertal adolescents, potentially explaining the physiology of pubertal insulin resistance (14
Lastly, a cluster of obesity-related amino acid and AcylCN metabolites have demonstrated an inverse relationship with IS among adults. Specifically, Newgard et al. (3
) observed associations between BCAA and insulin resistance. Likewise, Tai et al. (5
) reported that insulin resistance was associated with leucine/isoleucine, phenylalanine, tyrosine, and alanine, and a cluster of BCAA and related amino acids identified by principal components analysis. Thus, our second aim was to assess the relationship between IS or FOX with AcylCN species or amino acid concentrations. We found that once we accounted for the level of adiposity (BMI, %BF, or VAT), Tanner stage, and sex, the relationship between AcylCN species and IS or FOX disappeared. Furthermore, the observed correlations between BMI and IS and FOX (), suggest that the AcylCN associations are mediated by obesity in our pediatric population. In contrast, the relationship between IS and arginine, histidine, serine, and glycine concentrations remained significant after controlling for adiposity, Tanner stage, and sex.
The strengths of the current study are 1) the number of otherwise healthy NW and OB adolescents and youth with type 2 diabetes; 2) the extensive investigations that included the state-of-the-art use of stable isotopes together with indirect calorimetry and the hyperinsulinemic-euglycemic clamp to assess in vivo lipolysis, lipid oxidation, and IS; 3) the comprehensive evaluation of whole body and abdominal adiposity; and 4) a first-time metabolomics analysis in youth. A potential weakness is the absence of in vivo protein turnover examination. Another perceived weakness is that youth with type 2 diabetes were on treatment. However, ethically, patients cannot be evaluated without the provision of proper therapy. Moreover, the differing treatment modalities were driven by standard clinical recommendations and not by a research protocol.
In summary, our results show that in contrast to adults, OB youth with or without type 2 diabetes do not demonstrate impaired fatty acid or amino acid metabolism. In fact, the metabolomics and the in vivo data favor enhanced mitochondrial function, as obese adolescents with type 2 diabetes demonstrate lower concentrations of the later β-oxidation intermediates along with higher rates of FOX. It is our theory that these contrasting results between adolescents and adults is a reflection of the duration of obesity and the consequent and gradual evolution of failure of mitochondrial adaptive mechanisms as the obese individual transitions from youth to adulthood and forward with continued obesity. Future longitudinal studies of metabolomics will be required to test this theory and to determine if this early adaptive metabolic plasticity disintegrates over time and if intervention(s) could prevent such maladaptive progression into adulthood.