Diabetes and branched-chain amino acids: What is the link?

Branched-chain amino acids (BCAA) have increasingly been studied as playing a role in diabetes, with the PubMed search string “diabetes” AND “branched chain amino acids” showing particular growth in studies of the topic over the past decade (Fig. 1). In the Young Finn’s Study, BCAA and, to a lesser extent, the aromatic amino acids phenylalanine and tyrosine were associated with insulin resistance (IR) in men but not in women, whereas the gluconeogenic amino acids alanine, glutamine, or glycine, and several other amino acids (i.e. histidine, arginine, and tryptophan) did not show an association with IR.1 Obesity may track more strongly than metabolic syndrome and diabetes with elevated BCAA.2 In a study of 1302 people aged 40–79; higher levels of BCAA tracked with older age, male sex, and metabolic syndrome, as well as with obesity, cardiovascular risk, dyslipidemia, hypertension, and uric acid.3 Medium- and long-chain acylcarnitines, by-products of mitochondrial catabolism of BCAAs, as well as branched-chain keto acids and the BCAA themselves distinguished obese people having versus not having features of IR,4 and in a study of 898 patients with essential hypertension, the BCAA and tyrosine and phenylalanine were associated with metabolic syndrome and impaired fasting glucose.5 In a meta-analysis of three genome-wide association studies, elevations in BCAA and, to a lesser extent, in alanine tracked with IR, whereas higher levels of glutamine and glycine were associated with lesser likelihood of IR.6 Given these associations with IR, it is not surprising that a number of studies have shown higher BCAA levels in people with and prior to development of type 2 diabetes (T2D),7-12 although this has particularly been shown in Caucasian and Asian ethnic groups while not appearing to occur in African Americans.13 Similarly, higher BCAA levels track with cardiovascular disease.14, 15 The metabolism of BCAA involves two processes: (i) a reversible process catalysed by a branched-chain aminotransferase (BCAT), either cytosolic or mitochondrial, requiring pyridoxal to function as an amino group carrier, by which the BCAA with 2-ketoglutarate produce a branched-chain keto acid plus glutamate; and (ii) the irreversible mitochondrial process catalysed by branched-chain keto acid dehydrogenase (BCKDH) leading to formation of acetyl-coenzyme A (CoA), propionyl-CoA, and 2-methylbutyryl-CoA from leucine, valine, and isoleucine, respectively, which enter the tricarboxylic acid (Krebs) cycle as acetyl-CoA, propionyl-CoA, and 2-methylbutyryl-CoA, respectively, leading to ATP formation.16-18 The BCAA stimulate secretion of both insulin and glucagon and, when given orally, of both glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP), with oral administration leading to greater and more prolonged insulin and glucagon secretion.19 Insulin may particularly reduce BCAA turnover to a greater extent than that of other amino acids,20 and decreases the appearance and increases the uptake of amino acids.21 However, older studies of the effect of glucose or insulin on BCAA concentrations22 and rates of leucine appearance and oxidation showed no reduction in T2D,23 although the higher baseline levels of BCAA in obesity have long been recognized.22 Impaired function of BCAT and BCKDH has been posited, either as a primary genetic abnormality or due to effects of elevated fatty acids, proinflammatory cytokines, or insulin levels with consequent accumulation of branched-chain keto acids and metabolites such as diacylglycerol and ceramide,18 potentially contributing to the development of further insulin resistance,24 and decreased skeletal muscle BCAT and BCKDH expression has been shown in people with diabetes,25 supporting this concept. A Mendelian randomization study used measured variation in genes involved in BCAA metabolism to test the hypothesis of a causal effect of modifiable exposure on IR, showing that variants in protein phosphatase, Mg2+/Mn2+ dependent 1K (PPM1K), a gene encoding the mitochondrial phosphatase activating the BCKDH complex, are associated with T2D,26 but another such study suggested that genetic variations associated with IR are causally related to higher BCAA levels.27 Another hypothesis involves the mammalian target of rapamycin complex 1 (mTORC1), which is activated by BCAA, as well as by insulin and glucose via cellular ATP availability.28 If this is the relevant pathway, BCAA overload may cause insulin resistance by activation of mammalian target of rapamycin (mTOR), as well as by leading to increases in acylcarnitines,29 with mTOR seen in this scenario as a central signal of cross-talk between the BCAA and insulin.30 At this point, whether whole-body or tissue-specific BCAA metabolism is increased or decreased in states of insulin-resistant obesity and T2D is uncertain.31 Insulin action in the hypothalamus induces but overfeeding decreases hepatic BCKDH, leading to the concept that hypothalamic insulin resistance impairs BCAA metabolism in obesity and diabetes,32 so that plasma BCAAs may be markers of hypothalamic insulin action rather than direct mediators of changes in IR. A way to address this may be to understand the effects of changes in diet and other interventions on BCAA, as well as on IR and T2D. In an animal model, lowering dietary BCAA increased energy expenditure and improved insulin sensitivity.33 Two large human population studies showed an association of estimated dietary BCAA intake with T2D risk,34, 35 although another population study showed higher dietary BCAA to be associated with lower T2D risk.36 Ethnic differences, reflecting underlying differences in genetic variants, may be responsible for such differences. In the study of Asghari et al.37 in the current issue of the Journal of Diabetes, BCAA intake was associated with the development of subsequent IR. Studies of bariatric surgery suggest lower basal and post-insulin infusion BCAA levels are associated with greater insulin sensitivity,38 with reductions in BCAA not seen with weight loss per se with gastric band procedures, but occurring after Roux-en-Y gastric bypass,39 an intervention that may have metabolic benefits over and above those from reduction in body weight. The gut microbiota may be important for the supply of the BCAA to mammalian hosts, either by de novo biosynthesis or by modifying nutrient absorption.40, 41 A final fascinating preliminary set of observations is that of the effects of empagliflozin on metabolomics; evidence of increased Krebs cycle activation and of higher levels of BCAA metabolites, such as acylcarnitines,42 suggests that sodium–glucose cotransporter 2 (SGLT2) inhibition may, to some extent, involve BCAA metabolism. Certainly, we do not yet have a full understanding of these complex associations. However, the suggestion of multiple roles of BCAA in the development of IR promises to be important and to lead to the development of novel effective T2D therapies. 越来越多的研究调查了支链氨基酸(branched-chain amino acids,BCAA)对糖尿病的影响, 在PubMed上使用“糖尿病”与“支链氨基酸”这两个字符串搜索后发现在过去的十年中相关专题研究的数量增长特别多(图1)。在Young Finn的研究中发现BCAA以及在较小的范围中的芳香族氨基酸苯丙氨酸与酪氨酸都与男性胰岛素抵抗(insulin resistance,IR)相关, 但是与女性却不相关, 然而却没有发现糖异生氨基酸如丙氨酸、谷氨酸或甘氨酸以及其他几种氨基酸(如组氨酸、精氨酸与色氨酸)与IR之间具有相关性1。与代谢综合征和糖尿病相比, 肥胖与BCAA增多之间的相关性可能更强2。在一项纳入1302名年龄在40-79岁的受试者的研究中, 发现高水平的BCAA与老龄、男性性别、代谢综合征以及肥胖、心血管风险、血脂异常、高血压、尿酸之间都具有相关性3。具有IR特征的肥胖患者与不具有IR特征的肥胖患者相比, 他们之间的中链与长链酰基肉碱、BCAAs经过线粒体分解代谢后的副产物, 以及支链酮酸与BCAA本身都有差异4, 而在一项纳入898名原发性高血压患者的研究中, 也发现了BCAA、酪氨酸以及苯丙氨酸都与代谢综合征以及空腹血糖受损相关5。在一项纳入了3个全基因组关联研究的荟萃分析中, 发现BCAA水平升高以及在较小的范围中丙氨酸水平的升高都与IR相关, 然而高水平的谷氨酰胺以及甘氨酸却与更少的IR发生率相关6。考虑到这些与IR之间的相关性, 那么大量的研究都发现2型糖尿病以及糖尿病前期患者的BCAA水平都较高就不会令人奇怪了7-12, 虽然这种现象在高加索以及亚裔种群中表现得尤为明显, 但是在非洲裔美国人中却没有出现13。同样, 高水平的BCAA与心血管疾病之间也具有相关性14,15。 BCAA的代谢过程包括两个过程:(i)可逆的支链转氨酶(branched-chain aminotransferase,BCAT)的催化过程, 在胞浆中或者线粒体中, 需要以吡哆醛作为氨基载体,BCAA与2-酮戊二酸由此生成了支链酮酸与谷氨酸;以及(ii)在线粒体中不可逆的支链酮酸脱氢酶(branched-chain keto acid dehydrogenase,BCKDH)的催化过程, 亮氨酸、缬氨酸与异亮氨酸分别由此形成了乙酰辅酶A(CoA)、丙酰-CoA以及2-甲基丁酰-CoA, 它们分别形成乙酰CoA、丙酰-CoA与2-甲基丁酰-CoA后进入了三羧酸(Krebs)循环, 最终生成了ATP16–18。BCAA既可以刺激胰岛素分泌也可以刺激胰高血糖素分泌, 如果口服使用的话还可以刺激胰高血糖素样肽-1(GLP-1)以及葡萄糖依赖性促胰岛素释放肽(GIP)的分泌, 并且口服使用可使胰岛素以及胰高血糖素的分泌更为持久19。与对其他氨基酸的影响相比, 胰岛素可以更大程度地减少BCAA的转换20, 并且还可以减少氨基酸的出现以及增加它们的摄取21。然而, 既往在2型糖尿病患者调查葡萄糖或胰岛素对BCAA浓度影响的研究22以及调查亮氨酸的出现率与氧化率的研究发现它们并没有减少23, 虽然既往早已公认肥胖患者的基线BCAA水平更高22。BCAT与BCKDH功能受损, 无论是原发性遗传异常, 还是由于脂肪酸、促炎细胞因子或胰岛素水平升高以及随之而来的支链酮酸及其代谢产物如甘油二酯与神经酰胺蓄积后造成的影响18, 已确信会进一步导致胰岛素抵抗的进展24, 既往已经证实糖尿病患者骨骼肌中的BCAT与BCKDH表达减少了25, 这些证据都支持这个观念。有一项孟德尔随机化研究利用过去测定到的参与BCAA代谢的基因变异情况来检验这个假设, 亦即不同暴露情况对IR影响的因果效应, 结果发现蛋白磷酸酶出现了变异, 亦即Mg2+/Mn2+依赖性1K(PPM1K)基因变异, 它是负责编码线粒体磷酸酶的基因, 可以激活BCKDH复合物, 它们与2型糖尿病之间具有相关性26, 但是另外一项研究表明IR相关的基因变异与高BCAA水平之间具有因果关系27。另外一项假设涉及到了哺乳动物雷帕霉素复合体1靶点(mTORC1), 它可以被BCAA激活, 除此之外还可以被胰岛素以及葡萄糖通过细胞内可用的ATP所激活28。如果这就是相关的代谢途径,BCAA超负荷可通过激活哺乳动物雷帕霉素靶蛋白(mTOR)以及通过增加酰基肉毒碱导致胰岛素抵抗29, 在这种情况下可以将mTOR看作BCAA与胰岛素之间相互沟通的中央信号30。在存在胰岛素抵抗的肥胖以及2型糖尿病的状态下无论是全身性还是组织特异性的BCAA代谢, 此时的代谢是增加的还是减少的都是不确定的31。下丘脑中的胰岛素作用可以诱导肝BCKDH, 但过食会导至它的降低, 这引出了一个概念, 亦即在肥胖与糖尿病患者中下丘脑胰岛素抵抗可以损害BCAA代谢32, 因此血浆中的BCAAs可能是下丘脑胰岛素作用的标志物而不是反映IR变化的直接介质。 了解饮食与其他干预措施的变化对BCAA以及对IR与2型糖尿病的影响可能就是解决这个问题的方法。在一个动物模型中, 降低饲料中的BCAA含量并可以增加能量消耗, 最终可以提高胰岛素的敏感性33。两项大型人群研究结果都显示估计的饮食BCAA摄入量与2型糖尿病风险相关34,35, 虽然另外一项人群研究结果显示高饮食BCAA摄入量与低2型糖尿病风险相关36。种族差异, 反映了遗传变异的潜在差异, 可能是造成这种差异的原因。在本期Journal of Diabetes中,Asghari等7的研究结果显示BCAA摄入与随后发生的IR之间具有相关性。减肥手术相关研究结果表明, 更低的基础与餐时胰岛素输注后的BCAA水平与更高的胰岛素敏感性相关38, 并且没有看到胃束带手术本身带来的体重减轻会导致BCAA下降, 但是Roux-en-Y胃旁路术后却会出现BCAA下降39, 这种干预措施可能比体重减轻更具代谢益处。对于哺乳动物宿主来说肠道菌群可能是BCAA的重要供应来源, 无论是通过重新生物合成还是通过改变营养成分的吸收40,41。最后还有一项极其有趣的初步观察研究, 它调查了恩格列净对代谢组学的影响;结果证实Krebs循环的活化作用得到了增强并且BCAA代谢产物如酰基肉毒碱的水平更高42, 这意味着抑制钠-葡萄糖共转运体2可能在某种程度上也涉及到了BCAA代谢。 当然, 目前我们还不能够完全理解这些复杂的关系。然而, 重要的是要认识到在IR的发展过程中BCAA具有多种作用, 这将有助于我们研制出新的有效治疗2型糖尿病的药物。 图1 1967–2017年间每年发表的与糖尿病以及支链氨基酸相关的出版物数量。出版物通过使用“糖尿病”与“支链氨基酸”这两个搜索词在PubMed上面搜索获得。