Validation of Plasma Branched Chain Amino Acids as Biomarkers in Huntington Disease

While investigating body weight in a cohort of 32 patients at an early stage of Huntington disease (HD) and presymptomatic HD gene carriers, we found a significant decrease in the plasma branched-chain amino acids (BCAA) valine, leucine, and isoleucine in the HD group compared with 21 healthy controls. This systemic metabolic defect, which is indicative of hypercatabolism, was associated with early body weight loss in the HD group.¹ We wanted to (1) try to replicate our initial findings in a larger HD cohort, and (2) assess the feasibility of using plasma BCAA as a biomarker in HD, ie, in a less controlled research environment than the initial study.¹

After approval by the institutional ethics committees (Institut National de la Santé et de la Recherche Médicale, Recherche Biomédicale 03-48), we measured fasting levels of plasma BCAA as previously described¹ in 16 presymptomatic HD gene carriers and 70 patients with HD at a mild, moderate, or severe stage of the disease who were seen consecutively at our outpatient clinic. Our control group consisted of 21 healthy individuals, previously described.¹ To evaluate the multivariate associations, the LASSO (Least Absolute Shrinkage and Selection Operator) model selection technique was used with the adjusted r² statistic as the model selection criterion. The number of CAG codons was forced into these models with the Unified Huntington Disease Rating Scale (UHDRS), age, body mass index (BMI; calculated as weight in kilograms divided by height in meters squared), sex, and their 2-way interactions as possible covariates.

Plasma valine, leucine, and isoleucine levels were significantly lower in the HD group compared with controls (P = .02, <.001, and .002, respectively). In addition, we found a significant decrease of valine, leucine, and isoleucine in moderately affected patients and patients at a severe stage of the disease (Figure). Leucine was also significantly decreased in patients at a very mild stage of the disease and in presymptomatic individuals (Figure). Simple linear regression showed that plasma BCAA were not correlated with the UHDRS. However, we found that valine was negatively correlated with the number of CAG repeats (P = .01, adjusted r² = 0.086). After adjusting for BMI and sex, we confirmed that there was a negative association between valine and CAG repeats (P = .02, adjusted r² = 0.309). For every 1-unit increase in CAG, valine decreased by −3.33. We also found a significant association between leucine and CAG repeats (P = .04, adjusted r² = 0.372), with an association dependent on BMI due to the interaction. For every 1-unit increase in CAG, leucine decreased by −1.50 for the mean value of BMI (23.2).

We confirmed, in a larger cohort of patients with HD at different stages of the disease as well as presymptomatic HD gene carriers sampled consecutively, that plasma BCAA are relevant and accessible biomarkers in HD. After adjusting for BMI and sex, we showed an inverse correlation between the plasma levels of valine and leucine and number of CAG repeats, the primary determinant of HD severity. Of the 3 BCAA, leucine was of particular interest because it was significantly reduced in patients at an early stage of the disease and, more remarkably, in presymptomatic individuals. Leucine is a well-known activator of mTor (mammalian target of rapamycin),² which regulates protein synthesis and whose inhibition results in increased autophagic proteolysis.³ Akt, which has been shown to be altered in HD rat brain and in peripheral blood cells of patients with HD,⁴ also mediates the activity of mTor. Therefore, reduced leucine levels and altered Akt activation in HD may both result in mTor inhibition and exacerbation of proteolysis in HD.⁵ The significant correlation of BCAA with CAG repeats but not with UHDRS stages suggests that this biomarker is closely related to the primary defect in HD rather than to neuronal degeneration.

Correspondence: Dr Mochel, INSERM UMR S975, Department of Genetics, Hôpital de La Salpêtrière, 47 Boulevard de l'Hôpital, Bâtiment Nouvelle Pharmacie, 4ème étage, 75013 Paris, France (fanny.mochel@upmc.fr).

Financial Disclosure: None reported.

Additional Contributions: We wish to thank warmly the patients and presymptomatic individuals for their participation in this study. We are also thankful to Derek Blankenship, PhD, for statistical support.