Selected reaction monitoring as an effective method for reliable quantification of disease-associated proteins in maple syrup urine disease

Abstract

Selected reaction monitoring (SRM) mass spectrometry can quantitatively measure proteins by specific targeting of peptide sequences, and allows the determination of multiple proteins in one single analysis. Here, we show the feasibility of simultaneous measurements of multiple proteins in mitochondria-enriched samples from cultured fibroblasts from healthy individuals and patients with mutations in branched-chain α-ketoacid dehydrogenase (BCKDH) complex. BCKDH is a mitochondrial multienzyme complex and its defective activity causes maple syrup urine disease (MSUD), a rare but severe inherited metabolic disorder. Four different genes encode the catalytic subunits of BCKDH: E1α (BCKDHA), E1β (BCKDHB), E2 (DBT), and E3 (DLD). All four proteins were successfully quantified in healthy individuals. However, the E1α and E1β proteins were not detected in patients carrying mutations in one of those genes, whereas mRNA levels were almost unaltered, indicating instability of E1α and E1β monomers. Using SRM we elucidated the protein effects of mutations generating premature termination codons or misfolded proteins. SRM is a complement to transcript level measurements and a valuable tool to shed light on molecular mechanisms and on effects of pharmacological therapies at protein level. SRM is particularly effective for inherited disorders caused by multiple proteins such as defects in multienzyme complexes.

Keywords: BCKDH; MCAD (ACADM); MSUD; SRM; branched-chain amino acid catabolism; liquid chromatography; maple syrup urine disease; mass spectrometry; selected reaction monitoring; tandem mass spectrometry.

Figures

Figure 1

(A) Overall workflow of the process followed in the SRM assay. (B) Scheme of the proteins studied and the location of the selected peptides and mutations for each protein. The three letter code design the first three amino acids of each peptide, for E1α: VDG (VDGNDVFAVYNATK) and AVA (AVAENQPFLIEAMTYR); E1β: SGD (SGDLFNCGSLTIR) and LGV (LGVSCEVIDLR); E2: LSD (LSDIGEGIR) and LSE (LSEVVGSGK); E3: ALT (ALTGGIAHLFK) and EAN (EANLAASFGK); MCAD: EEI (EEIIPVAAEYDK), TGE (TGEYPVPLIR), AFT (AFTGFIVEADTPGIQIGR), and IYQ (IYQIYEGTSQIQR). Th BD: thiamine-binding domain; L BD: lipoyl-binding domain; NAD BD: NAD-binding domain.

Figure 2

Analysis of samples by SRM. Extracted chromatograms derived from the injection of peptide extracts are shown, where the blue color corresponds to the internal standard (heavy-labeled peptides) and the red color to the endogenous peptide. For quantification purposes the ratio endogenous peptide to internal standard was used. The y-axis corresponds to the intensity and the x-axis to the retention time. (A) Representative peptide analysis corresponding to the E1α (BCKDHA) subunit of the BCKDH complex. The peptide VDGNDVFAVYNATK was detected and quantified in the samples from healthy individuals (chromatogram on the left), but not in the samples from patients (chromatograms on the right). The detection level for the E1α subunit was determined as 23% respect to controls, so the E1α subunit levels in the samples from patients are below 23%. (B) Representative peptide analysis corresponding to the E1β (BCKDHB) subunit of the BCKDH complex. The peptide LGVSCEVIDLR was detected and quantified in the samples from healthy individuals (chromatogram on the left), but not in the samples from patients (chromatograms on the right). The detection level for the E1β subunit was determined as 19% respect to controls. Therefore, the E1β subunit levels in the samples from patients are below 19%. (C) Representative peptide analysis corresponding to the MCAD (ACADM) protein. The peptide IYQIYEGTSQIQR was detected and quantified in the samples from healthy individuals (chromatogram on the left) and in patients 1, 2, and 3, but not in the samples from patient 4. The detection level for the MCAD protein was determined as 9% respect to controls. Thus, the MCAD protein levels in the sample from patient 4 are below 9%. Two different healthy individuals were analyzed with three biological replicates for each of them. The detection levels for each protein were determined from the background noise in all control samples (n = 12).