Quantification of histone deacetylase isoforms in human frontal cortex, human retina, and mouse brain

Abstract

Histone deacetylase (HDAC) inhibition has promise as a therapy for Alzheimer's disease (AD) and other neurodegenerative diseases. Currently, therapeutic HDAC inhibitors target many HDAC isoforms, a particularly detrimental approach when HDAC isoforms are known to have different and specialized functions. We have developed a multiple reaction monitoring (MRM) mass spectrometry assay using stable isotope-labeled QconCATs as internal standards to quantify HDAC isoforms. We further determined a quantitative pattern of specific HDACs expressed in various human and mouse neural tissues. In human AD frontal cortex, HDAC1,2 decreased 32%, HDAC5 increased 47%, and HDAC6 increased 31% in comparison to age-matched controls. Human neural retina concentrations of HDAC1, 2, HDAC5, HDAC6, and HDAC7 decreased in age-related macular degeneration (AMD)-affected donors and exhibited a greater decrease in AD-affected donors in comparison to age-matched control neural retinas. Additionally, HDAC concentrations were measured in whole hemisphere of brain of 5XFAD mice, a model of β-amyloid deposition, to assess similarity to AD in human frontal cortex. HDAC profiles of human frontal cortex and mouse hemisphere had noticeable differences and relatively high concentrations of HDAC3 and HDAC4 in mice, which were undetectable in humans. Our method for quantification of HDAC isoforms is a practical and efficient technique to quantify isoforms in various tissues and diseases. Changes in HDAC concentrations reported herein contribute to the understanding of the pathology of neurodegeneration.

Fig 1. Purity and size of QconCATs.

(A) Purified QconCATs (30 pmol of QconCAT#1, 22 pmol of QconCAT#2, and 25 pmol of QconCAT#3) were separated on 10% SDS-PAGE and ImageJ software was used for estimation of purity. Purity was nearly 100% for QconCAT#1, ≥95% for QconCAT#2, and ≥97% for QconCAT#3. Protein concentrations were determined by DC Protein Assay. (B) Charge state masses were collected for all QconCATs (representative spectrum of QconCAT#1 shown) with an Agilent 6550 QTOF and deconvoluted with MagTran 1.0 (insert) to confirm protein mass was congruous with expected mass based on sequence. Observed mass of QconCAT#1 was 46126.7 Da (46125.6 Da theoretical), QconCAT#2 was 48931.2 Da (48929.7 Da theoretical), and QconCAT#3 was 43167.8 Da (43166.6 Da theoretical).

Fig 2. Calibration curves for quantification of peptide LEELGLAGR from HDAC6 standard using ¹⁵N-labeled peptide from QconCAT#2 as internal standard.

Area ratio of ¹⁴N-LEELGLAGR to ¹⁵N-LEELGLAGR for each of the three transitions for quantification was plotted versus pmol of HDAC6 standard. Individual transitions shown are (t1) 479.3/586.4 and 485.3/595.3, (t2) 479.3/715.4 and 485.3/725.4, (t3) 479.3/586.4 and 485.3/595.3. Three replicates for each were collected and presented as mean ± SD. Consensus for t1–t3 is represented in LEELGLAGR graph.