Quantitative mass spectrometry reveals changes in SNAP-25 isoforms in schizophrenia

Abstract

SNAP-25 and syntaxin are presynaptic terminal SNARE proteins altered in amount and function in schizophrenia. In the ventral caudate, we observed 32% lower SNAP-25 and 26% lower syntaxin, but greater interaction between the two proteins using an in vitro assay. SNAP-25 has two isoforms, SNAP-25A and B, differing by only 9 amino acids, but with different effects on neurotransmission. A quantitative mass spectrometry assay was developed to measure total SNAP-25, and proportions of SNAP-25A and B. The assay had a good linear range (50- to 150-fold) and coefficient of variation (4.5%). We studied ventral caudate samples from patients with schizophrenia (n=15) previously reported to have lower total SNAP-25 than controls (n=13). We confirmed 27% lower total SNAP-25 in schizophrenia, and observed 31% lower SNAP-25A (P=0.002) with 20% lower SNAP-25B amounts (P=0.10). Lower SNAP-25A amount correlated with greater SNAP-25-syntaxin protein-protein interactions (r=-0.41, P=0.03); the level of SNAP-25B did not. Administration of haloperidol or clozapine to rats did not mimic the changes found in schizophrenia. The findings suggest that lower levels of SNAP-25 in schizophrenia may represent a greater effect of the illness on the SNAP-25A isoform. This in turn could contribute to the greater interaction between SNAP25 and syntaxin, and possibly disturb neurotransmission in the illness.

Keywords: Neural plasticity; Neurotransmission; Postmortem; SNARE proteins; Synapse.

Figure 1

SNAP-25 isoform assay sample preparation and analysis workflow. See text for description.

Figure 2

Efficiency of SNAP-25 immunoprecipitation (IP) from human brain samples. For NPC and SCZ samples, equal total protein amounts of crude brain homogenate were solubilized and clarified by centrifugation. IP efficiency of the SCZ sample was monitored by immunoblot. An aliquot of the input was kept to compare with the post-IP supernatant (SN) to confirm that SNAP-25 was specifically removed from the sample. Efficiency of bead washes was also monitored, the input, SN and wash #1 and #5 lanes can be compared directly. A control IP (beads that are not coupled with anti-SNAP-25 antibody, “empty beads”) did not immunoprecipitate SNAP-25 from a different brain sample. Empty beads correspond to magnetic beads not coupled to anti-SNAP-25 antibody, the same bead conditions used in the control IP. Antibody-coupled beads correspond to SP12 (anti-SNAP-25 antibody) coupled beads, used for IP. High molecular weight bands correspond to antibody eluting off the beads. Minimal heavy-chain and no antibody light-chain was detectable in eluted samples (comparing the SCZ or NPC IP lanes with antibody-coupled beads illustrates bands originating from antibody). Abbreviations: SCZ – schizophrenia; NPC- non-psychiatric comparison subject; SN – supernatant; IP- immunoprecipitation.

Figure 3

SNAP-25 peptides for MRM based isoform assay. A: SNAP-25 protein sequence schematic. Approximate locations of isoform-specific amino acid residues are indicated by vertical lines in the region encoded by the alternative exons (residues 54-96). Peptides chosen for MRM-monitoring in the isoform assay are indicated in color (red = A-specific; blue = B-specific; green = peptide common to both isoforms) and correspond to the MS/MS spectra in panels B-D. (B-D): Peptides and MS/MS spectra used for the isoform assay. MS/MS spectra obtained from recombinant SNAP-25 for the peptides monitored in the MRM isoform assay are shown. Peptide fragments (transitions) observed by MS/MS are labeled on the spectra and peptide fragmentation diagram. Transitions monitored in the isoform assay are circled. CPS = counts per second, m/z = mass-to-charge ratio. (B): SNAP-25A peptide; (C:) SNAP-25-B peptide; (D): Common SNAP-25 peptide. (E): Extracted ion chromatograms of MRM signals for one human brain sample. Traces correspond to SNAP-25A (red), SNAP-25B (blue) and total SNAP-25 (green) ions detected in a VMC sample. For MRM quantification, peak areas for each co-eluting transition were summed for each peptide (A, B, and Common). (E inset): Extracted ion chromatogram for a balanced mixture (50/50) of recombinant SNAP-25A and B. The traces correspond to the most intense transition peak for each peptide. The A and B peptides show different inherent ionization efficiency, as illustrated by the difference in heights and areas of the A and B-specific peaks, thus necessitating the use of an external calibration curve to determine the corresponding percentage of A and B in the sample. (F): SNAP-25 isoform ratio calibration curve. Recombinant SNAP-25A and B were mixed in known amounts (%SNAP-25A, x-axis) and analyzed by MRM. The summed peak areas of the A and B transitions were expressed as a ratio (A/B, y-axis). The general relationship between A/B and %SNAP-25A is exponential, but is approximated by a semi-log linear fit in the 5-90% SNAP-25A range (filled circles only, line of best fit has been weighted by 1/y). Error bars represent standard deviation of two or three replicates at different total protein amounts (0.5, 1, 2 μg, determined by Bradford assay). The A/B values for the human ventromedial caudate samples analyzed fell within the linear range (shaded grey). (G): Standard curves used for absolute quantification of SNAP-25 isoforms in the ventromedial caudate. Three separate amounts of the isotopically labeled heavy peptide (20, 80, 320 fmol) were run with each batch of ventromedial caudate samples (4-12 samples per batch) to create an external standard curve, within the known linear range of peptide detection. All ventromedial caudate samples quantified fell within this range. The transition peak area ratios for each heavy peptide sample were normalized to the peak area for the 80 fmol sample and curve fit (log-log fit, 1/x weighting). The common peptide/heavy peptide transition peak area ratio for each ventromedial caudate sample was compared to the curve to interpolate the total amount of SNAP-25 in each sample.

Figure 4

Linearity and detection limit of SNAP-25 peptides. (A): Detection range of the SNAP-25 common peptide. The heavy peptide internal standard (H104-119) was serially diluted in 1× MS buffer. Summed transition peak areas were normalized against one point in the curve (∼80 fmol dilution step) and plotted against the absolute quantity of the peptide. Points omitted from the curve fit are unfilled. Relative peak areas were log-linear over 2049-fold range (2.4 – 5000 fmol, filled circles). The dilution range encompassed by the external standard curves used to quantify human brain samples is shaded in grey. (B & C): Detection range of SNAP-25A and B was assessed using recombinant SNAP-25 and in-gel trypsin digestion. Isoforms were serially diluted and run on an SDS-PAGE gel, stained with Coomassie (bands are illustrated) and digested in-gel. Sums of the raw peak areas for isoform-specific transitions are plotted (left axis, black points) against the amount of recombinant SNAP-25 loaded onto the SDS-PAGE gel (x-axis). Extracted peptides were run in the presence of the heavy peptide (H104-119) internal standard to allow quantification of the amount of SNAP-25 (common peptide) extracted from the gel, (right axis, grey shading). Isoform specific peptide MRM signals were log-linear over a 50 to 150-fold range, and were detectable by MRM even at protein amounts below the detection limit of Coomassie staining. Note that the amount of the A or B specific peptide measured is directly proportional to the total SNAP-25A or B measured using the common peptide, suggesting that peptide recovery is not biased. Note: X and Y parameters in equations refer to log values, and log values are used in plots.

Figure 5

The MRM isoform assay detects changes in SNAP-25 isoform expression. (A): Absolute quantities of SNAP-25 isoforms in samples of human ventromedial caudate. The fmol amount of each isoform was calculated using the A/B ratio for each sample to estimate the %SNAP-25A using the external ratio calibration curve (Fig. 3F), then the absolute amount of SNAP-25A was determined as (%SNAP-25A/100) × fmol of total SNAP-25 (calculated using the heavy peptide external standard curve, Fig. 3G). Fmol of SNAP-25B was the difference between fmol total SNAP-25 and fmol SNAP-25A for each sample. *Total SNAP-25 and SNAP-25A were 27% and 31% lower in SCZ, respectively (P = 0.002; two-tailed t-test). Abbreviations: NPC – non-psychiatric comparison subject; SCZ – schizophrenia. (B): Absolute quantities of SNAP-25 isoforms in rats with subchronic exposure to antipsychotic medications. SNAP-25 isoforms were measured in medial striatum of rats treated with medications or saline control for 28 days (n = 10 per group). Drug treatment did not have a significant effect on SNAP-25A or B levels. Horizontal lines represent group median, crosses represent means, boxes represent the 25^th and 75^th percentiles and whiskers represent 10^th and 90^th percentiles.