pubmed.ncbi.nlm.nih.gov

Quantitative analysis of the detergent-insoluble brain proteome in frontotemporal lobar degeneration using SILAC internal standards - PubMed

  • ️Sun Jan 01 2012

Quantitative analysis of the detergent-insoluble brain proteome in frontotemporal lobar degeneration using SILAC internal standards

Nicholas T Seyfried et al. J Proteome Res. 2012.

Abstract

A hallmark of neurodegeneration is the aggregation of disease related proteins that are resistant to detergent extraction. In the major pathological subtype of frontotemporal lobar degeneration (FTLD), modified TAR-DNA binding protein 43 (TDP-43), including phosphorylated, ubiquitinated, and proteolytically cleaved forms, is enriched in detergent-insoluble fractions from post-mortem brain tissue. Additional proteins that accumulate in the detergent-insoluble FTLD brain proteome remain largely unknown. In this study, we used proteins from stable isotope-labeled (SILAC) human embryonic kidney 293 cells (HEK293) as internal standards for peptide quantitation across control and FTLD insoluble brain proteomes. Proteins were identified and quantified by liquid-chromatography coupled with tandem mass spectrometry (LC-MS/MS) and 21 proteins were determined to be enriched in FTLD using SILAC internal standards. In parallel, label-free quantification of only the unlabeled brain derived peptides by spectral counts (SC) and G-test analysis identified additional brain-specific proteins significantly enriched in disease. Several proteins determined to be enriched in FTLD using SILAC internal standards were not considered significant by G-test due to their low total number of SC. However, immunoblotting of FTLD and control samples confirmed enrichment of these proteins, highlighting the utility of SILAC internal standard to quantify low-abundance proteins in brain. Of these, the RNA binding protein PTB-associated splicing factor (PSF) was further characterized because of structural and functional similarities to TDP-43. Full-length PSF and shorter molecular weight fragments, likely resulting from proteolytic cleavage, were enriched in FTLD cases. Immunohistochemical analysis of PSF revealed predominately nuclear localization in control and FTLD brain tissue and was not associated with phosphorylated pathologic TDP-43 neuronal inclusions. However, in a subset of FTLD cases, PSF was aberrantly localized to the cytoplasm of oligodendrocytes. These data raise the possibility that PSF directed RNA processes in oligodendrocytes are altered in neurodegenerative disease.

PubMed Disclaimer

Figures

Figure 1
Figure 1. Proteomic Analysis of TDP-43 positive FTLD cases

(A) Immunohistochemical analysis of a control frontal cortex using a panTDP-43 antibody (upper panel) shows almost exclusively nuclear staining, whereas a phosphorylated TDP-43 specific (pTDP-43, Ser409/410) antibody detects cytoplasmic inclusions in FTLD frontal cortex (lower panel). Control cases do not stain positive for pTDP-43 (middle panel). Hemotoxylin nuclear counterstain shown in blue. (B) Workflow for the SILAC quantitative proteomic approach in human brain tissue. IHC, immunohistochemistry.

Figure 2
Figure 2. Use of SILAC internal standards to quantify the detergent insoluble brain proteome

(A) SDS-PAGE analysis of Brain/SILAC mixtures. Molecular weight marker (M; lane 1), whole cell lysate of SILAC labeled HEK293 cells (lane 2), FTLD brain insoluble-proteome alone (lane 3), FTLD/SILAC mixture (lane 4), control brain insoluble-proteome alone (lane 6), Control/SILAC mixture (lane 7). 20 μg of protein loaded in each lane. The brain/SILAC samples (lanes 4 and 7) were, excised into gel slices based on MW (regions A-E), digested with trypsin, and analyzed by liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) on a hybrid linear ion-trap/Orbitrap mass spectrometer. (B) The SILAC labeled peptides from the HEK293 cells are chemically identical to their native/unlabeled counterparts and serve as internal standards for the measurement of protein abundance across control and FTLD samples. (C) Representative base peak elution profiles (MW region A) of control:HEK293 and FTLD:HEK293 peptide mixtures, respectively.

Figure 3
Figure 3. Histogram analysis of the brain/SILAC mixed proteome

(A) Log2 ratios (light/heavy) from a representative control brain/SILAC mixture (replicate 2) shows a bimodal distribution of proteins that can be fitted to two populations (HEK293 and brain). Upper panel is the residual associated with the fitted bimodal distribution using Igor software. (B-D) Full MS scans (top) and extracted ion-chromatograms (bottom) using a ±20 ppm isolation window of brain and SILAC labeled HEK293 specific peptide ion pairs. In B, a SET specific peptide ion (m/z 608.818 corresponding to amino acid sequence VEVTEFEDIK) was observed exclusively in HEK293 cells (light peptide expected m/z 604.809 was at or below noise). In C, light/heavy peptide ion pair for UBA1 (m/z 955.009 (light) and 955.059 (heavy), respectively corresponding to amino acid sequence SLVASLAEPDFVVTDFAK) were almost equally intense in both brain and SILAC labeled HEK293 cells. In D, a MAP2 specific peptide ion (m/z 766.903 (light) corresponding to amino acid sequence SDTLQITDLGVSGAR) was observed exclusively in brain (heavy peptide expected m/z 771.910 was at or below noise). Red and blue circles represent heavy (SILAC labeled HEK293) and light peptides (brain), respectively. Triangles represent expected heavy (red) or light (blue) peptide signals.

Figure 4
Figure 4. The SILAC labeled internal standards increase precision of protein quantitation within a discrete range

(A) scatter plot of raw (brain/SILAC) protein ratios (log2 transformed) measured for one or more peptides in both technical replicate LC-MS/MS runs. This data includes protein ratios derived from both control/SILAC and FTLD/SILAC peptide mixtures. (B) The mean of the replicate measurements represented by x, y pairs in A were plotted (x-axis) versus the standard error of the pair (y-axis). Median standard error for bins of width 0.4 beginning at bin midpoint −3.8 and ending at midpoint 5.8 are overlaid as transparent bars.

Figure 5
Figure 5. Relative changes in the FTLD detergent-insoluble proteome

(A) MS spectra of detected peptide ion pairs (light and heavy) from control (upper panel) and FTLD (lower panel) detergent-insoluble samples. Presented are RHOGDI-alpha, triosephosphate isomerase 1 (TPI1), annexin I (ANXA1) and SAM68 (a negative control). White and black circles represent light (brain) and heavy (SILAC labeled HEK293 cells) peptide ion pairs, respectively. (B) Histogram analysis of the log2 difference (FTLD-control) for shared proteins in replicate two. (C) Immunoblots (IB) for protein targets (see Table 1) in individual control and FTLD cases are shown. Ponceau S staining of total protein is provided to show equal loading in all lanes.

Figure 6
Figure 6. Biochemical and immunohistochemical characterization of PSF in control and FTLD brain

(A) Detergent-insoluble fractions of frontal cortex samples from individual FTLD and control cases immunoblotted for PSF. Native (~100 kDa) and truncated PSF species (~25 kDa) are shown (arrows). Ponceau S staining of membrane is provided to show equal protein loading. (B) Quantification of native PSF (~100 kDa) by densitometry shows significant enrichment in FTLD cases (p=0.035). The line represent the mean value in control and FLTD. The four FTLD cases at or above the mean are labeled. (C) Immunohistochemical analysis for PSF in frontal cortex from a representative control and FTLD case (case 1 and 10, respectively). PSF expression was mostly nuclear in control and FTLD. PSF was observed accumulated in the cytoplasm in frontal cortex of a subset of FTLD cases (5, 7 and 10), and in these cases the stained cell morphology was consistent with glia. Lower panels represent zoomed magnification of boxed regions in upper panels. Scale bars = 20 μM. (D) In neurons of the frontal cortex or hippocampus, PSF (red) was exclusively localized to the nucleus and did not colocalize with cytoplasmic pTDP-43 (green). Upper (frontal cortex) and lower (hippocampus dentate gyrus) panel images were captured on an Olympus BX51 florescent microscope (Scale bar = 20 μM) and Zeiss LSM 510 confocal microscope (scale bar represents 10 μM), respectively.

Figure 7
Figure 7. PSF is expressed in neurons, oligodendrocytes and astrocytes and shows developmentally programmed decline during myelin development

(A) Immunoblots for PSF, PTB and TDP-43 in primary cultured mouse neurons (Neu), astrocytes (Ast), oligodendrocytes (OL) and microglia (MG). Ubiquitin carboxyl-terminal hydrolase isozyme L1 (UCHL1), GFAP and Olig2 were used as neuron, astrocyte and oligodendrocyte specific markers, respectively. (B) PSF, PTB and TDP-43 are expressed in human (HOG) and mouse (CG4) oligodendrocyte cell lines. (C) Optic nerve samples from mice at post-natal day 10, 23 and 38 were immunoblotted for PSF, PTB and TDP-43 in biological replicate. 2′,3′-cyclic nucleotide 3′ phosphodiesterase (CNP) is an oligodendrocyte specific maker. Actin was used as a loading control in all immunoblots (A-C).

Figure 8
Figure 8. PSF is predominately localized to the nucleus of mature astrocytes and only rarely observed in the cytoplasm

Double-labeling immunofluorescence in free-floating sections in FTLD (case 7) frontal cortex for PSF (red) and GFAP (green) show PSF nuclear colocalization in the GFAP expressing astrocytes. (white arrows, left panel). Although the majority of cytoplasmic PSF does not co-localize with GFAP (center panel), a small percentage (~11 %) of GFAP positive astroctyes cells captured in our analysis did display some level of cytoplasmic PSF (right panel). All scale bars represent 50 μM

Figure 9
Figure 9. PSF and Olig2 show a similar cytoplasmic distribution in oligodendrocytes in FTLD

(A) Double-labeling immunofluorescence staining in paraffin embedded sections (frontal cortex) for PSF (red) and Olig2 (green) in a control case 2 shows PSF and Olig2 nuclear localization in oligodendrocytes (top panel). In a subpopulation of oligodendrocytes in FTLD (case 10), PSF and Olig2 both display cytoplasmic colocalization (middle and bottom panel). Scale bar represents 25 μM. (B) Immunohistochemisty in FTLD (case 10) on adjacent sections probed with either PSF or Olig2 antibodies showed cells with a comparable pattern of cytoplasmic distribution (scale bar represents 20 μM).

Similar articles

Cited by

References

    1. Taylor JP, Hardy J, Fischbeck KH. Toxic proteins in neurodegenerative disease. Science. 2002;296:1991–5. - PubMed
    1. Ross CA, Poirier MA. Protein aggregation and neurodegenerative disease. Nat Med. 2004;10 (Suppl):S10–7. - PubMed
    1. Neumann M, Sampathu DM, Kwong LK, Truax AC, Micsenyi MC, Chou TT, Bruce J, Schuck T, Grossman M, Clark CM, McCluskey LF, Miller BL, Masliah E, Mackenzie IR, Feldman H, Feiden W, Kretzschmar HA, Trojanowski JQ, Lee VM. Ubiquitinated TDP-43 in frontotemporal lobar degeneration and amyotrophic lateral sclerosis. Science. 2006;314:130–3. - PubMed
    1. Mackenzie IR, Rademakers R, Neumann M. TDP-43 and FUS in amyotrophic lateral sclerosis and frontotemporal dementia. Lancet Neurol. 9:995–1007. - PubMed
    1. Lagier-Tourenne C, Polymenidou M, Cleveland DW. TDP-43 and FUS/TLS: emerging roles in RNA processing and neurodegeneration. Hum Mol Genet. 19:R46–64. - PMC - PubMed

Publication types

MeSH terms

Substances

Grants and funding