For Initial Studies, Ventricular CSF Specimens from 5 LAD and 5 Age-Matched NC Subjects

METHODS

For initial studies, ventricular CSF specimens from 5 LAD and 5 age-matched NC subjects were thawed at room temperature and 500 µL was subjected to molecular weight (MW) fractionation by centrifugation at 13,000 x g through 50 kDa cutoff filters. The fraction greater than 50 kDa for each sample was subjected to isoelectric focusing on 17-cm immobilized pH gradient (IPG) (pI 3-10) strips at 300 V for 5 hours, 1000 V for 1 hour, 2500 V for 1 hour, and 5000 V for 80,000 voltage-hours. Focused IPG strips were loaded onto 8-16% SDS-PAGE gels and subjected to electrophoresis at 16 mA for 30 minutes and 24 mA for 5 hours.

Gels were stained overnight with Sypro ruby, imaged using a VersaDoc system, and analyzed with PDQuest software. Spots of interest were isolated, subjected to in-gel trypsin digestion using sequencing grade trypsin, and analyzed using MALDI mass spectrometry or HPLC tandem mass spectrometry. Proteins were identified using the MASCOT search engine with the following restrictions: monoisotopic peptides, oxidative modification of methionine, a mass tolerance of 100 ppm, and one missed trypsin cleavage.

For Western blot analysis, 500 µL CSF from age-matched NC, DC, MCI, and LAD subjects were subjected to MW fractionation and the > 50 kDa fraction resuspended in water, freeze-dried and resuspended in 30 µL water. Total fractionated protein samples were separated on 5% to 20% linear gradient SDS-PAGE gels and transferred to nitrocellulose membranes. The blots were blocked for 3 hours in 5% dry milk/Tris-buffered saline with 0.05% Tween-20 (TTBS) and incubated overnight with a 1:2000 dilution of mouse anti-PDS raised against recombinant human PDS. The blots were washed 3 to 5 times in TTBS and incubated in a 1:5000 dilution of horseradish peroxidase labeled goat anti-mouse IgG and bands visualized using enhanced chemiluminescence. Images were scanned and staining intensities measured using Scion image analysis. The blots were stripped and reprobed for TTR using rabbit anti-human TTR (1:2000) raised against amino acids 1-147 of human TTR. Four to five control specimens were included on each gel for comparison with DC, MCI, and LAD specimens. Results were expressed as mean ± SEM % control for each gel and each antibody. To estimate levels of the PDS/TTR complex the individual percent control values for PDS and TTR were multiplied together.

For ELISA analysis, 42 wells of a 96-well microtiter plate were coated overnight with mouse anti-PDS raised against recombinant human PDS (1 µg/mL PBS containing 0.05% NaN3 [PBSN]). Because the ELISA is based on quantification of TTR complexed to PDS that is trapped by the anti-PDS antibody, purified human TTR (0, 25, 50,125, 250 and 500 ng/mL) was simultaneously coated in triplicate in the remaining 18 wells to establish a calibration curve. The plates were rinsed 5 times with PBSN and all wells blocked for 1 hour using 5% dry milk/15% normal goat serum in PBSN. Wells coated with anti-PDS were incubated 2 hours with 50 µL unprocessed CSF in triplicate. Wells containing standards were incubated with PBS. Following 5 washes with PBSN, rabbit anti-TTR raised against amino acids 1-147 of human TTR (1 µg/mL in blocking buffer) was added to all wells for 1 hour to interact with TTR coated in wells as standards or complexed to PDS. The plates were washed 5 times with PBSN and incubated with horseradish peroxidase conjugated goat anti-rabbit secondary antibody (1:1000 dilution/blocking buffer) for 1 hour. After 5 washes with PBSN, color was developed using 3,3',5,5'-tetramethylbenzidine color reagent (150 µL/well) for 30 minutes followed by addition of 50 µL stopping solution. Absorbance was measured at 450 nm using a multi-well UV-Vis plate reader. To limit edge effects, the outer wells of the plate were not used.

RESULTS

Analysis of molecular weight fractions greater than 50 kDa from ventricular CSF specimens from 5 LAD and 5 age-matched NC subjects using 2D-electrophoresis and mass spectrometry showed PDS and TTR in the >50 kDa fraction were significantly different in patients with LAD compared with NC subjects. The presence of lower MW proteins in the greater than 50 kDa fraction suggested that the proteins may be part of an aberrant complex or aggregate. To test this hypothesis, ventricular CSF from 8 autopsy-verified LAD and 7 age-matched NC subjects was subjected to MW fractionation, freeze dried, resuspended in 30 µL PBS, analyzed using Western blot, and probed for PDS and TTR. To verify specificity of the antibodies, representative LAD CSF specimens were subjected to Western blot analysis using antibodies preincubated with immunizing proteins or cross incubated with the opposite proteins. In E-figure 1-A preincubation of the PDS antibody with 1 mg recombinant human PDS caused a loss of immunostaining in the native protein (21 kDa) and the band at 55 kDa. In contrast, preincubation of the PDS antibody with TTR did not diminish immunostaining of either band (E-figure 1-B). Similarly, preincubation of the TTR antibody with 1 mg purified human TTR caused decreased immunoreactivity for monomeric and dimeric TTR and the proposed complexed band at 55 kDa (E-figure 1-C). Preincubation of the TTR antibody with purified PDS did not alter immunostaining of the monomeric or dimeric proteins or the 55 kDa band (E-figure 1-D).

Western blot analyses of ventricular CSF samples from 4 subjects with MCI and 12 DC patients including Parkinson’s disease (PD) (n = 2), progressive supranuclear palsy (PSP) (n = 2), pure dementia with Lewy bodies (DLB) (n=5), and frontotemporal dementia (FTD) (n = 3) showed similar staining patterns for PDS, 21 kDa (native) and 55 kDa (complexed) and TTR 16 and 32 kDa (monomeric and dimeric) and 55 kDa (complexed) as observed in NC subjects. Staining intensity of the ~55 kDa band was calculated as percentage of control staining for each gel for each of the antibodies. Because we hypothesized the amount of the PDS/TTR complex is indicative of AD, we calculated the product of the two percentages and found levels of the complex were ~3 times NC levels in MCI subjects and ~10 times NC levels in LAD subjects. Levels of the PDS/TTR complex in DC subjects measured by Western blot showed levels that were comparable to those observed in NC subjects (1.4 times NC). .

Figure Caption

E-Figure 1. Immunoreactivity of the 55 kDa band and the band representing native proteins is blocked by preincubation of the prostaglandin-d-synthase (PDS) antibody with recombinant PDS (A) but not purified transthyretin (TTR) (B) and by preincubation of the TTR antibody with purified TTR (C) but not recombinant PDS (D).