The following antibodies were used in this study, including anti-actin monoclonal antibody (mAb) (Thermo Fisher Scientific, Rockford, IL, USA), anti-14-3-3 polyclonal antibody (pAb), anti-glycogen synthase kinase (GSK)-3β pAb, anti-heat shock protein 27 (Hsp27) mAb and anti-sirtuin 1 (Sirt1) pAb (all from Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), anti-GAPDH and horseradish peroxidase (HRP)-conjugated anti-rabbit or anti-mouse immunoglobulin G (Thermo Fisher Scientific).

The stored frozen brain tissue from normal healthy hamsters (n=3) and hamsters inoculated intracerebrally with hamster-adapted scrapie agent 263K (n=3), as well as the post-mortem cortex and cerebellum of patients with sporadic CJD (sCJD, n=1), FFI (n=3), G114V genetic CJD (G114V gCJD, n=1) and a healthy subject (n=1) were enrolled in this study. Brain homogenates (10%, w/v) were prepared based on a previously described protocol (27). Briefly, brain tissues were homogenized in lysis buffer (100 mM NaCl, 10 mM EDTA, 0.5% Nonidet P-40, 0.5% sodium deoxycholate, 10 mM Tris, pH 7.5) containing 1% protease inhibitor cocktail (Abcam, Cambridge, MA, USA). The tissue debris was removed by low-speed centrifugation at 2,000 × g for 10 min and the supernatants were collected for further analysis.

SNO proteins in the brain tissue were determined using a commercial SNO protein detection assay kit (Cayman Chemical, Inc., Ann Arbor, MI, USA) following the manufacturer’s instructions. Briefly, all tested samples were normalized to 100 mg (approximately 1 ml 10% brain homogenate/sample) total brain proteins and transferred to 1.5-ml Eppendorf tubes. The brain samples were homogenized in 0.5 ml buffer A containing blocking reagent (thiol-blocking reagent at 2 mg/ml). Following incubation at 4°C for 30 min in the dark, the samples were clarified by centrifugation at 4°C for 10 min. The supernatants were transferred to 2-ml tubes and mixed with pre-cooling acetone (1:4, v/v), and total proteins were precipitated at −20°C for 2 h. Following centrifugation at 4°C for 15 min, the pellet was resuspended in 0.5 ml buffer B containing reducing and labeling reagents, and further incubated at room temperature for 2 h on a rocker with gentle agitation. Acetone precipitation was conducted again as described above. The protein pellet was thoroughly dissolved in dissolving buffer (8 M urea, 30 mM HEPES, 10 mM DTT, 2 mM EDTA and 1 mM PMSF, pH 8.2) at the same volume as the brain homogenates initially used. The products were stored for further analysis.

To pull-down SNO proteins, streptavidin-conjugated magnetic beads (2.8 μm, Dynal magnetic beads; Invitrogen Life Technologies Corporation, Carlsbad, CA, USA) were used according to the manufacturer’s instructions. Briefly, the beads were washed 3 times with phosphate-buffered saline (PBS, pH 7.4) containing 0.01% (v/v) Tween-20 and added to the samples mentioned above. The mixture was rotated at room temperature for 1 h. Bound beads were washed 5 times in PBS containing 0.1% bovine serum albumin (BSA) (w/v) and then resuspended in the same volume of PBS as the initial volume of beads used.

Four different elution buffers with different elution conditions were tested for releasing SNO proteins from streptavidin beads: i) incubation in non-ionic water at 75°C for 5 min as previously described by Holmberg et al (28); ii) incubation in 10 mM EDTA pH 8.2 with 95% formamide at 90°C for 2 min as described in the instructions provided with the Dynabeads^® M-280 (Life Technologies); iii) incubation in HEPES elution buffer (20 mM HEPES-NaOH, 100 mM NaCl, 1 mM EDTA, 100 mM 2-ME, pH 7.7) at room temperature for 20 min as previously described by Xu et al (29); iv) incubation with various concentrations of sodium dodecyl sulfate (SDS; 0.1, 0.2, 0.5 and 1% SDS) at 100°C for 5 min. SNO proteins eluted with 1.0 ml of various elution buffers were precipitated with acetone at −20°C for 2 h. Following centrifugation at 10,000 × g at 4°C for 15 min, the pellets were resuspended with dissolving buffer at the same volume as the brain homogenates initially used.

Three different bead/homogenate ratios (v/v) were tested, including 1:6, 1:3 and 1:1. In addition, 3 different incubation times at 2 incubation temperatures were evaluated: 4°C for 30, 60 or 120 min; and 25°C for 30, 60 or 120 min. All other experimental conditions were maintained invariable.

The protein concentration of the redissolved products was determined in triplicate using the bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific) according to the manufacturer’s instructions.

SNO protein samples were mixed with 5X loading buffer and boiled for 8 min. Proteins were separated in 15% SDS-polyacrylamide gel electrophoresis (PAGE) and transferred onto nitrocellulose (NC) membranes (Whatman, Pittsburgh, PA, USA) by the semi-dry method in transfer buffer and immunoblotted with anti-biotin antibody (Cayman Chemical, Inc.). Individual SNO proteins were measured by specific antibodies, including anti-14-3-3 used at 1:1,000, anti-GSK-3β used at 1:1,500, anti-Hsp27 used at 1:2,000, anti-Sirt1 used at 1:800, anti-β-actin used at 1:5,000 and anti-GAPDH antibodies used at 1:2,000. Reactive signals were visualized using an enhanced chemiluminescence (ECL) kit (Amersham-Pharmacia Biotech, Piscataway, NJ, USA).

One microliter of the isolated SNO proteins was dotted onto NC strips, which were allowed to dry at room temperature. The strips were blocked with 2% BSA (w/v) in Tris-buffered saline containing 0.5% Tween-20 (TBST, w/v) and incubated overnight at 4°C to block the residual binding sites on the paper. Subsequently, the strips were further incubated for 1 h at 37°C with anti-biotin antibody diluted 1:1,000 in 2% BSA. The strips were washed 3 times with TBST and reactive signals were visualized using an ECL kit.

iTRAQ labeling for human brain specimens was carried out using the iTRAQ^® Reagent-8Plex kit (AB Sciex, Framingham, MA, USA) according to the manufacturer’s instructions. The purified SNO proteins of the cerebellum and cortex from the normal control were labeled with iTRAQ labeling reagent 113 and 117, while those of the cerebellum and cortex from the patients with sCJD, FFI and gCJD were labeled with iTRAQ labeling reagents 114 and 118, 115 and 119, 116 and 121. The amounts of the purified SNO proteins were 100 μg/label. The labeled products were digested with trypsin at a ratio of 1:20 (w/w, trypsin/protein) at 37°C for 36 h. The digested peptides were subsequently dried and reconstituted with 0.5 M triethylammonium bicarbonate (TEAB; Sigma, St. Louis, MO, USA) and 0.1% SDS. The dried peptides were then labeled with respective isobaric tags, and incubated at room temperature for 1 h before being combined.

To remove all interfering substances, such as dissolving buffer, ethanol, acetonitrile, SDS, excess iTRAQ reagents, strong cation exchange chromatography (SCX) was carried out for the combined iTRAQ-labeled peptides using the cation exchange system provided in the iTRAQ method development kit (AB Sciex). The eluted fraction was desalted using Sep-Pak C18 cartridges, dried and then reconstituted with 0.1% formic acid (FA) for nano-flow liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis coupled with the ultimate LC system using Q-Exactive Mass Spectrometer (Thermo Fisher Scientific). Peptide separations were performed in a 100×75 mm column (BEH130 C18) using mobile phase A (0.1% FA in LC-MS grade water) and mobile phase B (0.1% FA in LC-MS grade ACN). The flow rate was set at 400 nl/min. The LC fluent was directed to the electrospray ionization (ESI) source for quadrupole mass spectrometry (Q-MS) analysis, using precursor ions that were selected across the mass range of 350–2,000 m/z with 250 msec accumulation time/spectrum. A maximum of 20 precursors per cycle from each MS spectra was selected for MS/MS analyses with 100 msec minimum accumulation time for each precursor and dynamic exclusion for 15 sec.

The software used for data interpretation was proteome discoverer version 1.3 (Thermo Fisher Scientific) and Mascot version 2.3.0. The database searched was the peptide sequence library in the Swissport database restricted to the human protein sequence data set.

Usage of the stored human and animal brain specimens in this study was approved by the Ethics Committee of the National Institute for Viral Disease Prevention and Control, China CDC under protocol 2009ZX10004-101. All Chinese golden hamsters were maintained under clean grade. Housing and experimental protocols were in accordance with the Chinese Regulations for the Administration of Affairs Concerning Experimental Animals.

Several methods have been described for eluting biotin-labeled biological macromolecules from streptavidin beads, such as ddH₂O, EDTA, HEPES or SDS. To evaluate the eluting abilities of the different methods, 30 μl of brain homogenate of normal healthy hamsters (roughly 30 μg total proteins) were subjected to BST for biotin-labeling and bound with streptavidin-conjugated beads. The biotin-labeled SNO proteins bound on the beads were separately eluted with various eluting conditions and the biotin signals in the eluting fractions and beads were examined by western blot analysis for biotin. After being eluted with ddH₂O, only a small portion of biotin-positive signals was detected in the fraction of elution (Fig. 1A, left panel). More biotin signals were observed in the elution fraction following treatment with 10 mM EDTA with 95% formamide (pH 8.2), but there were still biotin signals left in the beads (Fig. 1A, middle panel). In the preparation of HEPES, almost all biotin signals were observed in the fraction of beads (Fig. 1A, right panel). These results indicate that it is not possible to completely elute the bound biotin-SNO protein from streptavidin beads with these 3 buffers.

Subsequently, the eluting capacity of the SDS buffer was evaluated based on the same experimental conditions. Following treatment with various concentrations of SDS buffer, the SNO proteins in the elution fraction and beads were evaluated by biotin-specific western blot analysis. The results revealed that the biotin signals were distributed almost equally in the fractions of elution and beads in the reaction of 0.1% SDS, but increased significantly in the fraction of elution in the reaction of 0.2% SDS. In the reactions of 0.5 and 1.0% SDS, all biotin signals were observed in the fraction of elution (Fig. 1B). Considering that a high concentration of SDS may affect the subsequent MS analysis (30) and other assays, we used 0.5% SDS as the elution buffer to break biotin-streptavidin interaction for further experiments.

To simplify the detection of SNO proteins, we used the biotin-specific dot blot technique instead of western blot analysis. In the reaction of 0.5% SDS, biotin signals were merely observed in the eluting products, but not in the beads (Fig. 1C); these results were consistent with those of western blot analysis.

To optimize the volume ratio of the streptavidin beads to the brain homogenate (10% w/v), the protein capturing abilities of the different working volume ratios (bead/homogenate, 1:6, 1:3 and 1:1) were examined. Different amounts of bead were mixed with 30 μl brain homogenates prepared from normal or 263K scrapie-infected hamsters after being labeled with biotin by BST. After being eluted with 0.5% SDS buffer, the protein concentrations in the eluted products were measured. As shown in Fig. 2A, only very small amounts of protein were detected in the eluted products at a ratio of 1:6, whereas a markedly larger amount of protein was eluted in the reactions at a ratio of 1:3 and 1:1. The protein-capturing abilities were approximately the same between the reactions of ratios 1:3 and 1:1 in either the normal or infected brain homogenates. Based on higher cost and effective ratio, we selected the ratio of 1:3 (bead/homogenate) as the optimal working volume ratio.

To optimize the incubation temperature and incubation time for the streptavidin beads with biotin-labeled brain homogenate, 10 μl of beads were mixed with 30 μl normal or 263K-infected hamster brain homogenates at the volume ratio of 1:3 (bead/homogenate). Following incubation at 2 different temperatures (4 and 25°C) for 3 different periods of time (30, 60 and 120 min), the biotin-labeled proteins were eluted from the beads with 0.5% SDS buffer and the protein amounts were evaluated using the BCA method. The results revealed similar alternative curves of the eluted protein contents in both the normal and infected homogenates, along with the changes in incubation temperature and duration (Fig. 2B). It was apparent that incubation at 25°C yielded higher protein concentrations in the eluted products than at 4°C. The increase in the incubation time, particularly with the incubation temperature of 25°C, yielded more proteins in the elution products. Based on these data, incubation at 25°C for 120 min was regarded as the optimal working condition.

To optimize the number of incubations in order to recover all SNO proteins from the brain homogenates, 15 μl of BST-treated brain homogenates of normal hamsters (roughly 15 μg total proteins) were incubated with 5 μl of streptavidin beads at the optimized working conditions. Following elution and the discarding of used beads, the brain homogenates were continually incubated with newly input beads. This process was repeated for at least 15 rounds and 3 adjacent elution products were pooled for further assays. Measurement of the protein concentration in the pooled elution products revealed that the protein contents gradually decreased along with the number of incubations, in which almost no protein was detected in the last pooled elution sample (incubation number 13–15) (Fig. 3A). The biotin-labeled proteins in various pooled samples were assayed with biotin-specific dot blots. In line with the results of the protein measurement, biotin-specific signals were observed in the first 4 pooled elution samples from 3 individual brain homogenates, but not in the last one (Fig. 3B). This indicates that at least 11 rounds of continual incubations at this experimental condition are required in order to recover all SNO proteins from the rear brain extracts of the hamsters.

Furthermore, the number of incubations for the recovery of the whole SNO proteins from human brain homogenates were assessed with the same protocol, including the brain homogenates prepared from the cortex and cerebellum of the healthy subject, and patients with sCJD, FFI and G114V gCJD. In agreement with the results obtained with the hamster brain homogenates, biotin-specific western blot analysis confirmed that 0.5% SDS was able to elute almost all biotin-labeled proteins from the bound beads in human brain extracts (Fig. 4A). A total of 15 rounds of successive incubations of the individual BST-treated brain homogenates with beads were conducted under optimized working conditions. In agreement with the observations of the hamster brain homogenates, almost no protein was detectable in the elution sample pooled from 13 to 15 rounds in all tested extracts (Fig. 4B). Dot blots identified similar patterns, according to which the intensities of the biotin signals gradually decreased in the first 4 pooled samples and diminished in the last of all tested samples (Fig. 4C). This suggests that the optimized working conditions based on brain homogenates of hamsters are suitable for the purification of SNO proteins from human brain extracts, from either healthy subjects or patients with various prion diseases.

To address whether the isolated SNO proteins with the newly optimized working conditions can be used for further experiments, the immunoreactivities of the purified SNO proteins from human cortex regions were subjected to several specific western blot analyses, including for 14-3-3, GSK-3β, Hsp27, Sirt-1, actin and GAPDH. Specific bands were observed at the expected positions in all tested blots (Fig. 5), strongly indicating that the SNO proteins extracted under the newly optimized working conditions possess reliable immunoreactivities. The signal intensities of all tested proteins in the samples of prion diseases were significantly stronger than those of the normal control.

Subsequently, the isolated SNO proteins from the human cortex and cerebellum under the newly optimized working conditions were subjected to iTRAQ-based quantitative proteomics analysis. A total of 69,896 spectra was matched from a total of 448,298 spectra. After searching with Mascot software, a total of 1,509 proteins was identified from 9,265 unique peptides with high confidence (FDR <1%). Fig. 6 illustrates the MS/MS spectra of the peptides of 3 different proteins, representing neuroblast differentiation-associated protein AHNAK isoform 1 (peptide sequence, VDIDAPDVDVHGPDWHLK) with a molecular weight of 628.7 kDa (Fig. 6A), heat shock protein gp96 precursor (peptide sequence, GVVDSDDLPLNVSR) with a molecular weight of 90.1 kDa (Fig. 6B) and S100-B (peptide sequence, AMVALIDVFHQYSGR) with a molecular weight of 10.7 kDa (Fig. 6C). The reporter region of each MS/MS spectrum illustrated the signals of the peptides labeled with the different iTRAQ tags in various brain extracts at the expected positions (m/z), including 113 and 117 for the normal control, 114 and 118 for sCJD, 115 and 119 for FFI, and 116 and 121 for G114V gCJD. This indicates that the SNO proteins purified with the present modified method are suitable for proteomics analysis.