Study design

In the current work we thoroughly characterized the inhibitory effect of human CSF on α-syn aggregation by means of several complementary techniques. The study design of the present work is summarised in Fig. 1. At first, we collected more evidence supporting the fact that CSF is naturally capable of inhibiting α-syn aggregation in both SAA reaction mix and phosphate buffered saline (PBS), both in seeded and unseeded conditions, and in a patient-dependent manner (Fig. 1A.1). Starting from the observation that two CSF samples collected from normal-pressure hydrocephalus (NPH) patients with marked difference in proteins content produced different α-syn ThT fluorescence profiles in PBS (Fig. 1A.2), we performed fractionation of a pool of CSF collected from neurological controls by mean of centrifugal filters (Fig. 1B.1). We then analysed different fractions by means of mass spectrometry and Protein aggregation assays (Fig. 1B.2–3). We determined that the fraction corresponding to molecular weight above 100 kDa, rich in apolipoproteins, albumin and transthyretin (TTR), retained most of the inhibitory effect of neat CSF. By means of Protein aggregation assays, Western blot (WB), dot blot (DB), transmission electron microscopy (TEM), and solution nuclear magnetic resonance (NMR) spectroscopy we determined that high-density and low-density lipoproteins (HDL and LDL) are highly inhibitory against α-syn aggregation and likely interact with α-syn oligomeric species (Fig. 1.C.1–3). Subsequently, we tested the impact of varying (within physiological ranges) concentrations of albumin, TTR, HDL and LDL on ultrasensitive diagnostic SAA (Fig. 1C.4). We then used immunoprecipitation (IP), to deplete the two most abundant CSF apolipoproteins (ApoA1 and ApoE) from a newly made CSF pool and tested the effect with Protein aggregation assays (Fig. 1D.1). Finally, we measured total protein content, ApoA1 and ApoE concentrations in SAA-negative CSF samples of a small cohort of neurological controls and repeated SAA on the same samples spiked with preformed aggregates to evaluate possible correlation between SAA kinetic parameters and CSF total protein content, ApoA1, and ApoE concentrations (Fig. 1D.2).

Protein expression and purificationRecombinant α-syn expression and purification (protein aggregation assay in PBS)

Escherichia coli BL21 (DE3) Gold were transformed with a pT7-7 vector cloned with the gene encoding α-syn. The overnight preculture of transformed cells was diluted 100-fold in LB medium and induced at an OD600 value of 0.6–0.8 with 1 mM isopropyl-β-d-thiogalactoside; after 5 h of incubation at 37 °C, the cells were harvested at 4000 rpm (JA-10, Beckman Coulter). The extraction was carried out through osmotic shock using 100 mL of the buffer Tris 30 mM, EDTA 2 mM, and sucrose 40%, at pH 7.2, according to Shevchik et al. [14] and Huang et al. [15].

The suspension was then ultra-centrifuged at 20 000 rpm (Type 70 Ti rotor, Beckman Coulter) for 25 min, and the pellet was collected and resuspended with 90 mL precooled ultrapure water containing 38 μL of 1 M MgCl2 and then ultra-centrifuged a second time. Supernatants derived from these two centrifugation steps were combined and dialyzed against 4 L of 20 mM Tris/HCl buffer at pH 8.0. The protein was then loaded in the fast protein liquid chromatography system, and anion-exchange chromatography was carried out with 0–50% linear gradient 1 M NaCl (GE Healthcare HiPrep Q HP 16/10 Column). The collected fractions were lyophilized and resuspended in 10 mM Tris/HCl, 1 mM EDTA, and 8 M urea at pH 8.0 for chemical denaturation. To eliminate all of the protein that formed aggregates, two size-exclusion chromatographies (HiLoad 16/600 Superdex 75 pg Column) were performed with 20 mM phosphate and 0.5 mM EDTA at pH 8.0 as the elution buffer. Purified α-syn was dialyzed against Milli-Q water and lyophilized in batches for long-term storage. The Roche complete protease inhibitor cocktail was added only during the extraction step in the quantity suggested by the producer.

15N-labeled wild-type α-syn was expressed in Escherichia coli grown in M9 minimal medium supplemented with 15NH4Cl and purified as started for E. coli in LB medium.

For 15N-labelled and unlabelled α-syn, protein expression and purification were performed as previously described [16]. 1H solution NMR, gel electrophoresis, and silver staining were used to check the quality of the purified protein. An image of two replicate silver staining experiments performed on the purified α-syn after one and two (used in Protein aggregation assays) size-exclusion chromatography steps is shown in Fig. S1 (Supplementary Material).

Recombinant transthyretin (TTR) expression and purification

For TTR, Escherichia coli BL21(DE3) RIPL PLysS cells were transformed with pET-28a( +) plasmid encoding TTR gene. The cells were cultured in LB Medium containing 0.1 mg/mL of Kanamycin, grown at 37 °C, until OD600 reached 0.6–0.8, then induced with 1 mM isopropyl β-D-1-thiogalactopyranoside. They were further grown at 37 °C overnight and then harvested by centrifugation at 6500 rpm (JA-10 Beckman Coulter) for 15 min at 4 °C. The pellet was suspended in 20 mM Tris–HCl, pH 8.5 (60 mL per litre of culture) and sonicated at 4 °C for 40 min. The suspension was centrifuged at 40,000 rpm (F15-6 × 100y Thermo Scientific) for 40 min and the pellet discarded. TTR was purified by anionic-exchange chromatography using a HiPrep Q FF 16/10 column (GE Healthcare Life Science). The protein was eluted in 20 mM Tris–HCl buffer at pH 8.6 with a linear 0–1 M NaCl gradient. Fractions containing pure TTR were identified by Coomassie staining SDS-PAGE gels, then joined and purified by Size Exclusion Chromatography using HiLoad Superdex 26/60 75 pg in 50 mM phosphate buffer at pH 7.5.

Neurological control CSF samples

The neurological control (NC) CSF samples used in this work had been previously collected and stored at -80 °C according to international guidelines [17]. NC were cognitively unimpaired subjects referring to the Centre for Memory Disturbances of the University of Perugia (Perugia, Italy), who underwent lumbar puncture for subjective memory complaints not confirmed by the neuropsychological assessment, or as part of a diagnostic work-up for minor neurological symptoms (i.e., headache, peripheral neuropathy, psychiatric disorders), showing no cognitive impairment after at least a 2-year follow-up. All the selected NC samples tested negative for both classical AD CSF biomarkers (amyloid 1–40 and 1–42 peptides ratio, total tau, and T181-phosphorylated tau) [18] and for α-syn SAA [19].

CSF samples used in fractionation experiments

CSF from 19 different NC subjects (10 females and 9 males, average age = 70 y, standard deviation = 8 y) were pooled (CSF pool 1) reaching a total volume of 8 mL that was split in 2 aliquots of 3 mL and 10 aliquots of 0.2 mL.

CSF samples used for immunodepletion experiments

A second CSF pool of 5 mL was prepared from different CSF samples collected form n = 10 NC subjects (5 females and 5 males, average age = 69 y, standard deviation = 5 y). This second pool was then split in 10 aliquots of 0.5 mL each, which were then used for ApoA1 and ApoE immunodepletion experiments and protein aggregation assays.

CSF samples used for ELISA and SAA experiments

Two aliquots of 0.5 mL relative to 31 CSF samples collected from NC subjects (7 females and 24 males, average age = 69 y, standard deviation = 8 y), were selected for ApoA1 and ApoE ELISAs, total protein measurement, and SAA experiments.

Healthy control and PD CSF reference samples

CSF collected from healthy control (HC) and PD subjects was used to perform initial α-syn seed spiking experiments and to test the impact of adding HDL, LDL, HSA, and TTR at Amprion Inc. (San Diego, CA, U.S.). The HC1-6 samples used in α-syn seed spiking experiments shown in Fig. 2A have been purchased from Biochemed Services (Winchester, VA, U.S.) and were collected from 3 male and 3 female subjects (age = 26–39 y). With reference to the experiments summarised in Fig. 8, HC22 (female, age = 35 y) and HC24, HC66, HC67 (all males, age = 30–35 y) were also purchased from Biochemed Services. All neat HC CSF samples tested negative in α-syn SAA. With reference to the same experiments, PD10, PD34, PD62 CSF samples (all males, age = 72–79) belonged to PD patients with an Hoehn and Yahr (H&Y) stage of 2 and were purchased from PrecisionMed (Carlsbad, CA, U.S.). PD47 (PD patient, female, age = 62 y, H&Y = 1.5) was instead purchased from BioIVT (Westbury, NY, U.S.). All neat PD CSF samples tested positive in α-syn SAA.

CSF Fractionation procedure

An aliquot of 3 mL of CSF pool 1 was resuspended in 1.5 mL of PBS 3 × in order to have 4.5 mL of human pooled CSF in PBS 1x. This volume was then subjected to a series of filtrations using Amicon® Ultra-4 molecular weight cut-off (MWCO) filters. The procedure used to fractionate human CSF is schematized in Fig. 4A. The aliquots collected in this way contained the different constituents of the starting 4.5 mL of CSF in PBS with different concentration factors, the volume and the relative concentration factors (with respect to fraction 1) of the aliquots depicted in Fig. 4A are summarised in Table S1. Aliquots 2, 3, 4 and 5 were washed 3 times with PBS before storage. The ability to interact with α-syn monomers was then tested for all the CSF fractions. The different concentration factors were adjusted by diluting the samples with PBS (see Supplementary Material Table S1).

Generation of preformed α-syn aggregates

Preformed α-syn aggregates were generated by incubating 1 mg/mL of α-syn in PBS for one week at 37 °C under vigorous double orbital shaking (500 rpm) in a sealed 1.5 mL polypropylene vial. The final products were subjected to cycles of sonication (20 s tip sonication, 20 s rest) with an amplitude of 12 μm. The polypropylene vial had been immersed in ice for the whole duration of the sonication procedure. The aggregates were then diluted at 0.25, 2.5, 25, 250 and 2500 pg/μL, considering the initial monomer concentration as reference. The generated α-syn aggregates were then aliquoted and stored at -80 °C.

α-Synuclein Seed Amplification Assay (αS-SAA)

Samples were analysed as previously reported [19, 20]. Briefly, CSF samples were evaluated in triplicate (40µL/well) in a 96-well plate (COSTAR 96, cat# 3916), in a reaction mix consisting of 0.3 mg/mL recombinant α-syn (Amprion, cat# S2020), 100 mM PIPES pH 6.50 (Sigma, cat# 80,635), 500 mM NaCl (Lonza, cat# 51,202), 10 µM ThT (Sigma, cat# T3516), and a 3/32-inch BSA-blocked Si3N4 bead (Tsubaki Nakashima). Beads were blocked with 1% BSA in 100 mM PIPES pH 6.50 for 1 h, washed twice with 100 mM PIPES pH 6.50, and dried overnight. This assay was performed in a BMG FLUOstar Omega shaker/reader at 37 °C, plates were shaken for 60 s every 30 min for 150 h. The assay outcomes of the assay are positive, inconclusive, or negative, based on a probabilistic algorithm that uses maximum fluorescence and kinetic parameters [19]. With respect to other Protein aggregation assays performed in this work, the ultrasensitive diagnostic SAA used a different αSyn substrate (C-terminal histag), which was purified by standard IMAC procedures (Amprion, cat#S2020) [19, 20].

ThT protein aggregation assays in PBS

The protein aggregation experiments used to characterize interaction between α-syn and CSF constituents were performed with a programmable BMG LABTECH ClarioStar® fluorometer in Greiner clear-bottom 96-well plates (cat# 655,906). The ThT fluorescence was read from the bottom using excitation and emission wavelengths of 450 and 480 nm, respectively. An incubation temperature of 37 °C was used for all the experiments. Slightly different gain values were used to avoid overflow of the analog-to-digital converter. In each experiment, lyophilized recombinant α-syn was thawed in 3 mM NaOH at the concentration of 3.5 mg/mL. The solution was brought to physiological pH by diluting it with concentrated PBS (4x) and distilled water. In all the experiments described, the final reaction volume was 200 μL, α-syn final concentration was 0.7 mg/mL and ThT final concentration was 10 μM. To avoid bacterial contamination, 0.1% NaN3 was also present in all the tested conditions. All reactions were performed at 37 °C in sealed plates. Each condition was tested at least in triplicate.

Testing of the inhibitory effect of CSF and CSF fractions

158 μL of the solution containing monomeric α-syn was poured in wells, each of them containing 6 glass beads of 1 mm diameter. Depending on the experiment, 40 μL of human pooled/NPH CSF, 40 μL of PBS or 40 μL of CSF fractions were added. In seeded experiments, 2 μL of PBS containing 0, 0.25, 2.5, 25, 250, and 2500 pg/μL of α-syn aggregates were added. Plates were sealed and subjected to cycles of shaking (1 min double-orbital shaking at 500 rpm, 14 min rest) inside the fluorometer.

Protein aggregation assays with human serum albumin (HSA) and high-density lipoproteins (HDL)

To compare the results with those of a previously published paper, the experiment was performed in the exact same way as it is described in Bellomo et al. [16]. With respect to the above-described experiments the only notable difference consisted in applying a shaking/incubation protocol of 29 min rest and 1 min shaking. In addition to HSA (Sigma Aldrich, A1653), wells containing 0.12 mg/mL and 0.57 mg/mL human serum HDL (Sigma Aldrich, LP3) with and without α-syn were also analysed.

Protein aggregation assays with HDL, low-density lipoproteins (LDL) and TTR

In these experiments 40 μL of solutions containing different dilutions (all the products were diluted with distilled H2O) of human serum HDL (Sigma-Aldrich LP3), human serum LDL (Sigma-Aldrich LP2), recombinant TTR were added. Each condition was replicated in 3 distinct wells in the presence and absence of α-syn. For the experiments in which we tested the effect of different concentrations of human serum HDL (0, 0.003, 0.03, 0.3 and 1 mg/mL) the plate was sealed and subjected to cycles of 1 min double-orbital shaking at 500 rpm, 14 min rest. For the experiments in which we tested the effect of different concentrations of human serum HDL (0.3 mg/mL), LDL (0.03 and 0.3 mg/mL) and TTR (1.0, 0.3 and 0.03 mg/mL) the shaking protocol was changed to 2 min double-orbital shaking at 500 rpm and 13 min rest at 37 °C.

Analysis of ThT fluorescence profilesProbabilistic algorithm for α-syn SAAs

During the ultrasensitive SAA experiments performed at Amprion Inc., fluorescence readings were collected every 30 min to estimate kinetic parameters with high accuracy. The following 4 parameter sigmoid was used to fit the raw fluorescence readings:

$$\mathrm\left(\mathrm\right)=}_}+\frac}_}-}_}}}_}}\right)}^}}$$

where Fmin is the minimum fluorescence, Fmax is the maximum fluorescence, T50 is the time to reach 50% Fmax, S is the slope, and t is time. The coefficient of determination (R2) was calculated for each fitting. The α-syn SAA result of each CSF sample was determined using a probabilistic algorithm:

$$}_}=\frac}^+\mathrm*}_+\mathrm*}_}}}^+\mathrm*}_+\mathrm*}_}}$$

where Ppos is the probability for a replicate to be positive, A = -4.02, B = 2.98, C = 1.87, and Fmax5000 and Rsquare93 are binary values depending on a threshold. If the Fmax of a given replicate is above 5,000 RFU, then Fmax5000 = 1, else 0. If the R2 for the fitting of the 4-parameter model to the fluorescence data is above 0.93, then Rsquare93 = 1, else 0. The coefficients A, B, and C were estimated using a database of more than 900 aggregation curves. Among the PD and HC samples in the database, Fmax and the R2 showed statistically significant differences (p-value < 0.001 for each of these variables). If the probability for positivity (Ppos) is higher than 0.12, the replicate is determined positive, otherwise it is determined negative. Since CSF samples were analysed in triplicates, if all 3 replicates were positive the CSF sample was called positive. If none or one replicate was positive the CSF sample was called negative. If two replicates were positive and the average Fmax of the 3 replicates was less than 5,000 RFU (or a.u.) or the coefficient of variation (CV) was higher than 110, the sample was also called negative.

Double sigmoid fitting

The average background fluorescence produced by three wells containing the analyte without α-syn was subtracted prior to the analysis from the ThT intensity profiles relative to the same analyte in the presence of α-syn. While analysing data relative to the sole α-syn the background fluorescence from well containing only the reaction buffer was subtracted. Each ThT kinetic trace was then fitted with a double sigmoid function using Origin Pro v9.0. In the fitting model, A2 fits the fluorescence value of the second plateau, A1 fits the fluorescence value of the first plateau and A0 fits the baseline fluorescence. The time parameters t1 and t2 fit the first and the second inflection points, respectively, while d1 and d2 represent the slopes of the sigmoids. In the non-linear fitting procedure used, the following bounds were applied: 0 < A0 < 1000, 500 < A1 < 5000, A2 > 2000, 0 < t1 < 100 h and t2 > 0. For some kinetic traces, a decrease in fluorescence was observed after reaching the second plateau. This known phenomenon is caused by the sequestration of ThT molecules by mature fibrils and by the sedimentation of HMW insoluble aggregates [21]. In these cases, the last descending part of the ThT profile was removed prior to fitting. Fitting was rejected when the adjusted determination coefficient R2 was below 0.3.

WB and dot-blot assays

Equal amounts of assay products (volumes containing 2 µg of α-syn) were added with Laemmli’s sample buffer without sodium dodecyl sulphate (SDS), without boiling them to prevent solubilization of SDS-sensitive aggregates. Samples were separated through SDS-PAGE on 4–20% polyacrylamide gels (Bio-Rad) and transferred into PVDF membranes (0.45 μm, Bio-Rad) by wet transfer at 100 V constant for 90 min using 25 mM Tris–HCl, 192 mM glycine, 20% methanol, and 0.015% SDS. Membranes were fixed with 4% PFA for 30 min prior to blocking with 5% non-fat milk in TBS-T (TBS with 0.1% Tween 20) for 1 h at room temperature. After blocking, filters were incubated with primary antibody against α-syn (211, sc-12767, Santa Cruz Biotechnology) O/N at 4 °C. The membranes were further incubated with a goat-anti-mouse IgG-HRP conjugate (Bio-Rad, 1,706,516; 1:5000) secondary antibody for 1 h at RT, and signals were visualized using an ECL reaction. Densitometric analysis was performed with ImageJ software (National Institute of Health). Dot-blot images were inverted, and window/level adjusted prior to the analysis. A rectangular region of interest (ROI) was used to measure the average grey level intensity in each band relative to monomeric α-syn. The average grey level of an equivalent ROI not containing bands was subsequently subtracted. To estimate the approximate percentage of remaining α-syn monomer, the average intensity of the WB bands was divided by the one at t = 1 h.

For dot blotting aliquots of products obtained in SAAs corresponding to 300 ng of monomeric α-syn in the initial reaction mixtures were spotted (2 μL/spot) on nitrocellulose membrane pre-equilibrated with TBS-T. Samples were dried at RT and fixed with PFA (0.4% in PBS) for 30 min, and then filters were blocked with 2% non-fat milk (in TBS-T). The membranes were incubated overnight at 4 °C with OC (1:1000) or A11 (1:1000) conformational antibodies [22] followed by incubation with goat-anti-rabbit IgG-HRP conjugate (1:5000; 1,706,515 Bio-Rad) secondary antibody for 1 h at RT. Membranes were developed with an ECL reaction and images were acquired through a ChemiDoc™ imaging system. Dot-blot images were inverted, and window/level adjusted prior to the analysis. The average grey level intensity was extracted from a circular ROI containing all the α-syn containing samples. For each dot, the average grey level intensity of an adjacent ROI not containing any dot was subtracted to remove the background intensity. The averaged and background-subtracted grey level intensity was then multiplied for the ROI area in cm2 to estimate the integrated density. No significant grey level fluctuations were found while repeating the same procedure for control samples not containing α-syn.

Depletion of CSF ApoA1 and ApoE by immunoprecipitation

Immunoprecipitation (IP) was performed to deplete ApoA1 and ApoE from CSF pool sample. 50 µL of settled Immobilized Protein A/G (100 µL resin slurry, Pierce™ Protein A/G Plus Agarose, ThermoFisher Scientific™, USA) and 60 µg of anti-ApoA1 antibody (MIA1404, ThermoFisher Scientific™, USA) were combined in a 2-mL tube. The beads-antibody slurry was incubated for 4 h at 4 °C (constant rotation). The tube was centrifuged at 1,000 × g for 2 min at 4 °C and the supernatant was discarded. The bead pellet was washed with Phosphate Buffered Saline (PBS) twice. Similarly, 50 µL of settled Immobilized Protein A/G (100 µL resin slurry, Pierce™ Protein A/G Plus Agarose, ThermoFisher Scientific™, USA) and 20 µg of anti-ApoE antibody (PA5-27,088, ThermoFisher Scientific™, USA) were combined a 2-mL tube. The beads-antibody slurry was incubated for 4 h at 4 °C (constant rotation). The tube was centrifuged at 1,000 × g for 2 min at 4 °C and the supernatant was discarded. The bead pellet was washed with PBS twice. Subsequently, ApoA1-antibody-bound beads and ApoE-antibody-bound beads were combined in a single 2-mL tube. The solution was gently mixed and centrifuged at 1000 × g for 2 min at 4 °C. The supernatant was discarded and 400 µL of CSF pool was added to the tube with beads. The sample underwent overnight incubation at 4 °C (constant rotation). At the end of incubation, the CSF-beads slurry was centrifuged (1,000 × g for 2 min at 4 °C) and the supernatant and bead pellet were collected separately for further analysis. In parallel, a control sample was prepared by following the same procedure on beads without anti-ApoA1 or anti-ApoE antibody bound to IP. Briefly, 50 µL of neat resin slurry was transferred to 2-mL tube and washed with PBS twice. After the second wash cycle, PBS was removed and 400 µL of CSF pool was added instead. The sample underwent overnight incubation at 4 °C (constant rotation). At the end of incubation, the sample was centrifuged (1,000 × g for 2 min at 4 °C) and the supernatant and bead pellet were collected separately for further analysis. Performance of ApoA1 and ApoE IP was assessed by Western Blot. Briefly, all the samples (1. IP CSF pool sample, 2. CSF pool sample undergoing procedure mimicking IP (without antibody), 3. bead pellet of IP CSF pool sample, 4. bead pellet of CSF pool sample undergoing procedure mimicking IP (without antibody), 5. neat CSF pool sample which did not undergo any treatment) were resuspended in a reducing loading buffer and boiled 5 min at 95 °C (thermoblock). Tubes with bead pellet (samples 3. and 4.) were centrifuged and the supernatant was collected and used for further analysis. A 12% polyacrylamide gel was prepared, and the samples were run in reducing conditions. All samples were transferred from the gel to a nitrocellulose blotting membrane (SF110B, Himedia, India). The quality of the transfer was evaluated by the reversible Ponceau S staining. The membrane was blocked and subsequently probed overnight at 4 °C (shaker) with the selected primary antibodies (MIA1404 for ApoA1, PA5-27,088 for ApoE, 1:1000) diluted in 5% bovine serum albumin (BSA), 0.02% NaN3 in Tris-buffered saline with 0.1% Tween®20 (Sigma-Aldrich, USA) (TBST) with addition of a phenol red as a pH indicator. After the incubation, the membrane was washed with TBST and the secondary antibodies diluted 1:5000 in a blocking buffer were applied for 1 h at RT. Depending on the used primary antibodies, goat anti-mouse (170–6516, Bio-Rad, USA) (for MIA1404) or goat anti-rabbit IgG-HRP (170–6515, Bio-Rad, USA) (for PA5-27,088) were added. Signal development was performed with use of the enhanced chemiluminescence (ECL) solution (SuperSignal™ West Pico Plus, ThermoFisher Scientific™, USA).

ApoA1 and ApoE levels in CSF

CSF levels of ApoA1 and ApoE were assessed with use of the commercially available ELISA kits—Human Apolipoprotein AI ELISA Kit, ab108803 and Human Apolipoprotein E ELISA Kit, ab108813 (Abcam, UK) in n = 31 SAA-negative NC CSF samples. Both assays were performed according to the manufacturer protocol. All the standard curve points and CSF samples were run in duplicate. Optical density (OD) was read at 450 nm (wavelength correction 570 nm) by the Clariostar (BMG Labtech, Germany) plate reader. The 4-parameter logistic model was applied to generate a standard curve and interpolate concentrations of analysed CSF samples.

Total protein content of CSF samples

Total protein content was evaluated in the same samples with use of the Pierce™ 660 nm Protein Assay Reagent cat. 22,660 (Thermo Scientific™, USA). The assay was performed according to the manufacturer protocol. A set of BSA dilutions of known concentrations served as the standard to calculate total protein content of CSF samples. All the standard curve points and CSF samples were run in duplicate.

Statistical analysis

One-way analysis of variance (ANOVA) and Tukey post-hoc test for mean comparisons were applied while assessing differences among fitted kinetic parameters for different samples. Correlation between added seed mass and T2 parameters were computed according to Spearman. Two-tailed Student’s t-test was applied when comparing adjusted integrated densities of dot blot images. Standard error of mean (SEM) is reported in each image showing bar plots and/or average fluorescence profiles. In the analysis of mass spectrometry data, an false discovery rate (FDR) of 1% was imposed and the criterion used to accept protein identification included probabilistic score sorted by the software. Correlations among SAA parameters and Log2-transformed CSF levels of ApoA1, ApoE and total protein were computed by means of Pearson’s correlation coefficients. Ward linkage criterion was applied for hierarchical clustering of correlations.