Ultra-sensitive detection of protein biomarkers for diagnosis of Alzheimer’s disease

Beta amyloid peptide, tau, and phosphorylated tau are well recognized as promising biomarkers for the diagnosis of Alzheimer’s disease (AD).


Synthesis of the turn-on fluorophore
Scheme S1. Synthetic route of indolium-derived cyanine dyes, SIM.
General Procedure All the solvents were dried by the standard methods whenever needed. 1 H NMR spectra were recorded using a Bruker-400 NMR spectrometer and referenced to the residue CDCl 3 7.26 ppm or DMSO-d 6 2.5 ppm. 13 C NMR spectra were recorded using a Bruker-400 NMR spectrometer and reference to the CDCl 3 77 ppm or DMSO-d 6 39.5 ppm.

Preparation of the silica-coated iron oxide nanoparticles
The iron oxide nanoparticles were prepared by co-precipitation of ferrous and ferric ion solutions in 1:2 molar ratios. The solution of 2.5 mM FeCl 2 4H 2 O and 5 mM FeCl 3 6H 2 O in 2 M HCl was added to 62.5 mL 0.7 M NH 4 OH under mechanical stirring. Stirring was allowed for 30 min. The nanoparticles were washed and redispersed in 12.5 mL distilled water, followed by the addition of three aliquots of 2.5 mL 1M tetramethylammonium hydroxide solution under mechanical stirring for 24 h. The nanoparticles were washed as follows: 800 µL nanoparticles solution was diluted with 4 µL 2M HCl, centrifuged and redispersed in distilled water. The silica coating was done by sol-gel reaction. An ethanolic solution of TEOS (98%, Aldrich) was added to a mixture of 4.85 mL NH 4 OH, 28.8 mL distilled water, 27.5 mL EtOH and 1.6 mL of the previously washed magnetic nanoparticles under mechanical stirring. Stirring was allowed for 4 h. The nanoparticles were washed with ethanol and distilled water respectively, and then redispersed in 5 mL distilled water.
Preparation of the iron oxide nanoprobes. A solution of 100 µL silica coated iron oxide nanoparticles were added to an ethanolic solution of (3-aminopropyl)triethoxysilane (Aldrich) under stirring at 70 °C for 24 h. The resulting nanoparticles were further functionalized by glutaraldehyde (70%, Aldrich) under stirring for 2 h. In between each step, the nanoparticles were washed twice and redispersed in PBS. The prepared nanoparticles were incubated with capture antibodies (11A50-B10 for Aβ 40 , 12F4 for Aβ 42 , BT2 for tau 441 and AT270 for ptau 181 ) for 2 h. The resulting nanoparticles were washed twice and redispersed in PBS.

Optimization of the immunoassay conditions.
To optimize the concentration of the turn-on fluorophore used for labelling, the immunocomplex were incubated with 1, 10, 20, 50, 100 and 200 µM dye. To determine the concentration of the detection probes that appropriate for the detection, 3 mg/mL, 600 µg/mL, 300 µg/mL and 30 µg/mL nanoprobes were used for the detection. To optimize the detection procedure, the co-incubation of the nanoprobes, target analyte and detection antibody and the separate incubation; with and without an extra washing step after the magnetic separation step were compared. To decide the concentration of the capture antibody for immunoassay, 100, 500 and 1000 pM capture antibody were added for the coupling with nanoparticles. To optimize the detection procedure, the coincubation of the nanoprobes, target analytes and detection antibody and the separate incubation; with and without extra washing step after the magnetic separation were compared.
To ascertain the reaction time for different step, the nanoprobes were incubated at 37 °C with the target analyte for 15, 30 and 60 min; the immunocomplex were incubated at 37 °C with the detection antibody for 15, 30, and 60 min. To study the matrix effect of the cerebrospinal fluid (CSF), the analyte was added into artificial CSF (aCSF, R&D Systems). To investigate whether excessive antibody places any significant were added for the immunoreaction.

Selectivity of the nanoprobes.
To study the selectivity of the detection assay, four samples were prepared, probes for the detection of Ab 42 incubated with (i) 0 fM beta amyloid proteins, (ii) 500 fM Ab 40 , (iii) 500 fM Ab 42 and (iv) mixture of the beta amyloid proteins with final concentration of 500 fM each at the optimal condition. To study the specificity of the antibody, four channels immobilized with (i) 50 µM Ab 40 , (ii) 50 µM Ab 40 with 500 nM Ab 42 antibody (iii) 50 µM Ab 42 and (iv) 50 µM Ab 42 with 500 nM Ab 42 were labelled with 500 µM SIM. The fluorescent images were captured by TIRFM-EMCCD system.

Detection of target protein biomarkers.
The calibration curve was established by correlating the average net intensity of fluorescent molecules in captured images at each concentration of spiked target analyte. The net intensity can be calculated by subtracting the intensity of 1×1 square pixel of the magnetic immunocomplexes from that of individual background area on the image. The average net intensity was obtained by taking average of the net intensities of 100 individual magnetic immunocomplexes. Tau 441 , phosphorylated tau at Thr181, 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP, Aldrich) treated Aβ 40 and Aβ 42 monomer with a final concentration of 0, 5, 10, 50, 100, 250, 500 or 1000 fM was incubated with the nanoprobes and detection antibody at the optimal condition, followed by the labeling of the dye. 10 µL of the dye labeled immune-solution was injected into the flow cell followed by the magnetic separation.
Quantification of the target protein biomarkers in human serum. The standard addition curve and external calibration curve were established by correlating the average net intensity of fluorescent molecules in captured images at each concentration of spiked target analyte and target analyte respectively. HFIP treated Aβ 42 monomer with a final concentration of 0, 250, 500, 750, 1000, 1500, 2500, and 5000 fM was incubated with the 3x nanoprobes, serum, and detection antibody at the optimal condition, the excessive reagent was discarded after the immunoreaction by the magnetic separation. The nanoprobes were then redispersed in PBS and labeled with the dye. The dye labeled immune solution of 10 µL was injected into the flow cell.
Pretreatment of human saliva and urine sample. Both of the saliva and urine samples were pre-treated prior use. The saliva sample was centrifuged at 1500 rpm at 4 °C for 5 min to remove debris. The sample was then stored at -20 °C. For the urine sample, it was placed in ice bath for 30 min. The urine sample was centrifuged at 13,500 rpm at 4 °C for 15 min. The supernatant of the sample was treated with 150 μL 100% (w/v) trichloroacetic acid, and left on ice for 1 h. The sample was centrifuged again at 13,500 rpm at 4 °C for 15 min. The pellets were washed with acetone for three times. The pellets were left to air-dry and stored at -20°C. The pellets were redispersed in PBS prior use.
Quantification of the target protein biomarkers in human CSF, urine and saliva. The external calibration curve was established by correlating the average net intensity of fluorescent magnetic immunocomplexes in captured images at each concentration of target analyte. Three independent calibrations for Aβ 42 monomer, tau 441 and p-tau 181 were established. In short, the target antigen of concentration ranges from 0-5 fM (Ab 42 ) and 0-2.5 fM (tau 441 and p-tau 181 ) was incubated with the nanoprobesand detection antibody at the optimal condition and the magnetic immunocomplexes were labeled with the dye. The dye labeled immune solution of 10 µL was injected into the flow cell. The magnetic separation was performed prior the imaging by the TIRFM. The CSF sample was diluted with PBS before the immunoassay. The detection assay for CSF, saliva and urine samples were performed using the same method as the standards. The fluorescence images were captured by TIRFM with an excitation wavelength 488 nm.

Verification of the developed assay with INNOTEST ELISA kits. INNOTEST β-
Amyloid1-42, Phospho-tau(181P) and hTau Ag were purchased from Fujirebio (Belgium). The detection of Aβ 42 , t-tau and p-tau 181 was performed in duplicate following the manufacturer's assay protocol. For the quantification of Aβ 42 , 25 µL of Aβ 42 standards and CSF sample were added into the capture antibody coated micro-wells followed by 75 µL biotinylated detection antibody. The mixtures were then incubated at room temperature for 1 hr. The wells were emptied and washed with 1x wash solution for five times. Then, 100 µL peroxidase-labeled streptavidin was added to the wells and incubated at room temperature for 30 min. The wells were emptied and washed with 1x wash solution for five times. Next, 100 µL of the TMB substrate was added to the wells and incubated in the dark at room temperature for 30 min. Finally, 50 µL of stop solution was added into each well and the plate was shaken carefully for 1 min. The absorbance at 450 nm was recorded by Benchmark Plus Microplate Reader. The quantification of t-tau was done as follows: 25 µL of t-tau standards and CSF sample were added into the capture antibody coated micro-wells followed by 75 µL biotinylated detection antibody. The mixtures were then incubated at room temperature for 16 hr. The wells were emptied and washed with 1x wash solution for five times. Then, 100 µL peroxidase-labeled streptavidin was added to the wells and incubated at room temperature for 30 min. The wells were emptied and washed with 1x wash solution for five times. Next, 100 µL of the TMB substrate was added to the wells and incubated in the dark at room temperature for 30 min. Finally, 50 µL of stop solution was added into each well and the plate was shaken carefully for 1 min. The absorbance at 450 nm was recorded by Benchmark Plus Microplate Reader. The level of p-tau was determined by following approaches. In short, 75 µL of t-tau standards and CSF sample were added into the capture antibody coated micro-wells followed by 25 µL biotinylated detection antibody. The mixtures were then incubated at 4°C for 16 hr. The wells were emptied and washed with 1x wash solution for five times. Then, 100 µL peroxidase-labeled streptavidin was added to the wells and incubated at room temperature for 30 min. The wells were emptied and washed with 1x wash solution for five times. Next, 100 µL of the TMB substrate was added to the wells and incubated in the dark at room temperature for 30 min. Finally, 50 µL of stop solution was added into each well and the plate was shaken carefully for 1 min. The absorbance at 450 nm was recorded by Benchmark Plus Microplate Reader.

Quantification of target protein biomarkers with spectrofluorimeter.
To explore the possibility of the quantification of the biomarkers with commercial fluorimeter, the external calibration curve of Ab 42 was established by correlating the fluorescent intensity at the emission maximum of the fluorophores against different concentrations of target proteins. Different concentrations of Ab 42 (0-1000 fM) was incubated with the optimal amount of nanoprobes and detection antibody under the optimal condition in 10% glycerol-PBS solution at 37 °C for 1 h. The resultant immunocomplexes were labeled with 50 µM SIM and the fluorescence spectra of the immunocomplexes was recorded by the fluorescence spectrophotometer (PTI QM-4/2005). The quantification of Ab 42 in human CSF was performed using the same method as the standards and the fluorescent signal was measured by the spectrophotometer.

Multiplex detection of target protein biomarkers.
In order to demonstrate the multiplexity of the detection assay, 5 pM of the target protein, Ab42 and tau441, were incubated with their corresponding probes labeled with two different fluorescent dye, SLAce and SIM, and detection antibody. The solution mixture was then incubated with SIM and injected into the flow cell. The first-order images were then observed under the TIRFM-EMCCD imaging system with a transmission grating.    Figure S3. The influence of the magnetic nanoparticles on the fluorescence response of the SLAce and SIM upon binding to Ab 42 in PB buffer.