A Quantification of Target Protein Biomarkers in Complex Media by Faradaic Shotgun Tagging

The progressive emergence of protein biomarkers promises a revolution in the healthcare industry and a shift of focus from disease management to much earlier intervention. Here, we introduce a facile shotgun tagging of ensemble proteins in clinically relevant media prior to specific target capture at antibody-modified electrodes. This facilitates a convenient voltammetric quantification of markers down to sub-pg/mL levels and across several orders of concentration. A translation of the methodology to an automated microfluidic platform enables marker quantification from 25 μL of sample in less than 15 min, demonstrated here with a simultaneous assaying of CRP and cardiac troponin I (cTnI). The assays show a good correlation with a standard immunoassay when applied to real patient serum samples. The platform is simple, generic, highly sensitive and requires no secondary labeling/binding or amplification.


pH and concentration optimization
To assess the effect of pH on the efficiency of coupling between the MB-NHS and proteins, the assay performance was tested using different pH buffers for the protein solution. These were pH 5.0, 6.0, 7.0 and 8.0 in acetate, MES and phosphate (pH 8.0, pH 9.0) buffers respectively. The results shown in figure S1 demonstrate the change in response as function of CRP concentration at studied buffers. The coupling in 100 mM MES buffer at pH 6.0 showed highest signal to noise ratio and a larger dynamic range. This was accordingly applied thereafter as the dilution medium.
Two different concentrations of methylene blue NHS were tested to investigate the effect on its concentration on the electrochemical behaviour of the assay. At high MB-NHS concentration (100 µg/mL) the detection limit was 500 pg/ml largely due to the high background signal. Reducing MB-NHS conc. to 10 µg/mL resulted in a much lower background decreasing the assay LOD down to 1 pg/mL. This ensured a low assay S/N ratio at the optimized experimental and electrochemical parameters reducing the background relevant to the amount of non-specifically adsorbed proteins. Figure S1: Optimization of pH buffer labelling conditions. Columns show average of response over two independent measurements with increasing CRP concentration for protein solutions buffered at different pH values. Buffers used were 100 mM Acetate buffer at pH 5.0, 100mM MES buffer at pH 6.0 and 100mM Potassium phosphate buffer at pH 8.0 and pH 9.0.  for free and MB-labelled CRP showed no significant difference with that of free CRP at 2.8 X 10 -10 M compared to 9.5 X 10 -10 M for MB-labelled CRP. Figure S3: SPR affinity measurements of native CRP (left) and MB-tagged CRP (right) measured on a gold surface with physisorbed CRP antibodies. Black lines depict recorded data, red lines show associated fits obtained from BicaoreX100 Evaluation Software.

S5
Assay specificity study Figure S4. Specificity study for physisorbed anti-CRP interface on GCE electrode and SPR gold chip. Left y-axis represent change in peak current and right y-axis represent change in SPR response units (RU) upon exposing the interface to 30 ng/mL CRP, 2mg/mL human serum albumin, 2 mg/mL bovine serum albumin, 2 mg/mL fibrinogen, 100 ng/mL cTnI, and 1% human serum. All measurements were in MES buffer. The concentrations used represent estimates for protein content of 10 to 100 times diluted human serum.
Recorded electrochemical faradic responses were within background noise indicating minimal non-specific binding while confirming that faradaic signals recorded on Ab-modified electrodes are induced by the specific recruitment of target species. Electrochemical response of BSA modified electrodes towards 100 µg/mL BSA, 100 µg/mL HSA, 10 ng/mL CRP, 100 ng/mL CRP, 10 ng/mL cTnI, and 100 ng/mL cTnI as compared to specific response of anti-CRP modified electrode to 30 ng/mL CRP and anti-cTnI modified electrode against 4.8 ng/mL cTnI. Raw unprocessed voltammograms of BSA electrodes against (B) 100 µg/mL BSA and 100 µg/mL HSA; (C) 10 ng/mL CRP, 100 ng/mL CRP; and (D) 10 ng/mL cTnI, and 100 ng/mL cTnI.

Microfluidic design for single protein quantitation
A microfluidic Y-shaped serpentine mixer was designed and investigated to improve sample/reagent mixing while housing a closed sample and reagent chambers that can be directly driven by a syringe pump. This setup allowed the whole assay to be run from sampleto-answer within 15 minutes without the need for separate mixing or washing procedures. In the proposed design, once reagent and sample are loaded, loading holes were blocked with adhesive tape promising an easy sample loading for a closed microfluidic system.

S9
Online assay specificity study Figure S8. Electrochemical Specificity study for (A) Anti-CRP modified electrodes and (B) Anti-cTnI modified electrode in online microfluidic multiplexed protein analyses. The response to studied interfering species was between 5-15% of the target-specific signal indicating a good specificity and low cross-reactivity towards other proteins. Error bars represent standard deviation from two independent measurements.

S10
Representative DPV peaks from on-line multiplexed protein analysis Figure S9: Representative DPV peaks for multiplexed on-line assay for CRP and cTnI as function of (A) increasing CRP connection on Anti-CRP decorated GCE and (B) cTnI increasing concentration on anti-cTnI decorated GCE.

Assay validation
The assay was validated by calculating the percent recovery of CRP spiked in 100% human serum. The electrochemical signal from un-spiked 100% human serum was measured and its CRP contents were estimated. Tested human serum was estimated to have 2.35 ng/mL of CRP; well within its normal expression range. Subsequently, two different concentrations of CRP were spiked and the recovered concentrations were calculated using the calibration data obtained in 1 % human serum (Figure 1). Spiked human serum samples showed percent recoveries between 80-120% of the spiked concentrations (