Fig 2: Validation of anti-GFAP and the detection of GFAP in seven patient plasma samples using ELISA (six patient samples and one healthy control). (A) Comparison between mean ± standard deviation (SD) of optical densities (OD) at 450 nm absorbance in assay buffer solution (pale orange) and control plasma solution (pale green) when spiked with known concentrations between 0.02 and 2.0 ng/mL of GFAP. All concentrations were measured in duplicate. ns = p > 0.05; * = 0.01 < p < 0.05. (B) Concentrations of GFAP detected in the healthy control sample (PS0) and six patient samples (PS1–PS6) measured using ELISA (left axis), and the assay LODs (right axis). All samples were measured in duplicate.

Fig 3: Multiplex of the IL-6 QLISA (fluorescent signal from QdotTM 525) and the GFAP QLISA (fluorescent signal from QdotTM 585) where the sample consisted of 25,000 pg/mL IL-6 and 10,000 pg/mL GFAP. (A) Representative images of ~4.5 µm beads trapped in the IL-6 detection band of the variable height device (left) and ~2.8 µm beads trapped in the GFAP detection band (right). Each image is an overlay of the QdotTM 525 and QdotTM 585 channels, and the brightness/contrast was set for ease of viewing. (B) The fluorescence intensity values from the QdotTM 525 (blue) and QdotTM 585 (yellow) channels for each variable height device detection band shown in (A) with the control values subtracted. The error bars are standard error of the mean and represent the variability in fluorescence intensity between the fields of view within each detection band.

Fig 4: Detection of GFAP in patient plasma using the GFET biosensor. (A) Calibration curve of the GFET biosensor for the GFAP detection in plasma. n = 3. (B) Signal intensity comparison between the tests in PBS and in the plasma for the same concentration order of magnitude illustrating excellent selectivity of the GFAP biosensor. (C) Real-time response of the GFAP biosensor for the detection of GFAP in plasma. Significant change seen in the curve for the sample of 0.56 pg/mL in comparison to the healthy control, suggesting that the sensor is able to respond to the sub-pg/mL (4 fM) level of GFAP in plasma. (D) Measurement results of six patient samples and one control plasma sample by the GFET. (E) Correlation of GFAP concentration measured by Simoa, ELISA, and GFET. GFAP concentration measured by GFET results showed significant correlation with those measured by Simoa and ELISA (p < 0.0001 and p < 0.001, respectively). The PS1 data are not fitted, as it is only available for Simoa. (F) Signal percentage of GFAP concentration measured by GFET and ELISA in comparison to Simoa as a reference. The GFAP concentrations in both PS0 and PS1 measured by ELISA and GFET are below their LODs.

Fig 5: Schematic of the methods for GFAP detection. (A) State-of-the-art Simoa technology relies on the effective binding between 500 K antibody-modified magnetic beads and the GFAP molecules at a low concentration. The GFAP concentration is determined by digital counting of the fluorescent signal from 216 K femtoliter-sized wells (for sample with high concentrations, there is also analogous signal quantification). (B) Classic sandwich ELISA uses an HRP-based colorimetric detection. The concentration is determined by the integration of TMB color changes. (C) On-chip GFET biosensing platform uses anti-GFAP functionalized graphene channel as a sensing element. The nonencapsulated reference electrode (orange) allows liquid gating without external electrode. Detection is based on the shift of Dirac point in response to the extent of antigen binding, which is linked to the GFAP concentration within a solution.