br Fig A Absorbance at nm of
Fig. 4. (A) Absorbance at 450 nm of PMGO and HRP in the presence H2O2 + TMB in PBS with diﬀerent pH levels (n = 3). (B) Comparison of the oxidization of PMGO and HRP with diﬀerent dilutions to H2O2 + TMB (n = 3). (C) Stability of HRP and PMGO stored at 25, 37, and 50 °C for a period of 1–90 days. The PMGO showed no decrease in oxidizability toward H2O2 + TMB after 90 days of storage at 37 °C (n = 3). (D) UV–vis spectra of PMGO after 7 days of storage at 70 °C. The results showed that the Benazepril peak at 680 nm attributed to PB disappeared (n = 3). (E) Analysis of the successful conjugation of diﬀerent antibodies (AbApoA1, mouse IgG, rabbit anti-mouse IgG, and goat anti-rabbit IgG) on the PMGO surface to form PMGO-1, PMGO-2, and PMGO-3 by SDS-PAGE Gel 12%. (F) Digital photos showing the color change with self-linkable PMGO for signal amplification.
3.5. Specificity of ApoA1 protein in spiked urine sample
During the interference study, we incubated a biochip with typical interfering species and other cancer biomarkers that exist in human plasma or urine, including urea, AA, VEGF, CA19-9, MUC1, FXYD3, BSA, and FBS. In our sensing system, the biochip could specifically capture the target protein (ApoA1) and prevent nonspecific adsorption of other species or self-linkable PMGO on the surface to induce un-wanted interference. No obvious A450 nm was observed when the in-terfering species were incubated with the biochip and then self-linkable PMGOs were added for signal induction. By contrast, a significant ab-sorbance intensity at 450 nm was observed when 20 ng/mL of ApoA1 was added for incubation with the biochip in the presence of self-linkable PMGOs (Fig. S3), indicating the PMGO-1 would only bind onto ApoA1 to induce self-chain reaction with PMGO-2 and PMGO-3 for signal amplification. To prove the accuracy of our immunosensing system for ApoA1 detection, we spiked various concentrations of ApoA1 into human urine, ranging from 20 ng/mL to 100 ng/mL, and the A450 nm was recorded; the results are presented in Table S1. The recovery rates of ApoA1 were found to be acceptable and in the range of 98.1–104.7%, and the relative standard deviation was lower than 3%, indicating that our immunosensing system is accurate enough for ApoA1 detection. As a result, we confirmed that the immunosensing system with self-linkable PMGO for signal amplification in ApoA1 de-tection had high sensitivity, selectivity, and accuracy.
3.6. Real urine sample test from BC patients
ApoA1 has been identified to be a potential biomarker for early diagnosis and clinical classification of BC with a sensitivity and speci-ficity of 89.2% and 84.6%, respectively (Li et al., 2014). The urine samples of aggressive BC showed significant increases in ApoA1 ex-pression compared with low malignant BC (Li et al., 2011). A total of sixteen urine samples were collected from the patients diagnosed with
BC (ten urine samples from BC patients were collected before surgical operation and chemotherapy; six urine samples from BC patients were collected after surgical operation and chemotherapy) and four healthy people to assess the utility of our immunosensor; the results were also compared with those obtained using a conventional detection method (i.e., an ELISA), and no significant diﬀerence was found between them (Fig. 6A). The results indicated that this immunosensor with self-link-able PMGOs for a signal amplification approach potentially oﬀers a more accurate and rapid alternative (within 60 min) than current tests that are based on immunological methods, such as an ELISA, im-munohistochemistry, or two-dimensional electrophoresis (2-DE) cou-pled with mass spectrometry.
The concentrations of ApoA1 in the urine of four healthy people using our immunosensor were 5.1 ± 1.2, 2.2 ± 0.5, 6.5 ± 1.1, and 4.8 ± 0.3 ng/mL, respectively, which were all lower than the cut-oﬀ value (11.16 ng/mL). By contrast, the concentrations of ApoA1 were detected using the immunosensor as being between 55.6 ± 2.3 and 104.7 ± 7.4 ng/mL for the patients who were diagnosed with high-grade BC, and between 19.3 ± 4.3 and 22.3 ± 4.9 ng/mL for the pa-tients who were diagnosed with low-grade BC. In addition, we found a notable trend; the ApoA1 concentrations (1.4 ± 0.8 ~ 10.3 ± 1.6 ng/ mL) were significantly decreased to normal level in the urine of the patients who were diagnosed with high-grade BC then received surgery and chemotherapy, and no recurrence was observed (Fig. 6B). The re-port of a pathological section also showed the results were negative for malignancy. Taken together, our colorimetric immunosensor with self-linkable PMGOs for signal amplification can be a potential tool to ra-pidly and accurately detect ApoA1 concentrations in urine for the early diagnosis, classification of tumor grade, and prognosis monitoring in BC.