Rapid, quantitative, and high-sensitivity detection of anti-phospholipase A2 receptor antibodies using a novel CdSe/ZnS-based fluorescence immunosorbent assay

Autoantibodies against M-type phospholipase A2 receptor (PLA2R) serve as specific biomarkers for idiopathic membranous nephropathy (IMN), and its quantification helps monitor disease activity. In this study, we describe a rapid and highly sensitive quantum dots-based immunochromatography assay (QD-ICA) for quantifying PLA2R autoantibodies. Serum samples from 135 biopsy-confirmed patients with nephrotic syndrome were analyzed for PLA2R autoantibodies using the novel QD-ICA as well as commercialized enzyme-linked immunosorbent assay (ELISA). Areas under the receiver operating characteristic curve (AUC-ROC) of QD-ICA were significantly greater than those of ELISA (91.1% [95% CI 85.9–96.3%] and 83.9% [95% CI 76.5–91.2%] respectively; p < 0.01). The detection sensitivity and specificity of QD-ICA (80.9% [95% CI 69.2–89.0%] and 100% [95% CI 93.2–100.0%], respectively) exceeded those of ELISA (72.1% [95% CI 59.7–81.9%] and 98.5% [95% CI 90.9–100.0%], respectively). The optimum cut-off value of QD-ICA was 18.18 relative units (RU)/mL, and the limit of detection was 2.86 RU/mL. The novel QD-ICA outperforms ELISA in detecting PLA2R autoantibodies, with shorter detection time, fewer steps, smaller equipment size, and broader testing application, suggesting its capability to improve IMN diagnosis and monitor patient response to treatment.


Results
Clinical characteristics of the patients. A total of 135 biopsy-confirmed patients with nephrotic syndrome (including 68 patients with IMN and 67 patients without IMN) who visited the Department of Nephrology, Shanxi Provincial People's Hospital (Shanxi, China) were enrolled. The median age of all patients was 48 (interquartile range  years. The ratio of females to males was 1:1.2. There were significant differences in serum total protein levels, serum albumin levels, 24-h proteinuria, Cystatin C content, and estimated glomerular filtration rate (eGFR) between the IMN and non-IMN groups (p < 0.05), but there was no significant difference in serum creatinine and urea concentrations ( Table 1).

Characterization of QD-mAb probes.
We designed hydrophilic core-shell CdSe/ZnS QDs bound to mAb. Supplementary Fig. 1a shows the absorption (356 nm) and photoluminescence spectra (615 nm) of waterinsoluble QDs. To change the hydrophobic properties of CdSe/ZnS QDs and enable their transfer to aqueous solution, QDs were encapsulated with amphiphilic polystyrene particles. TEM images of the QDs encapsulated with amphiphilic polystyrene particles are shown in Supplementary Fig. 1b. CdSe/ZnS QDs were evenly bound to the surface and inside of polystyrene particles. Prepared QDs exhibited a narrow particle-size distribution and highly homogeneous monodispersity, with a mean particle size of ~ 305 nm. We identified the effect of different coupling ratios between hydrophilic QDs and mAb on test performance. Specifically, a QDs:mAb coupling ratio of 1:0.2 had the strongest T/C fluorescence; therefore, we selected this ratio for coupling hydrophilic QDs to antibodies ( Supplementary Fig. 1c).
After optimizing the coupling conditions, we determined the sizes and fluorescence spectra of the QDantibody and hydrophilic CdSe/ZnS QDs ( Supplementary Fig. 1d,e). The fluorescence peak of the QDs-mAb solution did not differ significantly from that of hydrophilic QDs with respect to shape and position, although the fluorescence intensity declined as uncoupled components remained in the supernatant during this process (Supplementary Fig. 1d). After antibody labeling, the hydrodynamic analysis revealed that QDs-mAb size increased from 305 to 345 nm ( Supplementary Fig. 1e), indicating the successful formation of the QDs-mAb conjugate. The zeta potential of the QD-ICA was then detected to determine conjugate stability. Generally, a zeta potential > + 30 mV or < − 30 mV indicates stability in solution 22 . We determined a zeta potential of − 55.2 mV for the conjugate, suggesting that its carboxyl group can provide sufficient colloidal stability in an aqueous solution ( Supplementary Fig. 1f).
Optimization of the QDs-mAb probes. Next, we analyzed the stability and optical properties of the QDs-mAb probes under physiological conditions. We found that the QDs-mAb remained stable at a pH range of 6.0-9.0, with an optimal pH of 7.5 ( Supplementary Fig. 2a). The analysis of PL intensity showed that the fluorescence intensity of the QDs-mAb probe was the strongest in the MES buffer ( Supplementary Fig. 2b). Therefore, we evaluated the effect of PL intensity in the presence of MES buffer (pH 7.5) under conditions of six different ionic strengths and found that the QDs-mAb in MES buffer (0.005 M, pH 7.5) exhibited the highest fluorescence intensity; thus, this was selected as the diluent for the subsequent analyses ( Supplementary Fig. 2c).
Optimal proportions of coated PLA2R and QDs-mAb. We performed a checkerboard titration test to determine the optimal dilution ratios of the QDs-mAb probe and coated-antigen concentrations. The measured fluorescence intensity of varying concentrations of PLA2R and QDs-mAb demonstrated that the conditions were optimum at a QDs-mAb dilution of 1:100 and a coated antigen concentration of 0.5 mg/mL ( Table 2).

Determination of incubation time.
To determine the optimal time required for fluorescence development, we measured fluorescence intensities at different times. The results showed that the optimal fluorescencedevelopment time to establish dynamic equilibrium between the coated antigen and the QDs-mAb was 15 min (Fig. 2). As the QDs-mAb tends to be released from the immune complex, the maximum fluorescence intensity at 60 min began to decrease. The performance of QD-ICA. Limit of blank (LoB) was used to determine the highest concentration of anti-PLA2R antibodies detected in the blank samples, and limit of detection (LoD) was used to determine the lowest concentration that the method could reliably detect to determine the presence or absence of anti-PLA2R antibodies. The LoB (the 95th percentile of the detected data) was 1.94 RU/mL, and the LoD value (LoD = LoB + 1.645 SD) was 2.86 RU/mL, with a broader linear range of the sample between 3 and 1500 RU/mL. The precision and matrix effect of QD-ICA were investigated by recovery experiment. The inter-batch recovery rate was 93.3-105.4%, and the intra-batch recovery rate was 96.2-102.7% (Table 3). The evaluation of serum matrix effects showed that the recovery rate of three samples was 96.30-104.63%, indicating that no significant matrix effect was observed with this method (Table 4).
QD-ICA immunoassays targeting the anti-PLA2R antibody. ELISA (EUROIMMUN) was used to compare the QD-ICA results, generate scatter plot, and establish statistical regression equations. Human serum was diluted (1:100) with sample buffer according to Dähnrich et al. 16 . All serum samples were tested in triplicates, and the average values were used for further analysis. The Passing-Bablok regression equation was Y = 0.91x − 0.38, and Spearman's correlation coefficient r is 0.8 (Fig. 3a). The Bland-Altman plot (Fig. 3b) showed that the two assays are in good agreement with no obvious bias. The anti-PLA2R antibody sensitivity and specificity detected using QD-ICA    (Fig. 4). ROC analysis indicated that the optimal cut-off point for the QD-ICA was 18.18 RU/ mL. A comparison of the analytical performance of different methods for detecting the anti-PLA2R antibody is shown in Table 5. Conventional western blotting, IIFA assays, and commercial CBA-IFA assay are all semiquantitative analyses of anti-PLA2R antibodies. Although the sensitivities of these methods were comparable to those of quantitative detection methods, their complicated and time-consuming nature would hinder their application in clinical laboratories. Moreover, these methods were less reliable as the results depended on the observer's interpretation. ELISA, ALBIA, TRFIA, and LIPS quantitative detection methods improved detection speed, but the detection times remained longer than 2 h, making them unsuitable for use as POCT tests. Additionally, ALBIA and LIPS assays were less sensitive than the other methods listed in Table 5. The ChLIA method had similar sensitivity and specificity as the developed method here, but the detection system relied on expensive large-scale equipment, which was not applicable in community service points or areas with insufficient medical resources. These results showed that compared to other methods used to detect the anti-PLA2R antibody, QD-ICA was both rapid and highly sensitive.

Discussion
Detection of the anti-PLA2R antibody significantly improved early diagnosis and appropriate clinical treatment of IMN. Here, we developed a novel POCT test (QD-ICA) for the quantitative detection of PLA2R autoantibodies and compared its performance with commercially available ELISA. Compared to other assays targeting anti-PLA2R antibody detection, QD-ICA was highly sensitive and faster, with an observed clinical sensitivity Table 4. Evaluation of the serum matrix effects. Positive sample concentration is defined as the concentration of samples that were positive for PLA2R antibodies, as detected by ELISA (EUROIMMUN). The detection concentration of negative serum was 3.4 RU/mL. Recovery concentration (RU/mL) = detection concentration of samples-detection concentration of negative serum. Added positive sample concentration (RU/ mL) = positive samples concentration × added sample amount. Recovery rate (%) = recovery concentration / addition concentration × 100%.    14 . Increasing levels of PLA2R autoantibodies indicate the exacerbation of symptoms, whereas decreased levels correlate with disease remission 23 . Therefore, it is necessary to detect autoantibodies quantitatively. Previous studies used western blotting, which requires electrophoresis and immunoblotting, as well as the use of naturally purified glomerular glycoprotein or extracts from cells overexpressing PLA2R, which may not be available in conventional clinical laboratories. Similarly, immunofluorescence testing methods are semiquantitative and observer-dependent, making them unsuitable for clinical testing.
The EUROIMMUN ELISA is currently widely used in clinical laboratories. The cut-off value as per the manufacturer instructions is recommended at 20 RU/mL, and values exceeding this are considered positive and those < 14 RU/mL are considered negative. However, a value between 14 and 20 RU/mL represents an uncertain outcome. Moreover, 20 RU/mL results in specificity ranging from 89.7 to 100% 18,25,[32][33][34][35][36] . Some studies used different customized thresholds (e.g., 2, 2.6, or 14 RU/mL) to increase sensitivity 18,33,34,36 , although these sometimes adversely affected specificity. In the present study, ROC analysis revealed an optimal cut-off value of 17.84 RU/ mL with ELISA, which resulted in maximum sensitivity and specificity of 72.1 and 98.5%, respectively, which was still lower than the results obtained using the QC-ICA. Bobart et al. found that cases tested using ELISA with results ranging from 2 to 20 RU/mL and positive via IFA were confirmed as MN via biopsy. However, 57.5% of cases tested using ELISA with results ranging from 2 to 20 RU/mL and with negative values via IFA were confirmed as MN via biopsy 37 . Therefore, Bobart et al. suggested that samples with ELISA results ≥ 2 RU/mL and ≤ 20 RU/mL should be confirmed via IFA before performing MN-related biopsies to improve the diagnostic sensitivity. Among our 68 patients with IMN, confirmed using renal biopsy, samples that tested positive for anti-PLA2R antibodies using ELISA were also positive in the QD-ICA test. However, anti-PLA2R antibodies from six patients with IMN were not detected with ELISA, suggesting that QD-ICA was more sensitive than ELISA and did not require a confirmation test.
QD-ICA is a subset of lateral flow assay-based POCT devices category and is performed over a strip that consists of a sample pad, marker-binding pad, absorption pad, and NC membrane. Compared with ELISA, this method is easy to operate and has low operating costs 38 . The advantages offered by QD-ICA, including simplicity, user-friendliness, cost-effectiveness, the ability to quantify the results, and high reagent stability, would be a breakthrough for IMN diagnosis. QD-ICA exhibited improved sensitivity due to its reliance upon labeling using aqueous QDs rather than macromolecular enzymes, thereby allowing more efficient labeling. Additionally, the use of aqueous QDs increases the Stokes shift between the excitation and emission wavelengths, thus, effectively circumnavigating environmental interference arising from a relatively long fluorescence-emission period. There were some limitations to this study. All patients enrolled in the study were from a single center, and a limited number of cases were included in this study. In addition, only the performance of QD-ICA was compared with that of ELISA for the detection of PLA2R autoantibodies. To further determine the potential advantages and disadvantages of QD-ICA, additional detection studies are needed for comparative analysis.
In summary, this study constructed a rapid, quantitative, and highly sensitive anti-PLA2R antibody detection assay using a CdSe/ZnS-based immunosorbent assay. The novel QD-ICA has the advantages of faster detection time, fewer steps, wider testing scenarios, and greater overall feasibility for detecting PLA2R autoantibodies compared to ELISA. These findings suggest that QD-ICA can improve IMN diagnosis and monitoring of patient   www.nature.com/scientificreports/ China). Fluorescence spectra were measured using a fluorescence spectrofluorometer (Thermo Fisher Scientific, Waltham, MA, USA). Real-time dynamic light scattering signals of the QDs and the QDs-antibody probe were detected using a Zetasizer Nano ZS system (Malvern Panalytical, Malvern, UK). Selected mAbs and QD probes were dispensed on the sample pad using a gold-dispensing system (Jinbiao, Shanghai, China). Photoluminescence (PL) spectra of the control (C) line and test (T) line on the QD-ICA were recorded using the automatic fluorescence immunoanalyzer QD-S600 (Vazyme). All other reagents were purchased from Vazyme.

Study participants. A total of 135 biopsy-confirmed nephrotic syndrome patients [IMN (68), SMN (9),
and non-MN (58) with nephrotic syndrome (IgA nephritis (21), diabetic nephropathy (8), lupus nephropathy (7), focal segmental glomerular sclerosis (9), and immunocomplex-associated glomerulonephritis (13)] were enrolled from Shanxi Provincial People's Hospital. The patients were diagnosed based on immunohistochemistry, light microscopy, and electron microscopy findings. Serum and urine samples were collected the day before renal biopsy (before initiating therapy). Serum samples from all patients were centrifuged for subsequent use. Before the interview and blood collection, written informed consent from adults and the consent of the parents of minors was obtained. All methods were performed in accordance with relevant guidelines and regulations. The study was approved by the Ethics Committee of Shanxi Provincial People's Hospital. Participants were informed of the purpose of the study and their right to keep their information confidential.

Establishment of QD-antibody conjugates.
Hydrophobic core-shell CdSe/ZnS QDs were designed as reported previously, with appropriate modifications 39 (" Supplementary Information"). QDs-mAb conjugates were prepared via a conventional EDC/NHS coupling reaction between the amino groups of the antibodies and carboxyl groups on the hydrophilic surface of QDs (Fig. 1). First, 50 μL of QDs (10 mg/mL) was dispersed in 450 μL of MOPS (0.02 M, PH 6.5), followed by activation with 3 μL (50 mg/mL) EDC and 3 μL (75 mg/mL) NHS. The activated QDs on ice were subjected to ultrasound for 5 min and collected by centrifugation at 17,000 × g at 18 °C for 15 min. The QDs sediment was dispersed in 500 µL MES (0.02 M, pH 6.0) and incubated with different amounts of mouse anti-human IgG mAb for approximately 30 min at room temperature (20 ± 5 °C). Next, we added 5 µL (10 mg/mL) of the reference protein DNP-BSA for quality control and incubated for 45 min at room temperature. The complex was blocked using 0.5% casein and terminated using 10% ethanolamine for 30 min at the same temperature. The terminated sediment was obtained by centrifugation and stored in 50 µL Tris buffer (5 mM, pH 7.5).

Preparation of recombinant human PLA2R in HEK293 T cells.
The extracellular domain of human PLA2R (NP_031392.3) was expressed in HEK293 T cells as described previously 11 . After culturing for ~ 1 week, a stable strain was collected for Ni 2+ -affinity chromatography purification. Recombinant human PLA2R was identified by performing nonreducing western blotting using serum positive for antibodies against PLA2R, as identified using ELISA (EA 1254-9601 G, EUROIMMUN AG, Lübeck, Germany). SDS-PAGE and western blotting were performed to detect the purity and confirm the molecular weight of the protein.

Establishment of optimal proportions of coated PLA2R and QD-labeled antibody.
A checkerboard titration test was performed to determine the optimal dilution ratios for the QDs-mAb probe and coated antigen concentration. The coated PLA2R antigen was diluted to different concentrations (0.5, 1, and 2 mg/mL) and fixed on a microplate in a volume of 100 μL, after which different concentrations of standard serum (2, 20, 100, and 500 RU/mL) were added to the corresponding wells and incubated at 37 °C for 30 min. The QDs-mAb probes were then diluted at different ratios [1:50, 1:100, and 1:200 (v/v)], followed by incubation at 37 °C for 30 min. Optical density was measured for each concentration.
Design of the QD-based immunochromatography assay. The  www.nature.com/scientificreports/ to determine the PLA2R autoantibody concentration in the serum samples and select a sample (1556.84 RU/ mL) with concentration in the upper limit of the linear range of the kit (1500 RU/mL). Several serum samples with low anti-PLA2R levels were selected and mixed to prepare a "negative sample" (< 4 RU/mL). The samples were diluted to different concentrations with the negative serum sample (< 4 RU/mL) as a calibrator to develop a standard curve. The level of the PLA2R autoantibody in the serum was calculated using the established standard curve.
The accuracy and precision of QD-ICA were evaluated as follows. Serum was added at different concentrations to the negative serum samples (< 4 RU/mL) to prepare a mixture with final concentrations of 20, 40, and 80 RU/mL. The spiked concentration of each PLA2R antibody was tested six times on the same day, and the test was repeated every 3 days in three batches. The average spike recovery rate of each spike concentration and the intra-batch and inter-batch differences of test strips were calculated. The serum matrix effect was evaluated as follows. Three samples with high anti-PLA2R levels were selected and defined sample A (102.5 RU/mL), sample B (495.8 RU/mL), and sample C (1011.2 RU/mL). The positive serum and 60 µL of distilled water were added to 240 µL of the negative serum sample (< 4 RU/mL), and the recovery concentration and recovery rate of different samples were calculated. LoB, LoD 40 , and the linear range were determined as described in the "Supplementary Information". Statistical analysis. Statistical analysis was performed using SPSS (v.25; IBM Corp., Armonk, NY, USA).
Continuous variables were represented as means (standard deviation) or median (interquartile ranges), and categorical variables were described as frequency rates or percentages. Mann-Whitney test was used to analyze continuous variables, and the chi-square and Fisher's exact tests were used to analyze categorical variables. The Passing-Bablok regression equation and coefficient of determination (R 2 ) were calculated. The Bland-Altman plot of log-transformed difference was also used to describe the agreement between the two methods. The receiver operating characteristic (ROC) curve, area under ROC curve (AUC), and cut-off values of the QD-ICA were determined using clinical samples. A two-sided α of less than 0.05 was considered statistically significant.

Data availability
The datasets analyzed in this study are available from the corresponding author on reasonable request.