Early Detection of SARS-CoV-2 Seroconversion in Humans with Aggregation-Induced Near-Infrared Emission Nanoparticle-Labeled Lateral Flow Immunoassay.

An outbreak of coronavirus disease (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) poses great threats to human health and the international economy. To reduce large-scale infection and transmission risk of SARS-CoV-2, a simple, rapid, and sensitive serological diagnostic method is urgently needed. Herein, an aggregation-induced emission (AIE) nanoparticle (AIE810NP, λem = 810 nm)-labeled lateral flow immunoassay was designed for early detection of immunoglobulin M (IgM) and immunoglobulin G (IgG) against SARS-CoV-2 in clinical serum samples. Using a near-infrared (NIR) AIE nanoparticle as the fluorescent reporter (△λ = 145 nm), the autofluorescence from the nitrocellulose membrane and biosample and the excitation background noise were effectively eliminated. After optimization, the limit of detection of IgM and IgG is 0.236 and 0.125 μg mL-1, respectively, commensurate with that of the enzyme-linked immunosorbent assay (ELISA) (0.040 and 0.039 μg mL-1). The sensitivity of the proposed AIE810NP-based test strip for detecting IgM and IgG is 78 and 95% (172 serum samples), commensurate with that of ELISA (85 and 95%) and better than that of a commercial colloidal gold nanoparticle (AuNP)-based test strip (41 and 85%). Importantly, the time of detecting IgM or IgG with an AIE810NP-based test strip in sequential clinical samples is 1-7 days after symptom onset, which is significantly earlier than that with a AuNP-based test strip (8-15 days). Therefore, the NIR-emissive AIE nanoparticle-labeled lateral flow immunoassay holds great potential for early detection of IgM and IgG in a seroconversion window period.

An outbreak of coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has rapidly spread to produce a global pandemic.[1−5] The current standard method for detecting the SARS-CoV-2 RNA genome is based on reverse transcription polymerase chain reaction (RT-PCR).[6−10] However, the expensive equipment, highly skilled analysts, and time-consuming property limit its wide application.[11,12] Testing of specific immunoglobulin M (IgM) and immunoglobulin G (IgG) against SARS-CoV-2 in human serum is an alternative way to diagnose COVID-19.[13,14] Several serological immunoassays, such as lateral flow immunoassay,[15−18] enzyme-linked immunosorbent assay (ELISA),[19] and automated chemiluminescent immunoassay,[12,20] have been developed for the detection of IgM and IgG.

Because of the advantages of portability, rapidity, simplicity, and low cost, the traditional colloidal gold nanoparticle (AuNP)-based lateral flow immunoassay is widely used in point-of-care testing (POCT) of COVID-19.[17] For example, Brochot et al. developed a AuNP-based test strip that has a positive rate of 60–80% on day 10 and 100% on day 15 for the detection of IgG.[21] To improve the sensitivity of the rapid test strip, Wang et al. developed a surface-enhanced Raman scattering (SERS)-based lateral flow test strip, and the detection limit of IgM and IgG is 800 times lower than that of the AuNP-based test strip.[22] To further satisfy the need for the detection of IgM and IgG under challenging circumstances, such as community and port entry/exit, and to efficiently prevent the SARS-CoV-2 transmission, the lateral flow immunoassay must be sensitive and easy to operate. Fluorescence lateral flow detection platform has been recognized as an important POCT detection technology due to its advantages of high sensitivity and portable instrumentation. Therefore, a simple, rapid, and sensitive fluorescence lateral flow test strip is urgently needed to early detect IgM and IgG against SARS-CoV-2 in human serum.

In this study, we demonstrated a near-infrared (NIR) emissive lateral flow immunoassay with an aggregation-induced emission (AIE) dye-loaded nanoparticle as reported that could detect IgM and IgG against SARS-CoV-2 in 1–7 days after symptom onset. To avoid the interference of autofluorescence in a nitrocellulose (NC) membrane and biosample,[23−25] an AIE dye with NIR emission, namely, BPBT, was chosen as the fluorescent unit. To further amplify the fluorescent labeling signal of a detection ligand, a polystyrene (PS) nanoparticle with a size of 300 nm loaded with 3.18 × 106 dyes (AIE810NP) was developed to label the detection ligand (Scheme a). Additionally, a portable reader was developed to quantitatively read out the NIR fluorescence signal (Scheme c). Using AIE810NP-labeled SARS-CoV-2 antigen (AIE810NP-SARS-CoV-2 antigen) as the fluorescent probe (Scheme b), the test strip achieved a diagnostic sensitivity of 78 and 95% for IgM and IgG, respectively, superior to that of the commercial AuNP-based test strip (41 and 85%). More importantly, the AIE810NP-based test strip can detect IgM or IgG at 1–7 days after symptoms onset, earlier than that of the AuNP-based test strip (8–15 days). Overall, the developed AIE810NP-based test strip holds great promise for early detection of IgM and IgG against SARS-CoV-2 in clinical serum samples.

Scheme 1

Schematic Illustration of the NIR-Emissive AIE Nanoparticle-Labeled Lateral Flow Immunoassay for Detection of IgM and IgG

Conditions: (a) Synthesis of AIE810NP and conjugation path of SARS-CoV-2 antigen with AIE810NP and chicken IgY with AIE810NP. (b) Schematic of the developed test strip for the detection of IgM and IgG against SARS-CoV-2 in a human serum sample. (c) Schematic of the portable reader, including an LED lamp excited at 680 nm, a complementary metal oxide semiconductor (CMOS) camera, and a set of optics. (d) Interpretation of different test results. The fluorescent bands on the M line, G line, and C line represent IgM/IgG positive; the fluorescent bands on the M line and C line represent IgM positive; the fluorescent bands on the G line and C line represent IgG positive; the fluorescent band on the C line represents IgM/IgG negative; the absence of fluorescent bands on the M line, G line, and C line represent an invalid test strip.

Results and Discussion

Principle of the NIR-Emissive AIE Nanoparticle-Labeled Lateral Flow Immunoassay

On basis of the immunoreaction between IgM/IgG and the AIE810NP-SARS-CoV-2 antigen, the combined IgM–IgG lateral flow test strip is designed for the detection of IgM and IgG in a clinical serum sample (Scheme b). A portable reader was built to collect the NIR fluorescence signal from three lines, which comprises an LED lamp excited at 680 nm, a CMOS camera, and a set of optical elements (Scheme c). For human serum sample detection, the IgM and IgG are captured by an AIE810NP-SARS-CoV-2 antigen and then captured by the mouse anti-human IgM immobilized on the M line and mouse anti-human IgG immobilized on the G line (forming a sandwich immunocomplex), respectively. The AIE810NP-labeled chicken IgY (AIE810NP-chicken IgY) is specifically bound to goat anti-chicken IgY immobilized on the C line as a quality control signal. The test results for the detection of IgM and IgG using the AIE810NP-based test strip are displayed in Scheme d. In the presence of IgM and IgG, fluorescent bands appear on all of the M lines, G lines, and C lines under 680 nm excitation of an LED lamp (IgM/IgG positive); in the presence of IgM, fluorescent bands appear on both of the M line and C line (IgM positive); in the presence of IgG, fluorescent bands appear on both of the G line and C line (IgG positive). However, in the absence of IgM and IgG, no fluorescent bands on the M and G lines (IgM/IgG negative) could be observed. Notably, if the fluorescent band of the C line is not observed, the strip is invalid.

Synthesis and Characterization of the NIR-Emissive AIE Nanoparticle

In this test strip, a NIR-emissive AIE molecule (BPBT) was selected as a fluorescent unit (Figures S1–S4, Supporting Information).[26,27] The AIE properties of BPBT were checked by monitoring its fluorescence intensity (FL intensity) in THF/water mixtures with variable water volume fractions (fw). A significant increase in the fluorescence intensities of BPBT is observed upon the gradual enhancement of fw from 50 to 90% (water/THF, v/v), and gradual blue shifts are observed for the emission maxima (from 890 to 837 nm), owing to their twisted intramolecular charge transfer (TICT) state (Figure S5, Supporting Information).[28] The absorption spectra of BPBT with diverse fw were also investigated during the AIE titration experiment. The absorption maxima decrease gradually and shift from 633 to 665 nm (Figure S6, Supporting Information), owing to the formation of a different aggregated state.[29]

For in vitro diagnosis, BPBT molecules were encapsulated into PS nanoparticles using an organic solvent swelling method (Scheme a).[30] Transmission electron microscopy (TEM) images in Figure a,b show the morphologies of PS nanoparticles and AIE810NP, respectively. The physical photos of the PS nanoparticles and AIE810NP are shown in Figure a,b inset, which indicates that the BPBT are successfully embedded in PS nanoparticles. The hydrodynamic diameter of AIE810NP was similar to that of PS nanoparticles (Figure S7, Supporting Information), demonstrating no aggregation or rupture of the PS nanoparticle carrier after dye loading. The UV–vis absorption spectrum and fluorescence spectrum of the synthesized AIE810NPs were further characterized. Figure c shows that the absorption maximum of the AIE810NP is 665 nm. Moreover, by establishing the UV absorption standard curve of BPBT, the number of BPBT in an AIE810NP was calculated to be 3.18 × 106 (Figure S8, Supporting Information). Figure d shows that the emission maximum of the synthesized AIE810NP is 810 nm. At the same time, there was no fluorescence in PS nanoparticles with the same concentration (Figure S9, Supporting Information), which indicated that BPBT was successfully embedded in the PS nanoparticle. In addition, the emission maximum of AIE810NPs is blue-shifted relative to BPBT (from 890 to 810 nm), due to their TICT state.[28]

Figure 1

Characterization of AIE810NP. (a) TEM image of PS nanoparticles. Inset shows the physical photo of PS nanoparticles. (b) TEM image of AIE810NP. Inset shows the physical photo of AIE810NP. (c) Absorption spectrum of AIE810NP in water. The concentrations of all nanoparticles are 180 μg mL–1. PS nanoparticles have the same concentrations as previous works. (d) Fluorescence spectrum of AIE810NP in water (700 nm long pass filter), λex = 630 nm (25.5 mW cm–2). The concentrations of AIE810NP are 100 μg mL–1.

The quantum yield (QY), photostability, thermal stability, and colloidal stability of AIE810NPs were characterized. The QY of AIE810NP was determined using indocyanine green (ICG, QY = 1.6%) in water as a reference.[28] The QY of AIE810NP was calculated to be 2.1% (Figure S10, Supporting Information). The photostabilities of AIE810NP and BPBT were evaluated by monitoring the FL intensity under continuous radiation with a 730 nm laser (1 W cm–2). Given its wide utility in fluorescent bioanalysis, cyanine dye derivative (abbreviated as Cy7) was chosen as a reference for comparison (Figures S11 and S12, Supporting Information).[31−33] AIE810NP and BPBT show 6.5 and 3.3% loss of FL intensity after 60 min of continuous radiation, respectively, whereas Cy7 shows a 98% loss of FL intensity after 30 min of continuous radiation (Figure a), demonstrating that BPBT and AIE810NP have excellent photostability. The thermal stability of AIE810NP was evaluated by incubating it in water at 50 °C for 1 week. The loss in FL intensity of AIE810NP is 12.6%, whereas that of Cy7 is 67.5% (Figure b), revealing a better thermal stability for AIE810NP. In addition, the hydrodynamic diameter of AIE810NPs stored at 37 °C for 1 week is the same as before (Figure c), presumably due to the existence of a large number of negative-charged carboxyl groups on the surface of AIE810NP (Figure S13, Supporting Information), indicating that the AIE810NP has excellent colloidal stability. All of these results indicate that the synthesized AIE810NP is suitable to be used as a fluorescent reporter for test strip detection.

Figure 2

Characterization of photophysical properties of AIE810NP. (a) Changes in FL intensity of Cy7 (0.35 μM; λem = 770 nm), AIE810NP (50 μg mL–1; λem = 810 nm), and BPBT (10 μM, THF) under continuous radiation of a 730 nm laser (1 W cm–2). (b) Changes in FL intensity of Cy7 (0.35 μM) and AIE810NP (50 μg mL–1) upon incubation in water at 50 °C for 1 week. The standard deviation is calculated from the results of three independent tests. (c) Hydrodynamic diameter of AIE810NP in water before and after storage at 37 °C for 1 week (0.2 mg mL–1).

Optimization of Immunoreaction Conditions

To ensure the specific binding of AIE810NP to IgM/IgG, SARS-CoV-2 antigen (recombinant SARS-CoV-2 spike glycoproteins) has been modified on the surface of AIE810NP (AIE810NP-SARS-CoV-2 antigen) (Scheme a).[34,35] The hydrodynamic diameter of AIE810NP-SARS-CoV-2 antigen (324.3 ± 5.9 nm) is equal to the sum of the hydrodynamic diameter of AIE810NP (310.8 ± 4.4 nm) and SARS-CoV-2 antigen (14.5 ± 5.6 nm) (Figure a). The zeta-potential of AIE810NP-SARS-CoV-2 antigen (−11.7 ± 1 mV) is between the AIE810NP (−21.9 ± 1.1 mV) and SARS-CoV-2 antigen (−7 ± 1 mV) (Figure b). The AIE810NP-labeled chicken IgY (AIE810NP-chicken IgY) was prepared with a synthetic procedure similar to that used for AIE810NP-SARS-CoV-2 antigen. As shown in Figure c,d, the hydrodynamic diameter and zeta-potential of AIE810NP-chicken IgY (324.5 ± 2.6 nm and −15.9 ± 0.25 mV) are significantly higher than those of AIE810NP. The results indicate that SARS-CoV-2 antigen and chicken IgY are successfully modified on the surface of AIE810NP.

Figure 3

Characterization of AIE810NP-SARS-CoV-2 antigen and AIE810NP-chicken IgY. (a,b) Hydrodynamic diameter and zeta-potential of SARS-CoV-2 antigen, AIE810NP, and AIE810NP-SARS-CoV-2 antigen in water. The concentrations of SARS-CoV-2 antigen, AIE810NP, and AIE810NP-SARS-CoV-2 antigen are 50 μg mL–1. The standard deviation is calculated from the results of three independent tests. (c,d) Hydrodynamic diameter and zeta-potential of chicken IgY, AIE810NP, and AIE810NP-chicken IgY in water. The concentrations of chicken IgY, AIE810NP, and AIE810NP-chicken IgY are 50 μg mL–1. The standard deviation was calculated from the results of three independent tests.

Optimization of the preparation conditions is the key parameter to improve the detection performance of the test strip. Here, immunoreaction time before readout and the amount of AIE810NP-SARS-CoV-2 antigen were evaluated. As shown in Figure a,b, the IM/IC and IG/IC gradually reach a plateau at about 10 min, indicating that the optimal readout time is 10 min. For IgM detection, the IM/IC of the IgM sample (SM) increases significantly with the increasing AIE810NP-SARS-CoV-2 antigen concentration from 3.1 to 25 μg mL–1 and then no longer changes when the concentration is 50 μg mL–1 (Figure c). Moreover, the IM/IC of preCOVID sample (NM) enhances with the increase of AIE810NP-SARS-CoV-2 antigen concentration. The ratio of SM/NM reaches the maximum at the AIE810NP-SARS-CoV-2 antigen concentration of 12.5 μg mL–1 (Figure d). For IgG detection, the IG/IC of the IgG sample (SG) increases with increasing the AIE810NP-SARS-CoV-2 antigen concentration from 3.1 to 25 μg mL–1 and then decreases at 50 μg mL–1, whereas the IG/IC of the preCOVID sample (NG) increases significantly with the increase of the AIE810NP-SARS-CoV-2 antigen concentration (Figure e). The maximum value of the ratio of SG/NG occurs with the AIE810NP-SARS-CoV-2 antigen concentration of 6.3 μg mL–1 (Figure f). Considering the importance of IgM in early detection, 12.5 μg mL–1 is chosen as the optimal concentration of AIE810NP-SARS-CoV-2 antigen. In addition, to achieve the highest signal-to-noise ratio, the concentrations of mouse anti-human IgM (fixed on the M line) and IgG (fixed on the G line) were optimized. The mouse anti-human IgM and IgG were prediluted in phosphate buffer saline to prepare the reserve solution with the final concentrations of 0.3, 0.6, and 1.2 mg mL–1, respectively, and spotted onto the Pall90 membrane. The optimal concentrations of mouse anti-human IgM and IgG are 0.6 mg mL–1 (Figure S14, Supporting Information).

Figure 4

Optimization of parameters pertaining to the AIE810NP-based test strip performance. (a,b) IM/IC and IG/IC under different incubation time. The concentrations of the IgG and IgM calibrators are 0.15 and 0.5 μg mL–1, respectively. The standard deviation is calculated from the results of three independent tests. (c) IM/ICversus various amounts of AIE810NP-SARS-CoV-2 antigen used in the test strip. The concentrations of the IgM calibrator are 0.5 μg mL–1. The standard deviation is calculated from the results of three independent tests. (d) Ratio of the IM/IC of the IgM sample (SM) and the IM/IC of the preCOVID sample (NM). (e) IG/ICversus various amounts of AIE810NP-SARS-CoV-2 antigen used in the test strip. The concentrations of the IgG calibrator are 0.3 μg mL–1. The standard deviation is calculated from the results of three independent tests. (f) Ratio of the IG/IC of the IgG sample (SG) and the IG/IC of the preCOVID sample (NG).

Application to the Analysis of IgM and IgG in Clinical Serum Samples

To determine the thresholds of the AIE810NP-based test strip for detecting IgM and IgG in serum samples, 142 preCOVID samples (including 52 normal serum samples, 44 tuberculosis serum samples, 33 upper respiratory tract infection serum samples, and 13 rheumatoid arthritis serum samples) were tested (Figure a,b and Table S1, Supporting Information). According to the average value (142 preCOVID samples) of IM/IC or IG/IC plus 3-fold standard deviation, the threshold for the detection of IgM and IgG using AIE810NP-based test strip is 0.200 and 0.737, respectively. In other words, the serum samples with IM/IC > 0.200 or IG/IC > 0.737 are positive; otherwise, they are negative.

Figure 5

Analytical performance of AIE810NP-based test strip. (a,b) IM/IC and IG/IC for the detection of IgM and IgG in 142 preCOVID samples using AIE810NP-based test strip, including 52 normal serum samples, 44 tuberculosis serum samples, 33 upper respiratory tract infection serum samples, and 13 rheumatoid arthritis serum samples. The error bars represent the standard deviation of the values. The standard deviation is calculated from the results of 142 independent tests. (c) Calibration curve between IM/IC and IgM concentration in a range of 0.313–5 μg mL–1. The LoD of IgM is 0.236 μg mL–1. The standard deviation is calculated from the results of four independent tests. (d) Calibration curve between IG/IC and IgG concentration in a range of 0.156–5 μg mL–1. The LoD of IgG is 0.125 μg mL–1. The standard deviation is calculated from the results of four independent tests.

On this basis, a set of calibrators with IgM and IgG concentration of 0–5 μg mL–1 was analyzed using the AIE810NP-based test strip. Each sample was analyzed in quadruplicate. The experimental results show a good correlation between IM/IC and IgM concentration following an equation of y = −3.71 × e– + 2.787 (R2 = 0.97) (Figure c and Figure S15a, Supporting Information). Similarly, there is a good correlation between IG/IC and IgG concentration following an equation of y = −3.818 × e– + 3.563 (R2 = 0.91) (Figure d and Figure S15b, Supporting Information). The limit of detection (LoD) of IgM and IgG is determined to be 0.236 and 0.125 μg mL–1, respectively, commensurate with that of the ELISA (Figure S16, Supporting Information). The low LoD of the AIE810NP-based test strip could be attributed to the autofluorescence-free background of the NC membrane/biosample in the NIR region (enhancement of the signal-to-noise ratio) (Figure S17, Supporting Information).

A total of 172 serum samples from patients infected with SARS-CoV-2 (1–224 days after symptom onset) were analyzed using AIE810NP-based test strips for the detection of IgM and IgG (172 of samples were confirmed as SARS-CoV-2 infection by RT-PCR). Simultaneously, all samples were tested by ELISA and AuNP-based test strips as the control. As shown in Table , 146 cases of IgM positive and 164 cases of IgG positive were detected by ELISA; 70 cases of IgM positive and 147 cases of IgG positive were detected by AuNP-based test strips, and 135 cases of IgM positive and 164 cases of IgG positive were detected by AIE810NP-based test strips. The positive results of AIE810NP-based test strips for detecting IgM and IgG were consistent with that of ELISA, with only one case of IgG and two cases of IgM being different (Table S2, Supporting Information). The sensitivity of AIE810NP-based test strips for the detection of IgM and IgG is 78 and 95%, respectively, commensurate with that of ELISA (85 and 95%) and superior to that of the commercial AuNP-based test strips (41 and 85%). These results indicate that AIE810NP-based test strips have high sensitivity and are suitable for the detection of IgM and IgG in serum samples. Additionally, the sensitivity of AIE810NP-based test strips was comparable to that of the SERS-based test strips.

Table 1

Sensitivity for Detecting IgM and IgG in 172 Serum Samples Using ELISA, AuNP-Based Test Strips, and AIE810NP-Based Test Strips (“P” means “Positive”, “N” means “Negative” and “S” means “Sensitivity”)

	ELISA			AuNP-based test strip			AIE810NP-based test strip
target	P	N	S	P	N	S	P	N	S
IgM test (n = 172)	146	26	85%	70	102	41%	135	37	78%
IgG test (n = 172)	164	8	95%	147	25	85%	164	8	95%

Monitoring of IgM and IgG in Sequential Clinical Samples

Given its high sensitivity, the AIE810NP-based test strip was further extended for monitoring of IgM and IgG in sequential clinical samples. As shown in Table , the IgG or IgM can be observed 8–15 days after symptom onset using AuNP-based test strips, and the IgG or IgM can be observed 1–7 days after symptom onset using ELISA and AIE810NP-based test strips. These results demonstrated that the AIE810NP-based test strip could earlier detect IgM or IgG in the seroconversion window period than the AuNP-based test strip. Therefore, our developed AIE810NP-based test strip could become a promising alternative to AuNP-based test strips and ELISA for the early detection of IgM and IgG in the seroconversion window period.

Table 2

Detection of IgM or IgG in Sequential Clinical Samples Using ELISA, AuNP-Based Test Strips, and AIE810NP-Based Test Strip

Conclusion

In summary, we proposed an AIE810NP-based lateral flow immunoassay for early detection of IgM and IgG against SARS-CoV-2 in a seroconversion window period. The synthesized AIE810NP possesses a large Stokes shift of 145 nm, good photostability, high thermal stability, and strong colloidal stability. Using the NIR emission of AIE810NP as the detection signal, the interference of autofluorescence from the NC membrane and human serum was efficiently eliminated, and the sensitivity of the lateral flow test strip was improved. With the homemade portable reader, the LoD of detecting IgM and IgG with an AIE810NP-based test strip (0.236 and 0.125 μg mL–1) is comparable to that of ELISA (0.040 and 0.039 μg mL–1). The sensitivity of AIE810NP-based test strips (78 and 95%) in detecting IgM and IgG of 172 COVID samples are comparable to that of ELISA (85 and 95%) and better than that of commercial AuNP-based test strips (41 and 85%). Importantly, AIE810NP-based test strips (1–7 days) detect IgM or IgG in sequential clinical samples earlier than commercial AuNP-based test strips (8–15 days). Therefore, the AIE810NP-based lateral flow immunoassay can be used as an alternative method for early detection of IgM and IgG against SARS-CoV-2 and displays the great potential for point-of-care clinical diagnosis not only for SARS-CoV-2 but also for other virus outbreaks.

Experimental Section

Materials

PS nanoparticles, SARS-CoV-2 specific antigen (recombinant SARS-CoV-2 spike glycoproteins), mouse anti-human IgM, mouse anti-human IgG, a colloidal gold lateral flow immunoassay kit, PVC base, nitrocellulose (NC) membrane (Pall90), absorbent pads, sample pads, and phosphate buffer saline (PBS) were provided by Shanghai Taywell Biotech Co., Ltd. (Shanghai, China). Goat anti-chicken IgY and chicken IgY were purchased from Genstars Biotech CO., Ltd. Bovine serum albumin (BSA), N-(3-(dimethylamino)propyl)-N′-ethylcarbodiimide hydrochloride (EDC), and N-hydroxysulfosuccinimide sodium salt (Sulfo-NHS) were purchased from Sigma-Aldrich (Shanghai) Trading Co., Ltd. (Shanghai, China). Human anti-SARS-CoV-2 IgM and IgG were purchased from Novoprotein Scientific Inc. A total of 142 preCOVID samples (including 52 normal serum samples, 44 tuberculosis serum samples, 33 upper respiratory tract infection serum samples, 13 rheumatoid arthritis serum sample), 172 COVID samples, and 26 sequential clinical samples were collected from Anhui Medical University and treated in strict accordance with the standard operation for COVID-19 by the World Health Organization. This study was approved by the institutional review board of Anhui Medical University. All experimental procedures were completed under biosafety level II conditions. Solution preparation details can be found in the Supporting Information.

Characterization

The 1H NMR and 13C NMR spectra were collected on a Bruker AV-400 spectrometer. The UV–vis absorption spectrum was recorded on a Shimadzu UV-2600 spectrometer. The NIR emission spectrum was recorded with a modified EK2000-Pro back-thinned fiber spectrometer (Choptics Instruments, Shanghai, China). The dynamic light scattering and zeta-potential results were conducted with a Nano-ZS90 ZetaSizer (Malvern Instruments Ltd., UK). TEM images were acquired on a Hitachi HT7700 Exalens TEM (Hitachi Co., Ltd., Japan). A BioDot XYZ platform was supplied by BioDot (BioDot, Inc., USA).

NIR-Emissive AIE Nanoparticle Synthesis

NIR-emissive AIE nanoparticles were obtained by embedding BPBT into the PS nanoparticles by the swelling method.[30] Briefly, 50 μL of BPBT solution (dissolved by THF, 20 mg mL–1) and 250 μL of 40 mg mL–1 PS nanoparticle aqueous solution (containing 0.4 mg sodium dodecyl sulfate) were successively added to 500 μL of acetone aqueous solution. The mixture was stirred at room temperature for 6 h. Afterward, the resultant AIE nanoparticles were separated by centrifugation at 12,000 rpm for 10 min and resuspended in 1 mL of Milli-Q water for further use.

AIE810NP Conjugate Preparation

The AIE810NP-SARS-CoV-2 antigen conjugation and AIE810NP-chicken IgY conjugation were prepared with the EDC-NHS method.[34,35] Briefly, 50 μL of 10 mg mL–1 EDC solution and 50 μL of 10 mg mL–1 Sulfo-NHS solution (dissolved in borate-buffered saline (BBS), pH 7.4) were successively added to 500 μL of a 4 mg mL–1 AIE810NP solution (dissolved by BBS, pH 7.4). The mixture was stirred at room temperature for 30 min. After being activated, the AIE810NPs were separated by centrifugation and resuspended in 500 μL of BBS (pH 7.4). Then, 0.25 mg of SARS-CoV-2 antigen was added, and the mixture was stirred for 3 h at room temperature to form a tripartite complex of AIE810NP-SARS-CoV-2 antigen. Thereafter, the complex was further blocked by adding 50 μL of 100 mg mL–1 BSA aqueous solution for another 1 h at room temperature. After incubation, the AIE810NP-SARS-CoV-2 antigen was centrifuged to remove the unreacted reagent and redispersed in BBS containing 1% BSA and 0.05% proclin300 (pH 7.4) and stored in 4 °C for further use. The AIE810NP-labeled chicken IgY (AIE810NP-chicken IgY) were prepared with a synthetic procedure similar to that used for AIE810NP-SARS-CoV-2 antigen.

AIE810NP-Based Test Strip Preparation

The test strip was composed of five parts including a plastic adhesive backing pad, a sample pad, a conjugation pad, a NC membrane, and an absorbent pad. Mouse anti-human IgM (0.6 mg mL–1), mouse anti-human IgG (0.6 mg mL–1), and goat anti-chicken IgY (1.0 mg mL–1) prediluted in PBS (pH 6.8, containing 2% sucrose, 1% NaCl) were spotted onto the NC membranes as a test 1 line (M line), test 2 line (G line), and a control line (C line), respectively. The AIE810NP-SARS-CoV-2 antigen (5 mg mL–1) and AIE810NP-chicken IgY (2.5 mg mL–1) prediluted in PBS (pH 6.8, containing 2% sucrose, 2% BSA, and 2% NaCl) were spotted onto the conjugation pad followed by drying at 37 °C for overnight. Last, the well-assembled test strips were cut to a width of 3.8 mm and fitted into the plastic adhesive backing pad for further use.

AIE810NP-Based Test Strip for the Detection of IgM and IgG against SARS-CoV-2 in COVID Samples

The LoD of AIE810NP-based test strips was evaluated by detecting a set of calibrators with an IgM and IgG concentration of 0–5 μg mL–1. A total of 142 preCOVID samples were as negative controls. In preparation, 5 μL of the prepared serum sample was added into 95 μL of running buffer (consisting of BBS, pH 8.0; 0.25% Tween, 0.25% Thesit, and 2% BSA), and the mixture was subsequently added into the sample well. After a 10 min reaction, the fluorescence intensities of the M line (IM), the G line (IG), and the C line (IC) were recorded using a CMOS camera (Huatengwei Vision Technology Co., Ltd., China). The IM/IC and IG/IC were used to calculate the IgM and IgG concentrations in the serum, respectively. For 172 COVID samples and 26 sequential clinical sample detections, 5 μL of the serum sample was added into 95 μL of running buffer, and then the mixture was subsequently added into the sample well. After a 10 min reaction, the fluorescence intensities of the M line (IM), the G line (IG), and the C line (IC) were recorded.