Systematic evaluation of SARS-CoV-2 spike protein derived peptides for diagnosis of COVID-19 patients

Four independent cohorts of samples were collected and designed

To fully evaluate the diagnostic potential of the spike derived peptides, sera from four cohorts of COVID-19 patients and controls from multiple medical centers of China were collected (Table 1). 1) Cohort 1 consists of 55 sera from convalescent COVID-19 patients and 18 controls[17]; 2) Cohort 2 includes 2,360 sera from 784 in-hospital COVID-19 patients and 542 sera from a variety of controls. To accurately evaluate the peptides for diagnosis, for each patient, one serum sample that collected at least 21 days after onset of symptoms was selected according to the suggestion of WHO (World Health Organization) for antibody laboratory test[18]. As a result, in total, 729 sera were selected. The control groups include two types of samples. The first type is sera collected from hospitals, including sera from healthy people (n=92), upper respiratory infections (URI, n=104), patients with autoimmune diseases (AID, n=120), lung cancer patients (n=41) and patients with other diseases (n=112) that consist of cardiovascular or cerebrovascular diseases (34.2%), diabetes (9%), non-lung cancers (7.2%) and others. The second type is negative references provided by National Institutes for Food and Drug Control, including N11-N25 of National Reference Panel for 2019-nCoV (SARS-CoV-2) IgG Antibody Detection Kit (370096-202001) and other identified controls. 3) Cohort 3 includes 19 COVID-19 patients from another hospital and 50 healthy controls. 4) Cohort 4 consists of asymptomatic patients defined as positive either on NAT (nucleic acid test) or antibody test conducted by a commercial assay, and with a SARS-CoV-2 exposure history (see methods).

Peptide S2-78 and several other peptides exhibit high diagnostic values

To identify which section of S protein has diagnostic value, it is necessary to survey the entire protein on a systematic way. We took advantage of a previously constructed peptide microarray which has full coverage of S protein[17], analyzed 55 convalescent sera of COVID-19 patients along with 18 control sera (Cohort 1, Table 1). Overall, significant bindings of both IgG and IgM were observed in patient group, while the signals were low in the control group (Fig. 1a). Several peptides, e. g., S2-78 and S1-93 exhibit strong IgG antibody bindings and high response frequencies in patients, we define these peptides as “significant” peptides, these peptides may have diagnostic values (Fig. 1b). To test whether the peptide specific IgG antibody responses are concentration dependent, 3x serially diluted peptides (0.9, 0.3 and 0.1 mg/mL) were printed and immobilized on the microarray. As expected, the averaged signals of the patient group, but not the control group, are proportional to the concentrations of the peptides (Fig. 1c).

• Download figure
• Open in new tab

Figure 1. Peptides with strong IgG antibody responses were identified by Cohort 1.

A. Representative images of peptide microarray profiled by sera from a COVID patient and a healthy control. IgG (green) and IgM (red) were detected simultaneously. B. Box plot of IgG antibody responses against SI protein, N protein and some significant peptides for COVID-19 patient group (n=55) and control group (n=18) of Cohort 1. Each spot indicates one serum sample. Data are presented as box plots where the middle line is the mean value, and the upper and lower hinges are mean values ± SD. AUC (area under curve) values are labeled for each peptide or protein on the top of the box plots. C. Averaged signal intensities of IgG antibody responses against the indicated peptides at different concentrations, i.e. 0.1, 0.3 and 0.9 mg/mL, both for COVID-19 patients (blue line) and control group (orange line).

To extensively evaluate the peptides for diagnostic application, a larger cohort (Cohort 2, Table 1) of samples were screened by a revised peptide microarray that contains only one peptide concentration (0.3 mg/mL) for high-throughput analysis. To ensure the data generated from different microarrays are comparable, we prepared a positive reference sample by pooling 50 randomly selected patient sera. This reference sample was then tested on all the microarrays for normalization (see methods). Consistent results were achieved for most of the peptides. AUC (area under curve) values of IgG or IgM for each peptide were calculated. Eight peptides, i. e., S2-78, S1-97, S1-93, S1-101, S1-111, S2-97, 51-105 and S2-22 are of high performance, the AUCs of IgG or IgM against these peptides for both Cohort 1 and 2 are above 0.85. Specifically, for Cohort 2 the AUC values with 95% CI (confidential intervals) for S2-78, S1-97, S1-93, S1-101, S1-111, 52-97, S1-105 and S2-22 were 0.99 (0.986-0.995), 0.954 (0.942-0.965), 0.934 (0.92-0.948), 0.932 (0.917-0.947), 0.929 (0.915-0.943), 0.922 (0.907-0.938), 0.909 (0.893-0.926), and 0.866 (0.846-0.886), respectively (Fig. 2a). For IgM, only S2-78 has a AUC >0.85, i. e., 0.953 (0.941-0.964) (Fig. 2a). As expected, IgG and IgM for both N protein and S1 protein exhibited high performance. We further examined the antibody responses in different groups for each peptide (Fig. 2b-e, Fig. S1). Consistently, signals for COVID-19 group are significantly higher than that of the negative samples in all groups. It is noted that the signal intensity for S1 IgG is generally higher than that of any single peptide in the group of COVID-19 patients, this may because there are multiple antibody binding sites on S1 protein. However, slightly higher signal is also observed in control group for protein antigens, demonstrating non-specific binding while it is largely eliminated for synthetic peptides. It is suggested that more sensitive detection platforms or higher antigen concentration might improve the performance of peptides for diagnosis.

• Download figure
• Open in new tab

Figure S1. Evaluation of the significant peptides by Cohort 2 - other peptides.

Signal levels of the antibodies against the indicated peptides in COVID-19 patients (n=729), Healthy controls (n=92), upper respiratory infections (URI, n=104), patients with autoimmune diseases (AID, n=120), lung cancer patients (n=41), patients with other diseases (n=112) and negative reference samples (n=73) of Cohort 2.

• Download figure
• Open in new tab

Figure 2. Evaluation significant peptides by Cohort 2.

A. ROCs (receiver operating curves) of the peptides or proteins for discrimination of the COVID-19 group (n=729) and the control group (n=542). The AUC values with 95% Cl (confidential intervals) are provided for all the peptides and proteins. B-E. signal distributions of anti-S1 IgG (B), anti-S2-78 IgG (C), anti-S1-97 IgG (D) and anti-S1-93 (E) in COVID-19 patients (blue) and control groups (yellow). Sample size is indicated for each group. URI: upper respiratory infection, AID: patients with autoimmune diseases.

The diagnostic performance of S2-78 IgG is comparable to that of S1 IgG for COVID-19 patients

We next focused on S2-78, the peptide of best performance for detection. Optimal Youden index of ROC (receiver operating characteristic) curve was used to set the cutoff value. The specificity, sensitivity and overall accuracy (95% CI) of S2-78 IgG for detection of COVID-19 are 96.7 (94.8-98.0%), 95.5% (93.7-96.9%) and 96% (94.8-97%), respectively, which are slightly lower than that of S1 IgG (Table 2). Since serological test is essential for population screening, we calculated the PPV (positive predict value) and NPV (negative predict value) of two assumed prevalence rates. One is 0.04 for general population originated from the situations of Wuhan, China[19] and Netherlands[20]. The other is 0.5 for a high risk population[7]. For prevalence rate of 0.5, both PPV and NPV of S2-78 IgG is similar to that of S1 IgG, however, for prevalence rate of 0.04, the PPV is only 54.7%, although the NPV is extremely high, suggesting the antibody detection of S2-78 could effectively exclude negative ones but may generate high false positive rate at low rate. However, for low prevalence rate, since the number of real positive are very low, it might be acceptable to perform additional test by using other antigens to improve the overall performance.

We next investigated the consistency between S1 IgG and S2-78 IgG (Fig. 3a, b). The overall consistencies for COVID-19 group and Control group are 96.3% and 96.7%, respectively. Interestingly, there are eight samples from COVID-19 patients are negative for S1 IgG but positive for S2-78 IgG, suggesting it will be of diagnostic value to combine S2-78 and S1. It is known that the immune response may correlate to some key clinical parameters, such as gender, disease severity, age and the final outcome[5,9]. To test whether the positive rate of S2-78 IgG is associated with these clinical parameters, we analyzed the detection sensitivities among subgroups, i. e., male vs. female, ≥60 vs. <60 for age, severe vs. non-severe cases and survivor vs. non-survivor with critical diseases. Similar to that of S1 IgG (Fig. S2a), no significant difference was observed in all these subgroups.

• Download figure
• Open in new tab

Figure S2. Performance of peptides and SI for diagnosis in subgroups.

A. forest plot of sensitivities of SI IgG in different subgroups, i.e., age, gender, severity and outcome. The dots indicate the sensitivities while the error bars indicate the 95% CI. The exact values are also provided. P values were calculated with χ2 test. B. Graph of positive rates of IgG antibodies against the indicated peptides versus days after symptom onset in 2,360 serum samples from 784 patients.

• Download figure
• Open in new tab

Figure 3. S2-78 IgG for diagnosis of COVID-19.

A. Scatter plots of sera samples from COVID-19 (blue dots) and controls (orange triangle) for S1 IgG vs. S2-8 IgG. The grey lines indicate the cut-off values set as the optimal Youden index based on the ROC. The orange and blue numbers indicate the sample counts of control and patient on each quadrant, respectively. B. consistency between S1 protein and S2-78. The consistency values are provided with 95% CI. C. forest plot of sensitivities of S2-78 IgG among subgroups, i.e., age, gender, severity and outcome. The dots indicate the sensitivities while the error bars indicate the 95% CI. The exact values are also provided. P values were calculated with χ2 test. D-E. Graph of positive rates of IgG or IgM against the S, N proteins or S2-78 versus days after symptom onset in 2,360 serum samples from 784 patients.

Virus specific antibodies would persistently increase and usually conduct seroconversion within one or two weeks after symptom onset, and the portion of patients with positive antibodies reach ~100% after two or three weeks[21,22]. To test whether S2-78 IgG is suitable to detect the dynamic change of antibody response, we analyzed the positive rates at different time points after the onset of symptoms. Expectedly, the positive rate of S2-78 IgG continuously increased and reach the plateau about three weeks after onset, which is similar as S1 and N IgG (Fig. 3d). Similar trends were observed for S2-78 IgM (Fig. 3e) and IgG antibodies against other peptides (Fig. S2b). These observations suggest that the antibodies against S2-78 and other peptides could be applied for monitoring virus specific antibody dynamics.

The diagnostic performance of S2-78 IgG is comparable to that of S1 IgG for asymptomatic infections

Monitoring of asymptomatic infections is essential to put SARS-CoV-2 infection under control. It is thought that asymptomatic infection usually has a weak immune response[23]. To test whether S2-78 IgG could be used for detection of asymptomatic infection, we analyzed 63 asymptomatic patients (Cohort 4, Table 1) defined as positive either in NAT test or antibody test conducted by a commercial assay (see methods) but without obvious symptom. Four subgroups were divided according to positive or not for NAT, IgG or IgM[24]. It was shown that for S1 IgG and S2-78 IgG, all samples (n=4) of IgG^- group were negative, while of the 59 IgG⁺ asymptomatic infections, 47 and 45 were positive for S1 IgG and S2-78 IgG, respectively. (Fig 4a, b). The consistency between S1 IgG and S2-78 IgG was also high (93.7%). The contradictory results in IgG⁺ group between our data and the commercial assay may due to the differences of antigens involved. S and N recombinant proteins are used for the commercial assay, for which slightly lower specificity is common[7]. Overall, these results demonstrate the diagnostic and screening value of S2-78 IgG for asymptomatic infections.

• Download figure
• Open in new tab

Figure 4. S2-78 IgG for diagnosis of asymptomatic patients.

A-B. Levels of SI IgG and S2-78 IgG in asymptomatic patient groups divided by positive or negative among three tests, i.e., NAT (nucleic acid test), SARS-CoV-2 vims specific IgG and IgM antibodies. The dashed lines indicate the cut-off values. The number of positive samples as well as the total number of samples of IgG⁺ (IgG positive) groups were presented for SI IgG and S2-78 IgG. C. The consistency between SI IgG and S2-78 IgG.

The diagnostic value of S2-78 IgG was validated by ELISA

ELISA (enzyme linked immunosorbent assay) is common for commercial SARS-CoV-2 antibody assays[8,25]. To verify the efficacy and applicability of S2-78 IgG, we established an ELISA assay. Firstly, to test the consistency of the peptide microarray and ELISA, we randomly selected 31 sera from COVID-19 patients of Cohort 1 and tested by ELISA. High consistency was achieved with a Pearson correlation of 0.926 (Fig. 5a), demonstrating the validity of the microarray results. To further validate the performance of S2-78 IgG, we screened another independent cohort of samples collected from a different medical center (Cohort 3, Table 1). As expected, high performance of S2-78 IgG for specific detection of COVID-19 was achieved by ELISA (Fig. 5b, c).

• Download figure
• Open in new tab

Figure 5. ELISA validation of S2-78 by Cohort 3.

A. Correlation of the signals of S2-78 IgG response between peptide microarray and ELISA with Cohort 1 (n=31). B. Levels of S2-78 IgG in samples from COVID-19 patients either from Cohort 1 (n=31) or Cohort 3 (n=19) and healthy controls (n=50). C. ROC of S2-78 IgG for discrimination of 19 COVID-patients (Cohort 3) and 50 healthy controls.

A two-step strategy by combining S2-78 and other peptides for specific detection of SARS-CoV-2 infections

Because of the highly homologous genomes among human coronaviruses, one unneglectable risk for antibody based diagnosis of SARS-CoV-2 infection is the possible cross-reactivity with other coronaviruses, especially for those common cold causing coronaviruses, i. e., HCoV-OC43, HKU1, NL63 and 229E. The protein sequences of S protein among those viruses are of high similarity, selection of sections of S protein that are unique for SARS-CoV-2 is important for diagnosis[11,13]. To investigate whether the identified peptides are specific to SARS-CoV-2, we performed homology analysis among SARS-CoV-2 and 6 other coronaviruses (Extended data Fig 3). High homologies were observed for S2-78 and S2-22, suggesting they alone may not suitable for specific detection of SARS-Cov-2 infection. In contrast, other peptides, i. e, S1-93, 97,101,105,111 and S2-97, exhibit low similarities with other coronaviruses, particularly the four coronaviruses that cause common cold in human, suggesting they could be applied for specifically detection of SARS-Cov-2 infection. S1-97 with the best performance among these peptides was selected for further investigation. Considering the relatively low sensitivity (86.2%) of S1-97 IgG (Fig. 2, Fig. S4), we proposed a two-step strategy by combining S1-97 IgG and S2-78 IgG for detection (Fig. 6a). To simplify this strategy, there are three assumptions: 1) S2-78 IgG can detect the infection of SARS-CoV-2 and related coronaviruses with the same sensitivity (95.5%) and generates a false positive rate of 3.3% (1-speficicity, 1-96.7%) for healthy and patients of other diseases; 2) S1-97 IgG can specially detect SARS-CoV-2 infection with 86.2% sensitivity and no cross-reactivity with related coronaviruses, so it would generate a false positive rate of 6.8% (1-speficicity, 1-93.2%) for the infection of related coronaviruses, healthy control and patients with other diseases; 3) The positive rate of S2-78 IgG and S1-97 IgG in each group is independent to each other. According to the two-step strategy (Fig. 6a), for any given sample, the first step is to detect S2-78 IgG; For the positive ones, the second step is to detect S1-97 IgG. The samples positive for both S2-78 and S1-97 are defined as final positive, while the samples negative for either one is defined as negative. As a result, the sensitivity for SARS-CoV-2 infection detection would be 82.3%, the specificity for related coronavirus infections would be 93.5% (specificity 1), and the specificity for control group would be 99.8% (specificity 2). To further improve the performance of the two-step strategy, a panel (Panel-A) of peptides was composed by bivariate regression analysis based on the specificity of each peptide to SARS-CoV-2. The linear function for Panel-A is y =0.014* x₁ + 0.02* x₂ - 0.003* x₃ + 0.006* x₄-2.593, where y represents signals of Panel-A and x₁, x₃, x₃, x₄ represents the signals of IgG against S1-93, 97, 101 and 105, respectively. The sensitivity and specificity of Panel A, based on the data generated from Cohort 2, is 88.3% and 96.7%, respectively (Fig. 6b, c). Follow the two-way strategy by combining S2-78 IgG and Panel-A, the final sensitivity for the detection of SARS-CoV-2 infection, specificity for related coronavirus infection (specificity 1), and the specificity for control group (specificity 2) are 84.3%, 96.8% and 99.9%, respectively (Fig. 6d). When S1 protein and Panel-A are combined, the final sensitivity for the detection of SARS-CoV-2 infection, specificity for related coronavirus infection (specificity 1), and the specificity for control group (specificity 2) are 85.3%, 96.8% and 99.99%, respectively (Fig. S4b). These results demonstrate that combination of human coronavirus conserved peptides or S1 protein and other SARS-CoV-2 specific peptides (“significant” peptides) enable specific detection of SARS-CoV-2 infection with a high specificity and acceptable sensitivity.

• Download figure
• Open in new tab

Figure S3. Homology analysis of the significant peptides among SARS-CoV-2, SARS-CoV, MERS-CoV and other four human coronaviruses.

The line for each position indicates the gap. The bold letters indicate the positions with the same amino acids as that of SARS-CoV-2. The homology analysis was performed by an online tool called Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/).

• Download figure
• Open in new tab

Figure S4. Discrimination of related coronaviruses by combining SI and peptides of Panel-A.

A. ROC of SI-97 IgG based on the results of Cohort 2. B. The combination of SI IgG and Panel-A based on the results of Cohort 2 (729 patients and 542 controls).

• Download figure
• Open in new tab

Figure 6. Discrimination of related coronaviruses with combination of peptides.

A. Illustration of the strategy with combination of S2-78 IgG and S1-97 IgG based on the results of Cohort 2 (729 patients and 542 controls). For any given sample, the first step is to detect S2-78 IgG response. For the positive ones, the second step is to detect S1-97 IgG response. The blue numbers indicate the positive or negative rates for SARS-CoV-2 infections, while the green numbers indicated the assumed positive or negative rates for samples not belong the SARS-CoV-2 group. The last step summaries the overall sensitivities and specificities for different sample groups. B. Signals of Panel-A (S1-93, 97, 101 and 105 in COVID-19 patients and control groups. C. The corresponding ROC of Panel-A. D. The performance of the combination of S2-78 IgG and Panel-A.

Patients and samples

The study was approved by the Ethical Committee of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China (ITJ-C20200128), Institutional Ethics Review Committee of Foshan Fourth Hospital, Foshan, China (202005) and the Ethical Committee of The Fifth Affiliated Hospital of Sun Yat-sen University, Zhuhai, China (K14-2). Written informed consent was obtained from all participants enrolled in this study. COVID-19 patients were hospitalized and received treatment in multiple medical center during the period from 25 January 2020 and 28 April 2020. Sera of the control group from healthy donors, lung cancer patients, patients with autoimmune diseases were collected from Ruijin Hospital, Shanghai, China or Tongren Hospital, Shanghai, China. The negative reference samples were from National Institutes for Food and Drug Control. Sera of 63 asymptomatic patients were also from Tongji Hospital, Wuhan. The IgM and IgG antibodies against recombinant nucleoprotein and spike protein of SARS-CoV-2 in sera of these patients were detected by a commercial kit (YHLO Biotech, Shenzhen, China). According to the instruction of the kit, the antibody level ≥10 AU/ mL is positive, and <10 AU/mL is negative. All the samples were stored at −80°C until use.

Peptide synthesis and conjugation with BSA

The N-terminal amidated peptides were synthesized by GL Biochem, Ltd. (Shanghai, China). Each peptide was individually conjugated with BSA using Sulfo-SMCC (Thermo Fisher Scientific, MA, USA) according to the manufacture’s instruction. Briefly, BSA was activated by Sulfo-SMCC in a molar ratio of 1: 30, followed by dialysis in PBS buffer. The peptide with cysteine was added in a w/w ratio of 1:1 and incubated for 2 h, followed by dialysis in PBS to remove free peptides. A few conjugates were randomly selected for examination by SDS-PAGE. For the conjugates of biotin-BSA-peptide, before conjugation, BSA was labelled with biotin by using NHS-LC-Biotin reagent (Thermo Fisher Scientific, MA, USA) with a molar ratio of 1: 5, and then activated by Sulfo-SMCC.

Peptide microarray fabrication

The peptide-BSA conjugates as well as S1 protein, RBD protein and N protein of SARS-CoV-2, along with the negative (BSA) and positive controls (anti-Human IgG and IgM antibody), were printed in triplicate on PATH substrate slide (Grace Bio-Labs, Oregon, USA) to generate identical arrays in a 1 x 7 (for the peptide microarray with three concentrations) or 2 x 7 subarray format (for the peptide microarray with one concentration) using Super Marathon printer (Arrayjet, UK). The microarrays were stored at −80°C until use.

Microarray-based serum analysis

A 7 or 14-chamber rubber gasket was mounted onto each slide to create individual chambers for the 7 or 14 identical subarrays. The microarray was used for serum profiling as described previously with minor modifications[30]. Briefly, the arrays stored at −80°C were warmed to room temperature and then incubated in blocking buffer (3% BSA in 1×PBS buffer with 0.1% Tween 20) for 3 h. A total of 400 μL for 7-subarray format or 200 μL 14-subarray format of diluted sera or antibodies was incubated with each subarray for 2 h. The sera were diluted at 1:200 for most samples and for competition experiment, free peptides were added at a concentration of 0.25 mg/mL. For the enriched antibodies, 0.1-0.5 p,g antibodies were included in 200 μL incubation buffer. The arrays were washed with 1×PBST and bound antibodies were detected by incubating with Cy3-conjugated goat anti-human IgG and Alexa Fluor 647-conjugated donkey anti-human IgM (Jackson ImmunoResearch, PA, USA), which were diluted for 1: 1,000 in 1×PBST. The incubation was carried out at room temperature for 1 h. The microarrays were then washed with 1×PBST and dried by centrifugation at room temperature and scanned by LuxScan 10K-A (CapitalBio Corporation, Beijing, China) with the parameters set as 95% laser power/ PMT 550 and 95% laser power/ PMT 480 for IgM and IgG, respectively. The fluorescent intensity was extracted by GenePix Pro 6.0 software (Molecular Devices, CA, USA).

Data analysis of peptide microarray

For each spot, signal intensity was defined as the foreground subtracted by the background. The signal intensities of the triplicate spots for each peptide or protein were averaged. For the samples of Cohort 2 and 3, normalization among microarray slides were performed. For each slide, block #14 was incubated with the positive reference sample which was generated by pooling of 50 randomly selected sera from COVID-19 patients. For each slide, the data generated from the positive reference sample were subjected to build a linear regression function, and the data from other samples were subjected to linear normalization according to the function. Graphpad 6.0 was used to generate ROC plots and calculated AUC values.

ELISA

Briefly, 96-well microplates with high binding polystyrene surface (Corning, New York, USA) were coated with 100 μL BSA conjugated peptide (S2-78) at 100 μg/mL and incubated overnight at 4°C. The plates were washed once with PBST buffer (PBS buffer with 0.1% Tween 20), and blocked with 3% BSA (bovine serum albumin) for 1 h at room temperature, followed by one wash with PBST. The sera were diluted at 1:50 in PBST buffer with 1% BSA, 1% FBS and 3% horse serum, 100 μL of the preparation was loaded to each well, the incubation was carried out at 37C for 1.5 h in 100 μL. After six washes with PBST, the secondary antibody, i.e., anti-human IgG-peroxidase (Sangon Biotech, Shanghai, China) was diluted at 1:10000 and incubated at 37 ° C for 1 h, followed by eight washes. Tetramethylbenzidine substrate (Sigma-Aldrich, Missouri, USA) was then added and incubated for 20 min. Finally, 50 μL sulfuric acid (2 M) was added to stop the reaction. The optical density was read at 450 nm using a Behring EL311 ELISA microplate reader (Dade Behring Marburg Gmbh, Berlin, Germany). The assays were repeated twice for each sample.