Serum peptidome profiles immune response of COVID-19 Vaccine administration

Abstract

Background: Coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused significant loss of life and property. In response to the serious pandemic, recently developed vaccines against SARS-CoV-2 have been administrated to the public. Nevertheless, the research on human immunization response against COVID-19 vaccines is insufficient. Although much information associated with vaccine efficacy, safety and immunogenicity has been reported by pharmaceutical companies based on laboratory studies and clinical trials, vaccine evaluation needs to be extended further to better understand the effect of COVID-19 vaccines on human beings.

Methods: We performed a comparative peptidome analysis on serum samples from 95 participants collected at four time points before and after receiving CoronaVac. The collected serum samples were analyzed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) to profile the serum peptides, and also subjected to humoral and cellular immune response analyses to obtain typical immunogenicity information.

Results: Significant difference in serum peptidome profiles by MALDI-TOF MS was observed after vaccination. By supervised statistical analysis, a total of 13 serum MALDI-TOF MS feature peaks were obtained on day 28 and day 42 of vaccination. The feature peaks were identified as component C1q receptor, CD59 glycoprotein, mannose-binding protein C, platelet basic protein, CD99 antigen, Leucine-rich alpha-2-glycoprotein, integral membrane protein 2B, platelet factor 4 and hemoglobin subunits. Combining with immunogenicity analysis, the study provided evidence for the humoral and cellular immune responses activated by CoronaVac. Furthermore, we found that it is possible to distinguish neutralizing antibody (NAbs)-positive from NAbs-negative individuals after complete vaccination using the serum peptidome profiles by MALDI-TOF MS together with machine learning methods, including random forest (RF), partial least squares-discriminant analysis (PLS-DA), linear support vector machine (SVM) and logistic regression (LR).

Conclusions: The study shows the promise of MALDI-TOF MS-based serum peptidome analysis for the assessment of immune responses activated by COVID-19 vaccination, and discovered a panel of serum peptides biomarkers for COVID-19 vaccination and for NAbs generation. The method developed in this study can help not only in the development of new vaccines, but also in the post-marketing evaluation of developed vaccines.

Keywords: COVID-19; MALDI-TOF; immune response; peptidome; serum; vaccine.

Figure 1

Study design. (A) Ninety-five participants were recruited and received two doses of CoronaVac vaccines. Blood samples were collected at four time points before and after vaccine injection. (B) Serum separation. (C) MALDI-TOF MS and immunogenicity analysis on all the collected serum samples. (D) MALDI-TOF MS data processing using MALDIquant and Metaboanalyst. (E) Feature selection based on volcano plot and variable importance in projection (VIP) scores by PLS-DA. Hierarchical cluster analyses (HCA) showed the intensity distribution of the selected features among samples. Peaks with an asterisk are peaks of higher intensity filtered by the signal-to-noise ratio and are included in the analysis. Peaks without an asterisk are by default noise peaks and are not included in the analysis.

Figure 2

(A) The representative MALDI-TOF mass spectra of one participant at four time points pre-vaccination and post-vaccination. (B) Partial enlarged view of (A). The mass spectra were normalized against the strongest peak. r.i.: relative intensity.

Figure 3

Principal component analysis (PCA) of the MALDI-TOF MS-based serum peptidome profiles collected at 4 time points before and after vaccination: (A) between day 0 and day 21; (B) between day 0 and day 28; (C) between day 0 and day 42; (D) between day 21 and day 28; (E) between day 21 and day 42; (F) between day 28 and day 42. The shadow ovals represent 95% confidence interval.

Figure 4

Selection of feature peaks of vaccination. (A) A general scheme of sample collection, data processing and feature selection. (B) Cluster analysis of the 13 feature peaks of vaccination among all the samples. (C) The Gene Ontology (GO) enrichment analysis by Metascape involving all the identified features.

Figure 5

Classification of NAbs positive and negative based on MALDI-TOF MS serum peptidome. (A) PLS-DA analysis of all the 95 samples collected on day 42 to classify NAbs positive and negative; (B) PLS-DA analysis of 55 samples collected on day 42 from female individuals to classify NAbs positive and negative; (C) PLS-DA analysis of 40 samples collected on day 42 from male individuals to classify NAbs positive and negative; (D) a general scheme for feature selection and model establishment to classify NAbs positive and negative; (E) heat map of six feature peaks of NAbs generation among all the training samples; (F) performance of four classification models on the 14 test samples for NAbs generation prediction. N: negative, P: positive, RF: random forest, PLS-DA: partial least squares discriminant analysis, SVM: linear support vector machine, LR: logistic regression. The colored ovals represent 95% confidence interval.