95

cohort (C2 in Table 1, value = 0.079). Nevertheless, as predicted for serum-protein profiles (Figures 2C and 3A), a significant positive correlation was obtained between mean HBV ELISA O.D. 450 nm values and relative intensity of associated proteins A0A5C2FZ03 and A0A5C2GPZ0 only in asymptomatic and non-severe individuals, respectively (Figure 3C, Supplementary Materials Data File S2).

Autoantibodies in COVID-19 patients have been recently correlated with increased antiviral humoral and inflammatory immune responses [11]. In our study regarding autoantibodies, the survey identified 120 target proteins with significant signal-to-noise ratio (SNR ≥ 3; Figure 3D, Supplementary Materials Data File S3). The net signal intensity for each antigen subtracted by the local background and negative control signal and normalized to internal Ig controls (NSI-nor; Supplementary Materials Data File S3) showed a distribution by different cohorts with the highest NSI-nor average IgM+IgG value (IgM+IgG NSI-nor; Figure 3D). In accordance with serum-protein profiles, autoantibodies were identified mostly in PCR+ cohorts with only one protein (Histone H1), with the highest IgM+IgG NSI-nor value in PCR– individuals (Figures 3D and 4, Supplementary Materials Data File S3). As predicted by our analysis (Supplementary Materials Data File S2), most of the identified proteins reactive to autoantibodies were involved in the regulation of the immune system and/or associated with different diseases (Figure 4, Supplementary Materials Data File S3). For example, nuclear pore-membrane glycoprotein GP210 identified with the highest IgM+IgG NSI-nor value in the severe cohort is a prognostic marker in patients with primary biliary cirrhosis [12]. Another protein with significant signal-to-noise ratio in all PCR– and PCR+ cohorts (SNR ≥ 4.5) and the highest IgM+IgG NSI-nor in severe patients (SNR > 12) was the Type-1 angiotensin II receptor (AGTR; Supplementary Materials Data File S3), which during SARS-CoV-2 infection can recognize and internalize the soluble angiotensin-converting enzyme 2 (ACE2)–coronavirus spike protein complex through dynamin 2-dependent endocytosis [13]. In accordance with our results, these autoantibodies are associated with an unfavorable COVID-19 disease course [14].

The Igs identified as underrepresented in infected cohorts and thus with higher relative levels in PCR– individuals (Figure 5A, Supplementary Materials Data File S2) were used for the identification of vaccine-induced protective epitopes using a designed analytical workflow (Figure 5B, Supplementary Materials Data File S2). The results showed that the protective epitopes were not only identified in the SARS-CoV-2 spike (S) receptor-binding domain (RBD) associated with vaccine-protective capacity [15], but also in other virus proteins such as envelope small-membrane glycoprotein M (ORF3a), membrane-protein E, and nucleocapsid phosphoprotein N (ORF1ab) (Figure 5C, Supplementary Materials Data File S2). A correlation analysis was conducted between SARS-CoV-2-neutralizing antibodies (Table 1) and mass-spectra relative intensity of identified Igs with vaccineinduced protective epitopes (Supplementary Materials Data Files S1 and S2). The results showed no significant correlation for all cohorts together (R<sup>2</sup> = 0.182) but revealed a positive correlation in PCR– individuals in whom these proteins were overrepresented (R<sup>2</sup> = 0.826), thus providing support to the predicted protective epitopes in response to vaccination (Figure 5D).

**Figure 4.** Heatmap of IgM and IgG autoantibodies overrepresented in vaccinated infected cohorts. According to the NSI-nor value, log2 (NSI+1) is calculated, the data are normalized, and a heat map is generated. The antigens are clustered according to Euclidean distance. Additional information is in Supplementary Materials Data File S3. Abbreviations: S, severe; L, non-severe: A, asymptomatic; C, PCR– (Table 1). **Figure 4.** Heatmap of IgM and IgG autoantibodies overrepresented in vaccinated infected cohorts. According to the NSI-nor value, log2 (NSI+1) is calculated, the data are normalized, and a heat map is generated. The antigens are clustered according to Euclidean distance. Additional information is in Supplementary Materials Data File S3. Abbreviations: S, severe; L, non-severe: A, asymptomatic; C, PCR– (Table 1).

The Igs identified as underrepresented in infected cohorts and thus with higher relative levels in PCR– individuals (Figure 5A, Supplementary Materials Data File S2) were used for the identification of vaccine-induced protective epitopes using a designed analytical workflow (Figure 5B, Supplementary Materials Data File S2). The results showed that the protective epitopes were not only identified in the SARS-CoV-2 spike (S) receptorbinding domain (RBD) associated with vaccine-protective capacity [15], but also in other virus proteins such as envelope small-membrane glycoprotein M (ORF3a), membraneprotein E, and nucleocapsid phosphoprotein N (ORF1ab) (Figure 5C, Supplementary Materials Data File S2). A correlation analysis was conducted between SARS-CoV-2-neutral-These results further advance our knowledge on the antibody response in vaccinated uninfected (fully protected) and vaccinated SARS-CoV-2-infected (partially protected) individuals associated with host factors such as age, comorbidities, and coronavirus infection [16–19]. Whereas vaccinated PCR– individuals developed a protective response mediated by Igs against multiple SARS-CoV-2 proteins to prevent infection, PCR+ individuals showed overrepresented Ig profiles associated with COVID-19 symptomatology with protective Igs to control virus infection and thrombosis in asymptomatic cases and limited or no protective response against SARS-CoV-2 with Ig-associated risk of allergy and other diseases in non-severe and severe patients.

izing antibodies (Table 1) and mass-spectra relative intensity of identified Igs with vaccine-induced protective epitopes (Supplementary Materials Data Files S1 and S2). The results showed no significant correlation for all cohorts together (R<sup>2</sup> = 0.182) but revealed a positive correlation in PCR– individuals in whom these proteins were overrepresented (R<sup>2</sup> = 0.826), thus providing support to the predicted protective epitopes in response to

vaccination (Figure 5D).

**Figure 5.** Serum-protein profiles of Ig underrepresented in vaccinated infected cohorts and identification of vaccine-induced protective epitopes. (**A**) Heatmap of PCR+/− Log fold-change relative intensity (Z-scored original value) for Ig proteins underrepresented in infected cohorts. (**B**) Analytical workflow developed for the dentification of vaccine-induced protective epitopes. (**C**) Identification of SARS-CoV-2 proteins with predicted reactive epitopes to Ig underrepresented in infected cohorts and thus overrepresented in PCR– individuals. All methods and results are disclosed in Supplementary Materials Data File S2. (**D**) Correlation analysis between SARS-CoV-2-neutralizing antibodies (Table 1) and mass-spectra relative intensity of identified Igs with vaccine-induced protective epitopes (Supplementary Materials Data Files S1 and S2). **Figure 5.** Serum-protein profiles of Ig underrepresented in vaccinated infected cohorts and identification of vaccine-induced protective epitopes. (**A**) Heatmap of PCR+/− Log fold-change relative intensity (Z-scored original value) for Ig proteins underrepresented in infected cohorts. (**B**) Analytical workflow developed for the dentification of vaccine-induced protective epitopes. (**C**) Identification of SARS-CoV-2 proteins with predicted reactive epitopes to Ig underrepresented in infected cohorts and thus overrepresented in PCR– individuals. All methods and results are disclosed in Supplementary Materials Data File S2. (**D**) Correlation analysis between SARS-CoV-2-neutralizing antibodies (Table 1) and mass-spectra relative intensity of identified Igs with vaccine-induced protective epitopes (Supplementary Materials Data Files S1 and S2).

#### These results further advance our knowledge on the antibody response in vaccinated uninfected (fully protected) and vaccinated SARS-CoV-2-infected (partially protected) in-*3.2. Characterization of Non-Ig Protein Profiles and Correlation with COVID-19*

dividuals associated with host factors such as age, comorbidities, and coronavirus infection [16–19]. Whereas vaccinated PCR– individuals developed a protective response mediated by Igs against multiple SARS-CoV-2 proteins to prevent infection, PCR+ individuals showed overrepresented Ig profiles associated with COVID-19 symptomatology with protective Igs to control virus infection and thrombosis in asymptomatic cases and limited or no protective response against SARS-CoV-2 with Ig-associated risk of allergy and other diseases in non-severe and severe patients. As reported in previous proteomics studies [7,8], identified dysregulated non-Ig proteins and biological processes in vaccinated infected PCR+ cohorts when compared to vaccinated PCR– individuals were associated with SARS-CoV-2 infection and COVID-19 (Figures 6A,B and 7, Supplementary Materials Data Files S1 and S4). As expected, PCR+/− Log fold-change relative intensity was higher in individuals with severe symptoms (Figure 6A,B). Accordingly, protein–protein-interaction networks and components for non-Ig proteins over and underrepresented in infected cohorts showed a higher representation in the severe cohort when compared to PCR– cases (Figure 7). Gene-ontology (GO) categories with overrepresented proteins involved in the regulation of complement and coagulation cascades and antibody-mediated complement activation were the most represented in protein–protein interactions (Figure 7; identified with yellow stars). Hyperactivation of

−1 0 1

Row Z−Score

*Molecules* **2021**, *26*, x FOR PEER REVIEW 13 of 17

*3.2. Characterization of Non-Ig Protein Profiles and Correlation with COVID-19*

the complement and coagulation systems are associated with the clinical syndrome of COVID-19 [20]. vation of the complement and coagulation systems are associated with the clinical syndrome of COVID-19 [20].

As reported in previous proteomics studies [7,8], identified dysregulated non-Ig proteins and biological processes in vaccinated infected PCR+ cohorts when compared to vaccinated PCR– individuals were associated with SARS-CoV-2 infection and COVID-19 (Figures 6A,B and 7, Supplementary Materials Data Files S1 and S4). As expected, PCR+/− Log fold-change relative intensity was higher in individuals with severe symptoms (Figure 6A,B). Accordingly, protein–protein-interaction networks and components for non-Ig proteins over and underrepresented in infected cohorts showed a higher representation in the severe cohort when compared to PCR– cases (Figure 7). Gene-ontology (GO) categories with overrepresented proteins involved in the regulation of complement and coagulation cascades and antibody-mediated complement activation were the most represented in protein–protein interactions (Figure 7; identified with yellow stars). Hyperacti-

**Figure 6.** Multiple differential representation and enrichment ontology clusters for non-Ig proteins. (**A**) Heatmap of PCR+/− Log fold-change relative intensity (Z-scored original value). (**B**) Enriched GO/KEGG ontology clusters for proteins over- and underrepresented in infected cohorts when compared to PCR– cases. Accumulative hypergeometric *p*-values and enrichment factors were calculated and used for filtering. Remaining significant terms were then hierarchically clustered into a tree based on Kappa-statistical similarities among their gene memberships. Then, the 0.3 kappa score was applied as the threshold to cast the tree into term clusters. The term with the best p-value within each cluster was selected as its representative term and displayed in a dendrogram.**Figure 6.** Multiple differential representation and enrichment ontology clusters for non-Ig proteins. (**A**) Heatmap of PCR+/− Log fold-change relative intensity (Z-scored original value). (**B**) Enriched GO/KEGG ontology clusters for proteins over- and underrepresented in infected cohorts when compared to PCR– cases. Accumulative hypergeometric *p*-values and enrichment factors were calculated and used for filtering. Remaining significant terms were then hierarchically clustered into a tree based on Kappa-statistical similarities among their gene memberships. Then, the 0.3 kappa score was applied as the threshold to cast the tree into term clusters. The term with the best *p*-value within each cluster was selected as its representative term and displayed in a dendrogram.

The other GO identified in the protein–protein-interaction network of overrepresented proteins was the insulin-like growth factor (IGF) pathway as seen in Figure 7. Although an association has been proposed between low IGF1 levels and poor outcome in patients with COVID-19 [21], an epidemiological study provided evidence that higher IGF-1 concentrations are associated with a lower risk of COVID-19 mortality [22]. The results of our study suggested that activation of the IGF pathway may occur in response to vaccination by regulating immune-cell homeostasis to reduce risk for COVID-19 mortality [23].

**Figure 7.** Protein–protein-interaction networks and components for non-Ig proteins over- and underrepresented in infected cohorts when compared to PCR– cases. The MCODE algorithm was applied to networks to identify neighborhoods where proteins are densely connected. GO enrichment analysis was applied to the original protein–protein-interaction network and its MCODE network components to extract their "biological meanings," where the top three best p-value terms were retained. GOs with proteins involved in the regulation of complement and coagulation cascades and antibody-mediated complement activation are identified with yellow stars. **Figure 7.** Protein–protein-interaction networks and components for non-Ig proteins over- and underrepresented in infected cohorts when compared to PCR– cases. The MCODE algorithm was applied to networks to identify neighborhoods where proteins are densely connected. GO enrichment analysis was applied to the original protein–protein-interaction network and its MCODE network components to extract their "biological meanings," where the top three best *p*-value terms were retained. GOs with proteins involved in the regulation of complement and coagulation cascades and antibody-mediated complement activation are identified with yellow stars.

The other GO identified in the protein–protein-interaction network of overrepresented proteins was the insulin-like growth factor (IGF) pathway as seen in Figure 7. Although an association has been proposed between low IGF1 levels and poor outcome in patients with COVID-19 [21], an epidemiological study provided evidence that higher IGF-1 concentrations are associated with a lower risk of COVID-19 mortality [22]. The results of our study suggested that activation of the IGF pathway may occur in response to vaccination by regulating immune-cell homeostasis to reduce risk for COVID-19 mortality [23]. Another finding of our study was related to severe-cohort overrepresented proteins in the biological process involved in interaction with symbiont (GO:0051702) (Figure 6B, Supplementary Materials Data File S4). One of the proteins identified in this biological Another finding of our study was related to severe-cohort overrepresented proteins in the biological process involved in interaction with symbiont (GO:0051702) (Figure 6B, Supplementary Materials Data File S4). One of the proteins identified in this biological process, Apolipoprotein E isoform 1 (APOE1; A0A0S2Z3D5), was overrepresented in severe (Log fold-change = 0.157) and UCI (Log fold-change = 0.081) patients (Supplementary Materials Data File S1). The expression of ApoE proteins, including APOE1, is critical for the assembly of infectious Hepatitis C virus (HCV) in a strain-specific and cell-type dependent manner [24]. Related to COVID-19, higher disease risk has been associated with *apoE4* genetic variants [25], but this is the first possible implication of ApoE1 in this process. Therefore, APOE1-protein levels and genetic variants may be a biomarker associated with disease severity in vaccinated and SARS-CoV-2-infected individuals.

process, Apolipoprotein E isoform 1 (APOE1; A0A0S2Z3D5), was overrepresented in severe (Log fold-change = 0.157) and UCI (Log fold-change = 0.081) patients (Supplementary Materials Data File S1). The expression of ApoE proteins, including APOE1, is critical for the assembly of infectious Hepatitis C virus (HCV) in a strain-specific and cell-type dependent manner [24]. Related to COVID-19, higher disease risk has been associated with As in recent studies [26], the interacting underrepresented proteins in vaccinated and infected cohorts were apolipoproteins APOA1, APOA2, APOA4, and APOC1 involved in the regulation of cholesterol esterification and phospholipid efflux (Figure 7). Higher levels of APOA1 have been correlated with protection from COVID-19 severity [27]. Furthermore, cholesterol esterification may counteract the normally exacerbating effect of cholesterol on coronavirus cytopathology [28]. Consequently, our results suggest that higher levels of some apolipoproteins in PCR– individuals may be associated with a vaccine-protective effect.

#### **4. Conclusions**

In summary, novel findings of the study include (a) characterization of Ig and non-Ig protein profiles in vaccinated uninfected (fully protected) and vaccinated SARS-CoV-2 infected (partially protected) individuals with identification of disease and protectionassociated biomarkers; (b) identification of candidate-protective epitopes not only in SARS-CoV-2 RBD but also in glycoprotein M (ORF3a), membrane protein E, and nucleocapsid phosphoprotein N (ORF1ab); (c) analysis of autoantibody profiles that are associated with an unfavorable COVID-19 disease course even after vaccination; and (d) prediction on non-Ig serum biomarkers associated with vaccine-protective capacity or disease severity in vaccinated and SARS-CoV-2-infected individuals.

The main limitation of this study is that serum-proteomics analysis was conducted with five samples from each cohort, which may have reduced the effect of case-by-case differences in serum-protein representation. Nevertheless, the results of this study using a serum-proteomics approach to characterize host-associated factors to COVID-19-vaccine response suggest protective- and disease-associated mechanisms in vaccinated individuals. Despite differences in individual age, sex, vaccine provider, and doses, the results were consistent between different cohorts. These results may lead to studies with a higher number of individuals and including different vaccine formulations to improve vaccine efficacy and implementation against SARS-CoV-2.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/molecules27185933/s1, Data File S1. Serum proteomics analysis; Data File S2. Analysis of immunoglobulin proteins underrepresented and overrepresented in infected cohorts when compared to PCR– individuals; Data File S3. Human-autoantibody general survey in PCR– and PCR+ cohorts; Data File S4. Analysis of selected non-immunoglobulin proteins underrepresented and overrepresented in infected cohorts when compared to PCR– individuals.

**Author Contributions:** Conceptualization, J.d.l.F., J.M.U., M.V., and C.G.; methodology, M.V., and J.d.l.F.; formal analysis, M.V., J.d.l.F., J.M.U., F.J.R.-d.-R., and J.d.l.F.; investigation, M.V., S.A.-J., M.C., R.V.-R., and L.M.; writing—original draft preparation, J.d.l.F.; writing—review and editing, J.d.l.F., J.M.U., M.V., and C.G.; visualization, J.d.l.F.; supervision, J.d.l.F., M.V., and C.G.; funding acquisition, J.d.l.F., and C.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Ministerio de Ciencia e Innovación/Agencia Estatal de Investigación MCIN/AEI/10.13039/501100011033, Spain, and EU-FEDER (Grant BIOGAL PID2020- 116761GB-I00); and the Junta de Comunidades de Castilla-La Mancha (JCCM), Spain, and EU-FEDER (grant MYCOTRAINING SBPLY/19/180501/000174).

**Institutional Review Board Statement:** The study was conducted in accordance with the Declaration of Helsinki, and approved by the Ethical and Scientific Committees (University Hospital of Ciudad Real C-352 and SESCAM C-73).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study. Informed consent has been obtained from the patient(s) to publish this paper.

**Data Availability Statement:** All data are available in the main text or the Supplementary Materials. The mass-spectrometry proteomics data have been deposited in the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD031969 and 10.6019/PXD031969.

**Acknowledgments:** We acknowledge the contribution of other members of our SaBio group on discussing some of the results of this study.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

**Sample Availability:** Samples of the compounds are available from the authors.

### **References**

