Immunology of SARS-CoV-2 infection and vaccination

Comprehensive elucidation of humoral immune responses to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and vaccination is critical for understanding coronavirus disease 2019 (COVID-19) pathogenesis in general and developing antibody-based diagnostic and therapeutic strategies specifically. Following the emergence of SARS-CoV-2, significant scientific research has been conducted worldwide using omics, sequencing and immunologic approaches. These studies have been critical to the successful development of vaccines. Here, the current understanding of SARS-CoV-2 immunogenic epitopes, humoral immunity to SARS-CoV-2 structural proteins and non-structural proteins, SARS-CoV-2-specific antibodies, and T-cell responses in convalescents and vaccinated individuals are reviewed. Additionally, we explore the integrated analysis of proteomic and metabolomic data to examine mechanisms of organ injury and identify potential biomarkers. Insight into the immunologic diagnosis of COVID-19 and improvements of laboratory methods are highlighted.


Introduction
Since 2022, the emergence of variants of concern (VOCs) in severe acute respiratory syndrome Corona virus 2 (SARS-CoV-2) remains problematic in managing this outbreak worldwide.These VOCs include Omicron which exhibited higher transmissibility leading to increased disease severity despite the global vaccine roll-out and introduction of public health measures to slow their spread [1][2][3].
Currently, there are major concerns regarding Omicron including the efficiency and duration of protection that is offered by the currently available immunotherapeutic agents and vaccines [3,4], the possibility of reinfection, and unknown organ damage. To address these concerns, scientists have clarified the immunology of SARS-CoV-2 infection and vaccination and developed new strategies to counter the Omicron variant.
Protein-or peptide-based microarrays are important tools for epitope mapping and profiling of antibodies against the SPs and NSPs of SARS-CoV-2 [8]. These tools can help elucidate the human immune response to SARS-CoV-2, thereby facilitating the identification of SARS-CoV-2 epitopes and design of vaccines and therapeutic antibodies [9][10][11][12].
Six types of SARS-CoV-2 vaccines have been granted emergency use, including inactivated and live virus, protein subunit, adenovirus (Ad) vector, DNA, and mRNA vaccines [13,14]. As of April 23, 2023, more than 13 billion vaccine doses have been administered and reported to the World Health Organization (WHO) [15]; these vaccines provide 65-95% protection against COVID-19 [4,8,13,14,16,17]. Tackling the emerging SARS-CoV-2 variants, such as Omicron, that cause vaccine breakthrough infections in vaccinated and recovered individuals need to be addressed [4]. Many research projects have explored the efficiency of vaccines by analyzing the dynamic serological alterations in total antibodies, neutralizing antibodies (NAbs), and cellular immune responses to SARS-CoV-2 infection and vaccination [18][19][20][21][22]. Determining the effectiveness of a vaccine helps to improve our understanding of the role of NSPs and humoral immunity during SARS-CoV-2 infection as well as the immunology of the disease. However, there is difficult to make a thorough knowledge in B-and T-cell immunity for SARS-CoV-2.
Omics approaches have a growing influence on precision medicine as they contribute to the elucidation of the role of humoral immunity in SARS-CoV-2 infection and vaccination. Systematic screening of proteins, metabolites, and lipids has uncovered the mechanism of aberrant regulation of physiological processes, including the complement system activation, macrophage function, and platelet degranulation, in COVID-19 [22][23][24] and facilitated biomarker identification and development of effective therapies. Biological processes involved in the innate and adaptive immune responses, including the immune system, T-cell receptor (TCR) signaling, B-cell receptor signaling, and interleukin signaling, are enriched in COVID-19 [25][26][27] and may play a role in immune-mediated organ injury. A hypothetical model of immune dysregulation and increased reactive oxygen species (ROS) generation, which induces renal injury in patients with severe COVID-19, has been proposed [22,25]. Further, integrated research on omics data and developing diagnostic models and computer-aided detection systems can improve the practicability of immunological research on COVID-19 [26,27].
This review provides an overview of the immunology of SARS-CoV-2 infection and vaccination, focusing on the basic and clinical scientific findings, vaccine efficiency, and dynamics of B-and T-cell immune responses in patients over time. In addition, we summarize the role of humoral immunity in SARS-CoV-2 infection and vaccination based on omics approaches such as proteomics and metabolomics, Fig. 1. The purpose of our review is to address the importance of systematically studying viral immunity to develop new diagnostic biomarkers and applying this knowledge to other potential infectious and autoimmune diseases.

Mapping epitopes and NAbs of SARS-CoV-2
The B-and T-cell epitopes of SARS-CoV-2 have been predicted using bioinformatics or measured using T cell-based assays [28][29][30][31]. Currently, popular methods for mapping epitopes and NAbs of SARS-CoV-2 are based on full-protein or peptide array platforms and the VirScan phagedisplay platform. Here, we review the results of epitope mapping based on these platforms.

Fig. 1.
Overview of the scientific findings on the humoral and cellular immune responses to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and vaccination based on omics approaches and their contribution to improving the quality of laboratory methods for laboratory medicine we reviewed in the different article sections. First, we discussed mapping epitopes and neutralizing antibodies of SARS-CoV-2. Second, we reviewed the dynamics of the antibody and T cell responses to SARS-CoV-2 proteins. Third, we discussed the immunological evaluation of coronavirus disease 2019 (COVID- 19) vaccines. Finally, we discussed the profiling of immune responses in COVID-19 based on big data from proteomic and metabolomic studies.
VirScan, reported by Elledge in 2015 [35], is a high-throughput method for analyzing antiviral antibodies using immunoprecipitation and massive parallel DNA sequencing of a bacteriophage library; this method has been applied in the mapping of SARS-CoV-2 epitopes. Shrock et al. [36] identified 823 epitopes across the entire SARS-CoV-2 proteome, including 10 that were mostly derived from the S and N proteins likely recognized by NAbs. Zamecnik et al. [37] employed ReScan, which pans COVID-19 sera, to generate a focused SARS-CoV-2 antigen microarray that mapped nine antigens to specific regions of the N and S proteins and a single ORF3a peptide (residues 172-209). These antigens showed improved specificity over the whole protein panel and were validated using a luminex-based assay. Notably, these peptides were found in areas of low relative mutational diversity, indicating their potential for use in diagnostic settings.
High titers of S-specific antibodies have been reported in patients with COVID-19 [38,39]. The RBD is the most targeted region for the development of COVID-19 therapeutic antibodies [40,41]. Whether antibodies against NSPs provide protection or are associated with the survival of COVID-19 patients remains to be determined [42]. Li et al. [43] characterized three epitopes, S430-451, S470-481, and S501-515, in the RBD of the S protein and showed that the monoclonal antibodies from S470-486-immunized mice could inhibit SARS-CoV-2 replication. Furthermore, Tao et al. [34] discovered three hot epitope areas across the S protein, and six representative epitopes in areas/ epitopes of the S protein, instead of the RBD, were chosen as potential linear epitopes inducing NAbs. Surprisingly, antibodies against three epitopes, S553-564, S613-624, and S1148-1159, exhibited potent neutralizing activity (NA) in the pseudo-virus neutralization assay.
Using ELISpot, Yuan et al. [23] found that three epitopes of SARS-CoV-2, S375-394, S405-469, and S495-521, stimulated robust secretion of IFN-γ. The response of CD4 + T-cell epitopes in the RBD showed distinct patterns compared to those of B cells. Among the nine epitopes on the S protein, S370-394, S450-469, and S480-499 were identified as linear B-cell epitopes. These data suggest that the S370-394, S450-469, and S480-499 epitopes were more likely to serve as both T-and B-cell linear ID sites.
In summary, the abovementioned studies mapping the epitopes of SARS-CoV-2 provide serological evidence for developing broadspectrum antibodies, diagnostic tests, and vaccines against SARS-CoV-2 variants.

Broad-spectrum NAbs
In theory, broad-spectrum coronavirus NAbs can withstand viral mutations and can potentially be utilized in future campaigns against various coronavirus outbreaks [44].
When the mutant strain Omicron appeared [45], there was an urgent Table 1 The epitopes of SARS-CoV-2 discovered by proteomics. Mice immunized with the S protein of an original SARS-CoV-2 Serum of mice 9 S S3H3 On the subdomain1 (SD1) region of S trimer demand for the development of a new generation of COVID-19 vaccines and broad-spectrum NAbs that could provide broad-spectrum protection. In September 2021, Xia et al. [46] reported the broad-spectrum NAbs 7D6 and 6D6, the RBD mutation sites of the popular mutants Alpha, Beta, Gamma, Delta, Kappa, Iota, and Lambda were all located outside of the epitope targed by 7D6/6D6. Thus, 7D6/6D6 may be a good target for designing broad-spectrum vaccines. On January 26, 2022, a similar broad-spectrum antibody, S3H3, that targeted the SD1 region was reported by Xu et al. [47]. SD1 is located downstream of the RBD, similar to the switch in RBD. After the binding of S3H3, the RBD of the S-trimer showed a less open conformation, which interfered with the binding between the RBD and ACE2, thus blocking the virus from entering the cells. No mutation site was found in the binding region of S3H3, indicating a broad-spectrum epitope for NAbs.

Bispecific antibodies (bsAbs)
The term bsAb is used to describe a large family of molecules that are designed to recognize two different epitopes or antigens that differ from those of antibodies that target a single epitope [48]. bsAb creation has been important for the development of COVID-19 vaccines [10,49]. bsAbs can bind two highly conserved regions of the Omicron variant and, therefore, may improve the effectiveness of SARS-CoV-2 vaccines and therapeutic antibodies and overcome the limitation of continuous emergence of viral variants [50].
In September 2022, Ku et al. [49] engineered two bsAbs based on the COV2-06 and COV2-14 NAbs. They showed that bsAbs with the IgG-(scFv)2 design, but not the CrossMAb design, increased antigen binding and the virus-neutralizing capacity against multiple SARS-CoV-2 variants compared to the antibody cocktail. In October 2022, Wang et al. [11] constructed eight broad-spectrum bispecific NAbs using non-Omicron NAbs. The mechanism of action of the bsAb FD01 involved the synergistic induction of Omicron S protein-induced "trimer dimer" formation of the bsAbs with broad-spectrum inhibition ability against a variety of COVID-19 VOCs, including Omicron BA.1 and BA.2., SARS-Co-V, and other Sabe viruses. Therefore, rationally engineered bsAbs represent a cost-effective alternative to antibody cocktails and a promising strategy for improving broad-spectrum functionality.

Antibodies against SARS-CoV-2
Patients with COVID-19 exhibit features such as the extreme variability of clinical disease severity from asymptomatic to fatal [36] and the varied concentrations of nucleic acids, IgM, and IgG [51]; hence, strategies combining molecular and serological assays can be used to reveal the infection status and stage. The relation of antibodies to SPs is one of the main tools used for the diagnosis of COVID-19; consequently, many studies have revealed the humoral response to SPs, as listed in Section 2 [9,10,26,27,33,34,[38][39][40][41]43,46,47]. Antibodies of S, N, and accessory (ORF3b, ORF8) proteins of SARS-CoV-2 were elicited in patients with severe disease, demonstrating the association between disease severity and humoral immune responses [29,52,53]. Here, we explain the prevalence, clinical relevance, and dynamics of antibodies against NSPs and accessory proteins in patients from two representative studies [54,55].
Furthermore, high correlations of IgG responses to these NSPs and accessory proteins were revealed, indicating that the IgG responses against NSPs and accessory proteins show the same trend. However, there was no clear difference in the IgG responses to S1 and N, revealing a high association of NSPs/accessory proteins with disease progression. This suggests that the underlying mechanisms of NSP-, accessory protein-, and SP-triggered antibodies are different. IgG antibodies against NSPs/accessory proteins rapidly decline when they reach a plateau approximately 20 days after symptom onset compared with those against S and N proteins, which might be associated with a decrease in the protein expression level when viral replication is inhibited.
These results prove that the mechanisms by which the SPs and NSPs/ accessory proteins of SARS-CoV-2 elicit host humoral immune responses may be different, implying that B cells that produce IgG antibodies against NSPs/accessory proteins might be short-lived and the underlying mechanism may differ from that of generating IgG antibodies against S1 and N proteins.
In critically ill COVID-19 patients, Cheng et al. [55] mapped the NSPs generated from Orf1ab, which has the largest number of epitopes targeted by both IgM (84.7%) and IgG (58.0%). IgM targets more epitopes to NSPs than IgG. IgM is associated with a good prognosis and targets NSP3 and NSP5 proteases, whereas IgG antibodies are associated with high mortality and target SPs (N, S, ORF3a, ORF8). These data are inconsistent with the findings of Li et al. [43], who found that NSPs can strongly elicit IgG. The dynamics of IgG antibodies targeting NSPs/ accessory proteins (N, S, ORF3a) with clinical outcomes were consistent between the two research teams. Because critical COVID-19 patients were the target individuals to explore the immunity against SARS-CoV-2, the contradictory results confirmed the same idea as previous studies: the immune responses in critical COVID-19 patients might differ from those of patients classified as asymptomatic, mild, moderate, or severe. Altogether, these results indicate that the severity of the illness in the selected cohort may be a critical factor in determining the immune response to SARS-CoV-2.
The studies discussed above indicated that antibodies against NSPs could be associated with the survival of COVID-19 patients; thus, the immune system may protect against COVID-19 by generating antibodies against NSPs. One concern about antibodies against the S protein is the possible antibody-dependent enhancement [56], which causes uncontrolled release of proinflammatory cytokines such as IL-1, IL-6, IFN-γ, and TNF-α [57,58]. The severity of COVID-19 is highly associated with cytokine release syndrome or cytokine storms [58,59]. Antibodies against NSPs/accessory proteins may also trigger the production of cytokines when they bind the released antigens from the infected cells. These data help us understand the role of NSPs in SARS-CoV-2 infection and increase our understanding of humoral immunity during SARS-CoV-2 infection.

Dynamics of the antibody and T-cell responses to SARS-CoV-2 over time
Research regarding antibodies and T-cell responses to SARS-CoV-2 over time with follow-up cohorts has been reported [60]. A crosssectional study [60] conducted in China between April and December 2020 reported that NAb titers were lower in asymptomatic individuals than those in confirmed cases of COVID-19 as well as symptomatic individuals.
Although the antibody titers against SARS-CoV-2 decline over time after viral clearance, SARS-CoV-2-specific cellular immune responses remain detectable even eight months after infection [61][62][63]. T-cell immunity plays a central role in the control of SARS-CoV-2 infection, and its importance may have been relatively underestimated thus far [62,63]. Studies have shown that antigen-activated T cells and secreted cytokines may have more important effects on B cells than circulating antibodies [64].
Studies have focused on the protective role of cellular immunity against severe acute infection and reinfection as well as VOCs [18,50]. In acute SARS-CoV-2 infection, a cytotoxic CD8 + T-cell response is typically observed within 7 days and peaks at 14 days after the onset of symptoms, the duration is correlated with effective viral clearance [65,66].
Lymphocytopenia of peripheral blood is seen in acute SARS-CoV-2 [67] and may reflect impaired lymphocyte proliferation, apoptosis [65], or extravasation into tissue; resolution of lymphocytopenia correlates with recovery but can take several weeks [68].
The subsets and expressed molecules of T cells are also key determinants of the clinical outcome. The type 1 CD4 + phenotype is associated with effective viral control [65]. High expression levels of effector molecules in CD8 + T cells in acute COVID-19 may reflect an improved clinical outcome in patients [69].
T-cell memory against SARS-CoV-2 was discovered after initial infection; Wang et al. [12]  Both antibody and T-cell responses to D614G, Beta, and Delta viral strains were evaluated. The degree of reduced in vitro NAb response to the D614G and Delta variants was associated with NAb titers after SARS-CoV-2 infection [12].
IFN-γ-, IL-2-, and TNF-α-specific T-cell responses were cross-reactive to the Beta variant in most individuals [12]. Importantly, T-cell responses could be detected in all individuals who had lost the NAb response to SARS-CoV-2, 12 months after the initial infection. These findings indicate that SARS-CoV-2-specific cellular immunity decays slowly over time, memory T-cell responses to the original strain were not disrupted by the new variants.
These findings suggests that cross-reactive SARS-CoV-2-specific Tcell responses could be particularly important in protection against severe disease caused by VOCs, whereas NAb responses seem to decrease over time.
Global profiling of the antibodies and cellular immunity to SPs and NSPs of SARS-CoV-2 may imply a possible protective function, facilitate a deeper understanding of immunology of SARS-CoV-2 infection, and reveal biomarkers of disease progression, highlighting the importance of broad B-and T-cell immunity for future vaccine strategies targeting SARS-CoV-2.

Immunological evaluation of COVID-19 vaccines
The protection through homologous and heterologous immunization against COVID-19 has attracted significant attention, as the Omicron variant has unprecedented transmission speed. However, the cumulative protective effect of prior infection and vaccination also needs to be determined.

Effectiveness of homologous and heterologous immunization
Evidence has shown that heterologous immunization induces greater B-and T-cell immunity than homologous immunization ( Table 2). Mileto et al. [70] revealed that one month after the second dose of the BNT162b2 vaccine, NA against the Omicron variant reduced 7.3-and 3.15-fold compared with the EU and Delta variants, respectively.
After the third dose of BNT162b2, the anti-S IgG titer and NA increased by 75% and 3.03-fold, respectively. After booster immunization with an inactivated vaccine [71], the positive NA was 95.5% for the Omicron variant (B.1.1.529), 99.5% for the prototype, and 98.5% for the Delta variant. NA for the Omicron variant decreased by 4.9-and 3.0-fold compared with the prototype and Delta variants, respectively.
A heterologous primary immunization-booster strategy with a recombinant subunit, adenoviral vaccine, or mRNA vaccine presented better NA against the Omicron variant than homologous booster strategy [72][73][74]. Wang et al. [72] revealed that the heterologous RBD recombinant subunit vaccine Zifivax increased NAb titers 3.5-to 6.8-fold with a broader NA against six VOCs, including Omicron. Furthermore, the NAb titer, total anti-S + N or anti-RBD fractions, and the RBDspecific CD4 + T-cell population in the heterologous boosting group were higher than those in the homologous boosting group at 14 days post-vaccination, indicating that heterologous boosting targeted neutralizing epitopes more precisely than homologous boosting. The IFN-γ + and TFN-α + CD4 + T-cell population was found to be significantly greater after 3 months post-vaccination than that at day 0 but did not differ from that at day 14 and 1 month, indicating the maintenance of the T-cell response after RBD subunit booster vaccination.
In addition, Clemens et al. [73] reported that after heterologous immunization, the level of anti-S IgG greatly increased: 77-fold for Ad26, 152-fold for BNT162b2, 90-fold for ChAd, and 12-fold for Coro-naVac. All heterologous regimens showed anti-S IgG responses on day 28, which were superior to homologous booster responses.
Moreover, a phase 2 trial in the UK conducted by Munro et al. [74] reported that a booster vaccine increased the reactogenicity of pseudovirus neutralization, and T-cell assays against SARS-CoV-2: BNT162b2 after ChAd/ChAd or BNT/BNT and ChAd and Ad26 after BNT/BNT were emphasized.
The world is currently facing the highly transmissible Omicron variant; however, whether there is a need to rush the deployment of a fourth vaccination strategy remains controversial. Yang et al. [75] found that anti-SARS-CoV-2 antibody titers were significantly high at 9 months Table 2 Researaches of effectiveness about homologous and heterologous SARS-CoV-2 vaccine immunization.

References
Basic immunization Boosting immunization NAb after boosting vaccine Both strategies induced 3.5-fold to 6.8-fold higher NAb titers than homologous, with a broader neutralizing capacity against six variants of concern, including Omicron. after the third dose of vaccine and decayed slowly; therefore, the third dose of the CoronaVac inactivated vaccine could quickly evoke immune memory. However, a report on booster doses indicated that the fourth dose of the original vaccines may not generate robust protection against infection in the form of NAbs against Omicron [75]. Thus, Yang et al. [76] suggested that new vaccines may be more effective to combat the new variants. The combination of existing and new vaccines may provide an immunoprotective effect.

. Protection with vaccine superimposed on previous infection
A cohort study conducted when the Delta variant predominated [77] revealed that prior infection significantly boosted anti-S antibodies, antibody-dependent cellular cytotoxicity, and NAbs against D614G, Beta, and Delta variants; however, neutralization cross-reactivity varied by wave. After vaccination with Ad26.COV2, robust CD4 + and CD8 + T-cell responses were induced. T-cell recognition of variants was largely preserved; however, slight reduction in CD8 + T-cell recognition of the Delta variant was observed. After receiving the COVID-19 vaccine, people who were previously infected with SARS-CoV-2 had a 72% reduced risk of reinfection [78]. The greater the number of immunizations, the higher the level and broader the spectrum of NAb responses triggered by the body.

. Vaccine protection over time
NAbs were observed to decrease with time in recovered COVID-19 patients who were vaccinated one year after recovery [79,80]; this finding suggests the risk of reinfection with SARS-CoV-2 variants. The cellular basis of the protection provided by the COVID-19 vaccine needs to be further elucidated.
On October 22, 2022, Zhang et al. [81] revealed that a heterologous booster induced a faster and more robust plasma response characterized by the activation of plasma cells than the homologous booster, based on single-cell transcriptomes and single-cell B-cell receptor sequencing. The production of antibodies was positively correlated with antigen presentation by conventional dendritic cells, which provides support for Bcell maturation through the activation and development of follicular helper T cells.
From the studies we reviewed, we found that heterologous boosting strategies could boost NAbs against SARS-CoV-2 as well as VOCs. Different vaccination regimens induce broad and long-lasting spikespecific CD4 + and CD8 + T-cell immunity to SARS-CoV-2. These data provide important evidence for vaccination strategies based on available vaccines and may guide the development of future global vaccination plans.

Proteomic and metabolomic profiling of immune responses in COVID-19
The immune response evoked post-infection has been a major area of interest to scientists. Mass spectrometry (MS)-based proteomics and metabolomics studies have been performed on a diverse range of clinical sample types since the COVID-19 pandemic emerged, including plasma, sera, nasopharyngeal swabs, gargle solutions, urine, and postmortem tissues [21,82], making them accessible and desirable biospecimens for biomarker search. These studies can help explain the innate and adaptive immune response features.

Role of complement proteins and inflammation in COVID-19
The proteome of patient sera explicitly implicates overactivation of the complement system. Shen et al. [26] reported an upregulation of the complement system proteins C5 and C6 and their regulators properdin and carboxypeptidase N catalytic chain in patients with COVID-19. A study by Messner et al. [81] on 31 hospitalized COVID-19 patients also reported activation of the classical complement pathway, C1r, C1s, C8A, alternative pathway factor B, and complement modulators such as CFI and CFH. Georg et al. [83] described C3a-driven induction of activated CD16-expressing T cells in severe COVID-19. These T cells display increased immune complex-mediated TCR-independent cytotoxicity, causing the activation and release of chemokines by lung endothelial cells. Thus, complement system proteins are effector components of both innate and adaptive immunity and a COVID-19 hallmark. Inflammation is another attribute of the activation of the innate immune system. Messner et al. [82] reported the upregulation of serum amyloid A1 and A2, the monocyte differentiation antigen CD14, and other factors implicated in IL-6 signaling.
Proteomics of peripheral blood mononuclear cells [84] also demonstrated a clear upregulation of CXCR2, PRG3, LBP, MMP25, CRP, and NLRP1, indicating the activation of the IL-6/IL-8-directed innate immunity inflammation pathway in mild cases of COVID-19. Proteomics of COVID-19 serum stratified by rising levels of IL-6 revealed upregulated levels of the complement factors HI and C5, implicating complement system activation, which positively correlated with enhanced IL-6 levels [52,85].

Mechanism of antiviral immune evasion of SARS-CoV-2
Few in vitro studies have revealed the mechanism of evasion of the immune system by the interactome between viral proteins and human cell lines. Biotin-streptavidin-based affinity MS studies showed that the viral protein ORF9b physically interacts with the mitochondrial protein TOM70 to inhibit IFN-I responses of the host [86], indicating that the virus can evade the immune system.
Thus, SARS-CoV-2 infection may induce dysregulation of signaling pathways, and the virus may evade the immune and complement systems and cause inflammation (IL-6 signaling); these facts must be considered while studying innate immune responses during viral infection.

Developing proteomic and metabolomic biomarkers of COVID-19
Studies have adopted individual or multi-omics approaches to understand the pathobiology of COVID-19. Breakthroughs based on multiomics and the integrated analysis of multi-omics data in conjunction with machine learning algorithms have helped explore the complexity and immunogenicity of COVID-19, aiding the development of feasible diagnostic indicators and prediction models for disease severity [88].
Rahnavard et al. [87] developed a deep neural network model with an accuracy score of 81.78% by combining proteomic and metabolomic data to predict disease severity. This may reflect aberrant levels of basic organic molecules that influence viral replication and immune responses, such as cytosine and branched-chain amino acids. The COVID-19 Multi-omics Blood ATlas Consortium [88] presented a comprehensive multi-omics blood atlas for developing immune signatures as hallmarks of COVID-19 severity. A set of cross-modality features to predict severity included the acute-phase proteins serum amyloid A2 and CRP, an immunoglobulin (IGHG4), chemokines (CCL20 and CCL2), IL-6, and C5a, and the combined performance of these features in the hold-out validation set showed a balanced accuracy of 75-80% to predict the WHO category group.
Gao et al. [27] performed machine learning based on three datasets, the model led to the effective classification of severe cases with an area under the curve value of 0.965, outperforming the models with proteins only.
A meta-analysis conducted by Aggarwal et al. [21] provided an integrated view of perturbations in host responses at both the proteome and metabolome levels. Significantly altered proteins (n = 88) and metabolites (n = 40) were selected in the COVID-19-positive group. Four enriched pathways were affected: the immune system, TCR signaling, Bcell receptor signaling, and signaling by interleukins. These data provide a better understanding of the role of inflammation and alterations in the levels of IL-6, cytokines, lipids, and amino acids in disease progression.
All developments in the area of omics have not been translated into practical applications for disease diagnosis; further experiments should be conducted to profile the diagnostic responses in cytokine, protein, and metabolite levels.

Immune-mediated organ injury in COVID-19
Multi-organ damage in patients with COVID-19 contributes to immune responses, such as cytokine storms, macrophage recruitment from peripheral blood to the lungs that leads to alveolar macrophage infiltration, lung damage, and respiratory failure [89,90]. Guo et al. [26] used multi-omics data to reveal inflammation-induced renal injury. The study demonstrated the enrichment of 20 immunity-related pathways prominent in both serum and urine, supporting the hypothesis that immune dysregulation and increased ROS levels induce renal injury in patients with severe COVID-19.
On the one hand, some significantly changed metabolites in the urine of COVID-19 patients [26], such as downregulated N-acetylcysteine and upregulated quinolinate, activated ROS production, which could lead to a variety of immune-mediated tissue injuries in patients with COVID-19. On the other hand, inflammation can induce renal injury. The inflammation-involved pathway and leukocyte extravasation signaling pathway enriched significantly among the 23 enriched serum pathways in this study. Inflammation may initiate coagulation, an innate immune process. F5 and F9 are key proteins in coagulation [90] that are both upregulated in the sera of patients with COVID-19, coagulation may also activate downstream fibrosis [22,91,92]. Thus, multi-omics data have revealed renal injuries in patients with severe COVID-19.

Concluding remarks
Molecular tests, such as PCR, rapid antigen detection, and antibody tests, play a significant role in diagnosing COVID-19 at different stages of epidemic prevention and control [93]. Clinical laboratories are expected to provide tests to assess the disease severity of patients with COVID-19 considering that SARS-CoV-2 invades the lungs, heart, and kidneys, disturbing the function of the immune system, especially in patients with higher disease severity awaiting clinical decision-making.
The data presented herein provide methods to evaluate the possibility of reinfection and considerable guidance for developing COVID-19 diagnostic strategies as well as managing future pandemic threats.
Lastly, the immune trajectory of SARS-CoV-2 infection and vaccination reviewed herein may motivate researchers to conduct studies focusing on the clinical transformation application of these diagnostic markers. The use of artificial intelligence algorithms for processing multidimensional data would play an important role in laboratory medicine.

Ethics approval and consent to participate
Not applicable.

Consent for publication
All the authors consented for publication.

Availability of data and material
Not applicable.

Funding
This review was funded by The National Key Research and Development Program of China (2018YFE0207300), "345 Talent Project" of Shengjing Hospital of China Medical University (2022146), Emergency Scientific Research Project for Prevention and Control of COIVD-19 of Liaoning Province (4th batch), Development of Emergency Intelligent Control and Decision Support System for Major Infectious Diseases of Liaoning Province Science and Technology Plan Project (2021JH2/ 10300136). There was no participation of funders in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability
No data was used for the research described in the article.