*2.1. The Blood Sera Study of COVID-19 Convalescents of Varying Severity with a Favorable Outcome*

During our study, it turned out that in mild cases, the median values of titers of IgG specific to the inactivated SARS-CoV-2 antigen were in the range of 1:200–1:12,800, with the most common values being 1:800 and 1:3200. High titers were found in an asymptomatic patient as well as in the case of a patient who recovered after a moderate case of COVID-19 (with unilateral pneumonia)—1:6400 (Numbers 1 and 15 on the Appendix A Figure A1). Three more recovered cases with bilateral pneumonia, or those who were treated in the intensive care unit, showed the highest titers—1:12,800, in the neutralization test with the inactivated SARS-CoV-2 antigen (Appendix A Figure A1).

The cross-activity of antibodies specific to SARS-CoV-2 with the inactivated heterologous SARS-CoV (strain Frankfurt 1, 2002) antigen (the third row on the Appendix A Figure A1, marked in green) increased in accordance with the titer found on the SARS-CoV-2 homologous antigen.

The use of recombinant SARS-CoV-2 proteins in ELISA allowed us to determine the individual profile of antibodies specific to individual viral proteins. According to ELISA data in median values (Appendix A Figure A1) and the mean ± SD of titers of specific antibodies (Table S1), the antibodies in this panel of convalescents' blood sera, from mild to severe cases in specificity and quantitatively, are mainly distributed in the following descending order: N > S trimer > RBD.

Higher IgG levels with median values in the range of 1:400 to 1:3200, detected in ELISA using the recombinant RBD protein (Appendix A Figure A1), contributed to more effective protection of infected Vero cells from the development of a cytopathic effect by 50%—from 1:20 to 1:160 in mild cases (Numbers 3, 5, 6–14), and from 1:80 to 1:≥320 in cases of moderate severity (Numbers 15, 16) and seriously ill patients (Numbers 18–32). In an asymptomatic patient (Number 1, age 32), the titer of protective antibodies was 1:80 (data on neutralization) (Table S2).

In the presence of paired sera from patients with severe cases of COVID-19 (Numbers 17–28), an increase in IgG titers specific to RBD in ELISA was observed (Appendix A Figure A1).

Obviously, the presence of antibodies specific to RBD correlates with the data on the neutralizing activity of blood sera. So, for example, the 1:200 titer of IgG specific to this site of the S glycoprotein detected in ELISA either did not contribute to protection in a severe case (Numbers 17 and 23) or corresponded to the presence of neutralizing antibodies in a titer of 1:10–1:20 in the mild cases of Numbers 2 and 4 (Table S2). In patients with a severe course of the disease from whom only single serum samples were studied in the neutralization reaction (Numbers 29–32), the median titers of protective antibodies turned out to be 1:160. In the presence of paired blood sera in patients with a severe course of the disease, a tendency towards a sharp increase in the titers of protective antibodies was revealed within a few days from complete absence to 1:160 (Numbers 23, 24) or to 1:≥320 (Numbers 17, 18). In other patients with a severe course of the disease, the titers of protective antibodies also steadily increased—for example, from 1:160 to 1:≥320 (Numbers 19, 20 and 27, 28) and from 1:80 to 1:160 (Numbers 21, 22 and 25, 26) over the course of several days, respectively (Table S2).

The immunoblotting of the SARS-CoV-2/human/RUS/Nsk-FRCFTM-1/2020 antigen with the blood serum of an asymptomatic patient, as well as those with mild and moderate cases, made it possible to detect viral target proteins for specific antibodies. Moreover, it was also possible to visually determine the efficiency/intensity of the detection of these viral proteins in this enzyme immunoassay on the nitrocellulose membrane with antibodies used for testing the same dilution of blood serum—1:100. It was clearly demonstrated that the N nucleocapsid protein of SARS-CoV-2 is the main protein target for antibodies in mild cases. Numbers 1–32 include the antibodies of asymptomatic (Number 1), mild (Numbers 2–14), and moderate patients (Numbers 15 and 16), as well as patients from the intensive care unit (Numbers 1–32) (Appendix A Figure A2A).

The antibodies specific to linear epitopes of the S-protein were not detected in all blood sera. For example, the SARS-CoV-2 S-protein dimer and trimer were detected in the serum of an asymptomatic patient (Number 1) at 180 and 270 kDa, respectively. Additionally, among mild cases of the disease (Numbers 2–14), possibly due to background levels, there are no clear bands of the S protein. Antibodies were also detected in those with moderate cases of COVID-19 (Numbers 15 and 16). In severe cases of the disease (intensive care unit) (Numbers 17–32), antibodies specific to linear epitopes of the S-protein were detected in the sera (Numbers 24–32). These findings strongly demonstrate that the intensity of the immune response significantly increased in paired sera—Numbers 23–24, 25–26, and 27–28.

The SARS-CoV (2002) nucleoprotein mostly turned out to be the only target protein for cross-interaction of antibodies specific to SARS-CoV-2 in asymptomatic, mild, moderate (Numbers 1–16), and severe cases (Numbers 17–32).

In some cases, the antibodies of some blood sera detected the M protein SARS-CoV-2 (Appendix A Figure A2A). The M protein SARS-CoV (2002) was also the target of weak cross-interaction for these antibodies (Appendix A Figure A2B).

#### *2.2. The Blood Sera Study of Patients with a Fatal Outcome of Disease*

In our study, we used sera from 13 patients with severe cases of COVID-19 with a lethal outcome in the hospital. It was shown that according to the ELISA results, the median values of specific IgG titers in the sera of these patients ranged from 1:200 to 1:25,600 (Table S3).

In the case of the paired blood sera, just as in the case of convalescents, in patients who subsequently died in the hospital, an increase in antibody titers was observed when using the inactivated SARS-CoV-2 antigen. This is especially evident in the example of three blood sera from one patient with an increase in IgG titers in ELISA from 1:6400 to 1:12,800 and 1:25,600 in blood sera for Numbers 39, 40, and 41, taken at the 15th, 10th, and 7th days before death, respectively. Moreover, these blood sera contain antibodies specific to RBD in titers of 1:1600, 1:3200, and 1:6400, corresponding to the neutralizing antibodies in the range of 1:160 to ≥1:320 a week before death (Table S3). The neutralizing antibodies at the median values of titers ≥1:320 were also found in single blood sera from two patients (Numbers 33 and 35), who subsequently died a day after the blood was taken. The sufficiently high serological titers on the inactivated antigen (1:25,600 and 1:12,800) and on the recombinant RBD (1: 25,600 and 1:3200) in ELISA were also shown (Appendix A Figure A3).

The antibodies of the blood sera of 13 patients that died from severe COVID-19 in the hospital actively cross-interact with the SARS-CoV antigen (2002) in ELISA (Appendix A Figure A3). A dependence on the value of titers with a homologous antigen was also shown, i.e., a higher titer of antibodies to SARS-CoV-2, causing a higher value of cross-activity.

According to immunoblotting data (Appendix A Figure A4A), the N protein of SARS-CoV-2 is detected by antibodies of all patients in the intensive care unit (who later died in the hospital), except for the blood serum Number 36. The presence of paired blood sera taken at different times before the death of patients makes it possible to clearly demonstrate the results of ELISA (Appendix A Figure A3) on the increase in antibody titers and the presence of antibodies specific to the linear epitopes of the S protein (dimer and trimer), and, in some cases, to the M protein. In one case, antibodies specific to the linear epitopes of the S protein monomer (90 kDa) were observed (blood sera Numbers 44 and 45). The same antibodies were also cross-reactive with analogous epitopes of the S protein monomers of SARS-CoV (2002) (Appendix A Figure A4B).

As shown in Appendix A Figure A4, in many patients in the intensive care unit (who later died in the hospital), cross-reactive antibodies were specific to both the N protein and to the S protein trimer. In serum Number 36, there were antibodies that detected linear epitopes of the S protein on the SARS-CoV antigen (2002) that did not manifest themselves in the neutralization reaction with SARS-CoV- 2 and did not reveal any bands in the SARS-CoV-2 immunoblotting.

### **3. Discussion**

It is a fact that SARS-CoV-2 continues to spread rapidly on a global scale. Because of this, a better understanding of the relationship between the immune response and the severity of infection is needed. Serological tests are considered to be one of the main tools for epidemiological surveillance. If reliable serological data are available, the number of deaths associated with COVID-19 can help to determine the infection fatality ratio (IFR) for the disease. This knowledge can help to establish the relationship between fatalities and the total number of infections [8]. In our work, we thus provide a basis for further analysis of the pathogenesis of COVID-19, especially in severe cases, and protective immunity to SARS-CoV-2. It may also be implicated in designing an effective vaccine to protect against and treat SARS-CoV-2 infection.

Serological tests are considered to be a powerful complement to nucleic acid tests, especially for COVID-19 patients with undetectable viral RNA. Most of these tests based on the method of immunochromatography or ELISA and recombinant SARS-CoV-2 proteins have different sensitivity and specificity. In this regard, scientific groups studying the immune response of patients to COVID-19 use both commercial diagnostic kits and kits of their own production [9].

We have developed an author's ELISA test system based on antigens obtained by our scientific group. These include inactivated purified concentrated native viral preparations (SARS-CoV-2/human/RUS/Nsk-FRCFTM-1/2020 and SARS-CoV strain Frankfurt 1, 2002) and the recombinant proteins N, S, and RBD. This test system was used to determine the median titers of specific and cross-reactive antibodies in the blood sera of convalescents and patients who subsequently died from COVID-19 infection. It is shown that the crossactivity of antibodies (IgG resulting from COVID-19) with the inactivated heterologous SARS-CoV antigen (2002) increased in accordance with their titer, determined based on the homologous SARS-CoV-2 antigen in ELISA. The use of recombinant SARS-CoV-2 proteins allowed us to determine the individual profile of antibodies specific to individual viral proteins. According to the ELISA data, these antibodies in this panel of blood serum convalescents from mild to severe COVID-19, as well as in subsequent patients who died in the hospital, are distributed by specificity and quantitatively (in median titer values), mainly in the following descending order: N > trimer S > RBD.

The immunoblotting allowed us to identify linear epitopes of the main SARS-CoV-2 target proteins. In the case of a mild case of COVID-19, it is mainly a nucleoprotein. In moderate and severe cases with a favorable outcome, these are the nucleoprotein and the S trimer (270 kDa) and dimer (180 kDa). In a severe clinical course and a fatal outcome, antibodies specific to both the nucleoprotein and the S trimer (270 kDa) and dimer (180 kDa) were found in patients at different intervals before death. The literature data confirm that these coronavirus proteins are the most immunogenic of the four structural viral proteins, in addition to E (envelope) and M (membrane), to which the humoral immune response of the infected body is formed [10]. When the antibodies of some sera from surviving patients of severe COVID-19 do not interact with the linear epitopes of the S protein in immunoblotting, this cannot indicate the absence of such antibodies. In fact, it is most likely that they may be specific to the significant conformational epitopes of this viral protein, which are destroyed by abrupt electrophoretic separation.

The SARS-CoV nucleoprotein (2002) in immunoblotting was the major protein for cross-reactive antibodies in asymptomatic, mild, moderate, or severe COVID-19 with a favorable outcome.

The antibodies of some blood sera detected the M protein, which also causes the production of specific antibodies in patients, as shown by Guan M. et al. while testing the blood sera by immunoblotting, both when using an inactivated viral preparation and recombinant M protein received in eukaryotic expression system [11]. In our previous report [12], we confirmed the data of Sturman L. et al. [13], who showed that in order to visualize the matrix protein of mouse coronavirus A59, the viral preparation should not be heated prior to the electrophoretic separation of proteins.

Therefore, in this work, a viral preparation was used without heating, and it was found that not all patients suffering from COVID-19 develop an immune response to this protein. The M protein of SARS-CoV (2002) also was the target of weak cross-interactions for these antibodies.

Ouyang J. et al., summarizing the literature data, reported that titers of neutralizing antibodies in the range of 1:40–1:640 are detected in patients with COVID-19 and that the titers for donor plasma should be at least 1:80–1:160. The same authors emphasized that high titers of antibodies do not prevent a severe case of COVID-19 [14]. The results of serological testing using the recombinant S and N proteins showed that the majority of patients develop stable antibody responses to these structural viral proteins between the 17th and 23rd days after the disease started. A slower but stronger antibody response has been observed in severe cases. It was noted that in patients with more severe cases of COVID-19, antibodies specific to the S protein showed high titers [15].

According to our data, it is shown that the presence of antibodies specific to RBD mostly correlates with the data on the neutralizing activity of blood sera. With almost all deaths, barring a few exceptions, the titers of neutralizing antibodies were at the level of severely ill convalescents and even higher, but this did not save this group of patients. Most likely, this happened, as described in the literature, due to multiple organ disorders in the infected body as a result of SARS-CoV-2's immunopathogenesis. This was expressed in the fact that viral activation of dendritic cells and macrophages in lymphoid tissues leads to an excessive uncontrolled anti-inflammatory response, the so-called cytokine storm [16]. In addition, data have already been obtained on the dependence of the severity of COVID-19, which was established at the beginning of the pandemic, with increased IgG titers when compared with a low level of antibodies of this class in convalescents. The mechanism responsible for the immunopathological IgG response remains unclear [17].

It is not known whether patients from our random sample had previous contact with other seasonal coronaviruses (CoVs), for which antibodies are widely present in the human population. Several conserved regions have already been identified in the S2 domains of four known CoVs, namely 229E, NL63, OC43, HKUI, and SARS-CoV-2 [18], which is believed to be the cause of the antibody-dependent increase in the infection [19]. However, under ELISA and immunoblotting, when using the N recombinant protein of SARS-CoV-2 obtained in *E. coli*, cross-activity with this protein was observed only for antibodies of blood sera of patients who had previously recovered from SARS-CoV, but not with other CoVs [20]. In addition, the authors analyzed the amino acid sequence (aa) of the N protein of known CoVs isolated from humans. The N protein of SARS-CoV-2 turned out to have a homology of 19.1, 20.0, 26.5, 27.6, 46.1, and 90.5% with 229E, NL63, OC43, HKUI, MERS-CoV, and SARS-CoV, respectively [20].

In this study, we did not take into account the comorbidities of the patients. This aspect is also important for understanding viral pathogenesis. It is known that risk factors for a fatal outcome in COVID-19, in addition to male sex and old age [12], include diabetes mellitus and hyperglycemia that develops when infected with SARS-CoV-2 [21], cardiovascular diseases [22], etc. The analysis of the neutralizing activity of specific antibodies is an important stage in the study of the immune response of patients with COVID-19. Therefore, an extensive study of the immune response to SARS-CoV-2 in the population must be associated with such data for each patient with COVID-19, and this shows a further necessity for deeper investigations into the role of humoral immunity in COVID-19 disease.

A varied course of coronavirus infections was observed over the course of a year and a half of fighting the pandemic. In this study, we highlighted the immune response in patients with a different course of COVID-19 at the very beginning of the pandemic in one of the major cities of Russia under restrictions (the absence of air and railway connections with neighboring countries). It may reflect the situation at a specific point and is valuable precisely in a retrospective analysis of the development of the epidemic process. We tried to get closer to understanding the role of humoral immunity in the development of pathogenesis, which still requires careful study.
