Article

ISSN 0006-2979, Biochemistry (Moscow), 2024, Vol. 89, No. 5, pp. 872-882 © Pleiades Publishing, Ltd., 2024.

872

Humoral and Cellular Immune Response

to SARS-CoV-2 S and N Proteins

Zulfiia E. Afridonova

, Anna P. Toptygina

1,2,a

*, and Ilya S. Mikhaylov

G.N.Gabrichevsky Research Institute for Epidemiology and Microbiology, 125212 Moscow, Russia

Lomonosov Moscow State University, 119991 Moscow, Russia

Moscow Power Engineering Institute, 111250 Moscow, Russia

e-mail: toptyginaanna@rambler.ru

Received September 2, 2023

Revised October 12, 2023

Accepted November 1, 2023

Abstract— The pandemic of a new coronavirus infection that has lasted for more than 3 years, is still accompanied

by frequent mutations in the S protein of SARS-CoV-2 and emergence of new virus variants causing new disease

outbreak. Of all coronaviral proteins, the S and N proteins are the most immunogenic. The aim of this study was

to compare the features of the humoral and T-cell immune responses to the SARS-CoV-2 S and N proteins in peo-

ple with different histories of interaction with this virus. The study included 27 individuals who had COVID-19

once, 23 people who were vaccinated twice with the SputnikV vaccine and did not have COVID-19, 22 people who

had COVID-19 and were vaccinated twice with Sputnik V 6-12 months after the disease, and 25 people who had

COVID-19 twice. The level of antibodies was determined by the enzyme immunoassay, and the cellular immunity

was assessed by the expression of CD107a on CD8

high

lymphocytes after recognition of SARS-CoV-2 antigens. It was

shown that the humoral immune response to the N protein was formed mainly by short-lived plasma cells synthe-

sizing IgG antibodies of all four subclasses with a gradual switch from IgG3 to IgG1. The response to the S protein

was formed by short-lived plasma cells at the beginning of the response (IgG1 and IgG3 subclasses) and then by

long-lived plasma cells (IgG1 subclass). The dynamics of antibody level synthesized by the short-lived plasma cells

was described by the Fisher equation, while changes in the level of antibodies synthesized by the long-lived plas-

ma cells were described by the Erlang equation. The level of antibodies in the groups with the hybrid immunity

exceeded that in the group with the post-vaccination immunity; the highest antibody content was observed in

the group with the breakthrough immunity. The cellular immunity to the S and N proteins differed depending

on the mode of immune response induction (vaccination or disease). Importantly, the response of heterologous

CD8

Tcell to the N proteins of other coronaviruses may be involved in the immune defense against SARS-CoV-2.

DOI: 10.1134/S0006297924050080

Keywords: COVID-19, SARS-CoV-2, N protein, S protein, antibodies, vaccination, hybrid immunity, cellular immunity,

breakthrough immunity

Abbreviations: BAU, binding antibody unit; COVID-19,coro-

navirus infectious disease 19; SARS-CoV-2,acute respiratory

syndrome coronavirus2.

* To whom correspondence should be addressed.

INTRODUCTION

Recent pandemic of the new coronavirus infection

that has lasted for more than 3 years, is still accompa-

nied by frequent mutations in the S protein of SARS-

CoV-2 (acute respiratory syndrome coronavirus2) and

emergence of new virus variants causing further dis-

ease outbreaks, suggesting that this infection will re-

main with humanity for many more years. Although

it might become less severe, the fight against it may

turn into a permanent problem [1]. The most immuno-

genic of coronavirus proteins are the S and N proteins

[2] that induce generation of large amounts of anti-

bodies in response to the SARS-CoV-2 infection [3, 4].

Anti-SARS-CoV-2 vaccines target the S protein, since an-

tibodies against this protein provide strong protection

against the infection. However, frequent mutations in

IMMUNE RESPONSE TO SARS-CoV-2 S AND N PROTEINS 873

BIOCHEMISTRY (Moscow) Vol. 89 No. 5 2024

the S protein have led to a decrease in the efficacy of

existing vaccines [5, 6]. The interest of researcher in

the S protein has left the N protein somewhat in the

shadows. This protein is highly conserved among coro-

naviruses and is one of the most abundant structural

proteins in the virus-infected cells [7]. The main func-

tion of the N protein is to package the viral genomic

RNA into a long helical ribonucleocapsid complex and

participate in the virion assembly through interaction

with the viral genome and membrane protein  M  [8].

The location of the N protein in the center of the coro-

navirus virion explains why even high levels of anti-

bodies against this protein do not protect against the

disease, as the antibodies cannot enter the assembled

virion and contact the N protein. Moreover, the effects

of anti-N protein antibodies are poorly understood [9].

At the same time, the N-protein is a representative

antigen in the T-cell response against the infection.

Itwas shown that the T-cell response formed against

SARS-CoV-1 persists for many years [10-12]. Since the

N protein is highly conserved, it contains epitopes

that could generate T-cell immune responses that are

cross-reactive across SARS-CoV-2 and other human

coron aviruses [13].

The purpose of this study was to compare the

characteristics of humoral and T-cell immune respons-

es to the SARS-CoV-2 S and N proteins in people with

different histories of interaction with the virus.

MATERIALS AND METHODS

Analyzed cohorts and collection of biological

material. Simple open-label comparative study in-

cluded 97 adult volunteers aged 18-73 years. Of these,

27 people had a history of mild to moderate COVID-19

(coronavirus disease 19) confirmed by at least one

positive PCR test (group  1; post-infectious immunity).

They were examined 4 to 7 times over 1-18 months

from the disease onset. Group  2 (post-vaccination im-

munity) consisted of 23  people who were vaccinated

twice with the Sputnik  V vaccine and did not have

COVID-19. Group 3 (hybrid immunity) included 22  peo-

ple who had COVID-19 and were vaccinated twice

with Sputnik  V 6-12 months after the disease. Group  4

(breakthrough immunity) included 25 people who had

COVID-19 twice: the first time in 2020-2021 and again

in 2022 (omicron strain). Blood for the study was taken

from the ulnar vein into two vacuum tubes (4  ml each)

containing either heparin (cellular immunity studies)

or coagulation activator and gel for isolating blood se-

rum (assessment of the humoral immune response to

the SARS-CoV-2 antigens), respectively. The study was

approved by the Ethics Committee of the Gabrichevsky

Research Institute for Epidemiology and Microbiology

(protocol no.  58; December15, 2021). Informed volun-

tary consent was obtained from each participant in-

cluded in the study.

Evaluation of antibody levels. Blood serum was

obtained by centrifugation, transferred into Eppendorf

tubes, and stored at –70°C until the study. The antibody

content was determined by the enzyme immunoassay

using the SARS-CoV-2-IgG quantitative-ELISA-BEST

kit (JSC Vector-Best, Novosibirsk, Russia) for the anti-

S  protein antibodies and N-CoV-2-IgG PS kit for the

anti-N  protein antibodies (Saint-Petersburg Pasteur

Institute, St. Petersburg, Russia). The subclasses of IgG

antibodies to the SARS-CoV-2 antigens were studied

using a previously developed modification of ELISA

method [14,  15]. Briefly, we used 96-well panels with

adsorbed full-length S antigen from the SARS-CoV-2-

IgG quantitative-ELISA-BEST kit or with the N  protein

from the N-CoV-2-IgG PS kit. Instead of anti-IgG conju-

gates included in the kit, peroxidase-labeled anti-IgG1,

IgG2, IgG3, and IgG4 monoclonal antibodies (Polignost,

St.Petersburg, Russia) were used at a concentration of

1  μg/ml. All other stages of the assay were carried out

according to the kit instructions.

Assessment of cellular immunity. Mononuclear

cells were isolated from heparinized blood under ster-

ile conditions using gradient centrifugation (ρ = 1.077;

PanEco, Russia) and washed from platelets. The cells

were transferred to the wells of a sterile 96-well plate

(2.5  ×  10

cells per well) containing RPMI-1640 medium

supplemented with 2  mM L-glutamine, gentamicin, and

10% fetal calf serum (PanEco). Monensin (final con-

centration, 10 μM) and PE-Cy5-labeled monoclonal an-

tibody against CD107a (final dilution, 1  :  100) (control

samples) were added to the wells; the final volume

in the well was 200 μl. In experimental samples the

cells were stimulated with the SARS-CoV-2 S and N

antigens using the plates from the corresponding an-

tibody ELISA kits (see above) that had the S or N pro-

teins adsorbed at the bottom of the wells. Since the

ELISA plates were not sterile, they were sterilized by

ultraviolet irradiation for 30  min before the experi-

ment. All ingredients were added equally to both the

experimental and control wells, according to the pre-

viously developed method. Experimental and control

samples were incubated at 37°C in a humidified atmo-

sphere with 5%  CO

for 20  h, transferred into the tubes

for cytofluorimetry, washed with CellWash (300g for

5  min), stained with FITC-labeled antibodies against

CD8 for 20 min in the dark at 4°C, washed again un-

der the same conditions, and immunophenotyped us-

ing a BD FACS CantoII flow cytometer (Becton Dickin-

son Technologies and Software, USA). When analyzing

the results, we established the lymphoid gate and the

gate for lymphocytes highly expressing the CD8 anti-

gen (CD8

high

) within it and calculated the percentage

of CD8

high

CD107a

cells, i.e., cytotoxic T lymphocytes

that recognized the S or N antigens and responded by

AFRIDONOVA et al.874

BIOCHEMISTRY (Moscow) Vol. 89 No. 5 2024

releasing the content of cytotoxic granules (cytotoxic

attack). The level of 1% was considered as the limit

of spontaneous expression of the CD107a molecule on

CD8

high

lymphocytes [15].

Statistical analysis. The normality of data distri-

bution was determined by the Kolmogorov–Smirnov

method. The levels of anti-S  protein and anti-N  protein

IgG antibodies did not show the normal distribution.

The antibody content was expressed in antibody bind-

ing unit (BAU) per ml and presented as the median

(1st-3rd quartile) [Me (LQ-HQ)]. The differences be-

tween the groups were assessed using the Mann–Whit-

ney U  test. The percentage content of IgG subclasses

and cellular immunity parameters showed normal dis-

tribution and were presented as mean± standard er-

ror of mean (M  ±  SEM). The correlations were assessed

using the Pearson method. The differences were con-

sidered significant at p<0.05.

Modeling. The data on the changes in the con-

tent of anti-S protein and anti-N protein IgG antibod-

ies over time elapsed from the disease onset were ap-

proximated. The observed changes corresponded to a

distribution with the following characteristics: the ini-

tial value corresponded to the origin of coordinates; a

sharp increase in the function value to a certain max-

imum followed by its smooth decrease. As is known

from mathematical statistics, this type of distribution

is described by the Fisher and Erlang distributions.

These functional dependencies are widely known in

statistical analysis, probability theory, and biology and

belong to the Pearson typeIII distribution group (gam-

ma distributions). Our study examined two of them.

The first one was the Erlang distribution(1)

f(x; k, λ, c) = c ·

·x

k−1

·e

−λx

(k − 1)!

, (1)

where k is the shape parameter and λ is the rate pa-

rameter. The normalization coefficient c was intro-

duced to scale the function values. The combination

of k and λ determines the graph extremum position,

the function value at the extremum point, and the

function inflection smoothness.

The second one was the Fisher distribution(2):

f(x; n, m, c) =

{

c ·

−1

(

1 +

n·x

)

n+m

x ≥ 0

(2)

where coefficients n and m affect the shape of the

graph and the position and value of the function max-

imum; the coefficient c is also normalizing.

The Erlang distribution is characterized by a

smoother increase and decrease in the function val-

ues, while the Fisher distribution allows to describe

the functions whose values drop sharply after reach-

ing the maximum and then smoothly and asymptoti-

cally tend to zero.

x < 0

In this study, the general form of the approximat-

ing function was chosen based on the nature of ob-

served dependencies. We used an algorithmic process

to sort through the coefficient values to achieve the

best estimate of the standard deviation of the resulting

approximation curve from the experimental data.

RESULTS

The changes in the content of IgG antibodies

against SARS-CoV-2 antigens in the blood serum of

patients who had recovered from COVID-19 (group 1)

are presented in Fig.  1. The figure shows that the IgG

antibodies recognizing the N protein (curve 1) ap-

peared earlier and their concentration increased fast-

er in the blood with a sharp peak at 1033.2 (807.04-

1215.6)  BAU/ml 3 months after the disease onset, flowed

by a rapid decrease to 75.5 (25.5-182.2) BAU/ml after

18 months (kit cut-off value, 33.5 BAU/ml). The con-

tent of anti-S protein antibodies (curve  2) increased

more slowly and reached a plateau at 819.5 (614.7-

1538.3)BAU/ml 4 months after the disease onset, where

it remained for a year, and then decreased to 551.6

(372.5-757.5) BAU/ml by 18 months. The curve reflect-

ing the concentration of anti-N  antibodies was well ap-

proximated by the Fisher distribution (Fig. 1, curve 3)

according to the formula(3):

N−protein

(x) = 8 · 10

(0.2x)

5.5

(1 + 0.52x)

. (3)

Table  1 compares the data obtained by analyz-

ing the content of anti-N protein IgG antibodies in the

blood serum of individuals who had recovered from

COVID-19 and the results calculated using formula(3).

The experimental and calculated data differ by less

than 15%, while the calculation results strictly fall into

the LQ-HQ interval for all studied time points.

The attempts to approximate curve  2 reflecting

the concentration of anti-S protein antibodies with any

known distribution were unsuccessful. We assumed

that the curve is the result of two processes: forma-

tion of early antibody response to the S protein by

short-lived plasma cells and generation of antibodies

by long-lived plasma cells (curves4 and5 in Fig.1, re-

spectively). Both formation and death of plasma cells

are reflected in the concentration of antibodies they

produce. This process has the following properties: the

initial value corresponds to the origin of coordinates;

the antibody concentration then sharply increases to a

certain maximum value followed by a smooth decrease

in the function value. As is known from mathematical

statistics, this curve can be described by the Fisher

and Erlang distributions. Therefore, it was necessary

to adjust the two distributions so that their sum cor-

responded to curve2 in Fig.1. In this case, formation

IMMUNE RESPONSE TO SARS-CoV-2 S AND N PROTEINS 875

BIOCHEMISTRY (Moscow) Vol. 89 No. 5 2024

Fig. 1. The content of IgG antibodies to SARS-CoV-2 virus antigens. Curves: 1)IgG antibodies to the N protein (experimental

data); 2)IgG antibodies to the S protein (experimental data); 3)Fisher approximation of the content of IgG antibodies against

the Nprotein (short-lived plasma cells); 4)Fisher approximation of the content of IgG antibodies against the S protein (short-

lived plasma cells); 5)Erlang approximation of the content of IgG antibodies against the S protein (long-lived plasma cells).

Table 1. Comparison of the experimental data on the content (BAU/ml) of IgG antibodies against SARS-CoV-2

N protein in the blood serum from recovered individuals and the results of calculations using formula(3)

Time from

the disease onset

IgG antibodies against

N protein [Me (LQ-HQ)]

Calculation according

to the Fisher formula (3)

Deviation of the calculated value

from the experimental value, %

1 month 230.9 (118.4-430.6) 264.34 –14.5%

2 months 899.2 (497.8-1225.9) 846.74 5.83%

3 months 1033.2 (807.04-1215.6) 1020.38 1.2%

4 months 870.1 (319.7-3790.36) 940.00 –8.0%

6 months 579.3 (374.3-2524.9) 637.55 –10.1%

12 months 211.5 (116.9-507.3) 180.55 14.6%

18 months 75.5 (25.5-182.2) 66.76 11.6%

of antibodies by the early producers was also approx-

imated with the Fisher distribution (curve  4 in Fig.  1)

according to the formula(4):

S−protein Fisher

(x) = 7 · 10

(0.2x)

4.5

(1 + 0.66x)

13.75

, (4)

while the curve describing the synthesis of antibodies

by the long-lived plasma cells (curve 5 in Fig. 1) was

approximated by the Erlang distribution(5):

S−protein Erlang

(x) = 1100 · 

3.1

5.1

 · (0.11x)

4.1

 · e

−3.1x

. (5)

AFRIDONOVA et al.876

BIOCHEMISTRY (Moscow) Vol. 89 No. 5 2024

Table 2. Comparison of the experimental data on the content (BAU/ml) of IgG antibodies against SARS-CoV-2

Sprotein in the blood serum from recovered individuals and the results of calculations using formulas(4) and(5)

Time from

the disease

onset

S protein

[Me (LQ-HQ)]

Calculation

according to the

Fisher formula (4)

Calculation

according to the

Erlang formula (5)

Sum of calculations

obtained using

formulas(4) and (5)

Deviation of the sum

of the calculated values

from the experimental

data, %

1 month

103.3

(73.37-189.1)

89.60 1.23 90.83 12.1%

2 months

456.2

(199.2-1027.1)

439.30 14.96 454.26 0.4%

3 months

787.4

(356.1-1190.2)

687.89 56.07 743.96 5.5%

4 months

819.5

(614.7-1538.3)

718.62 129.69 848.31 –3.5%

6 months

841.8

(614.9-1420.7)

491.38 345.69 837.07 0.6%

12 months

810.9

(504.5-1215.3)

70.82 766.21 837.03 –3.2%

18 months

551.6

(372.5-757.5)

10.98 522.10 533.08 3.4%

When searching for the best values of approxi-

mating curve coefficients, we excluded the data corre-

sponding to 6 months from the disease onset. Instead,

this data point was used as a test point to avoid model

overtraining.

Table  2 compares the experimental data on the

levels of anti-S  protein IgG antibodies and approxima-

tion results obtained using the formulas (4) and (5).

In can be seen that the sums of the value calculated

using formulas (4) and (5) deviate from the experi-

mentally obtained values by no more than 13% and

strictly fall into the calculated LQ-HQ intervals at all

time points. When selecting the optimal coefficients

for the Fisher and Erlang formulas(3),(4), and(5), in

order to assess the quality of the applied models, we

calculated the root mean square error (RMSE) and the

mean absolute percentage error (MAPE) as the quality

metrics. When modeling the changes in the content of

anti-N protein IgG antibody with time occurring from

the disease onset using the coefficients presented in

formula (3), we obtained the minimum values of the

quality metrics (RMSE, 16.505; MAPE, 9.408%), indicat-

ing a good quality of the proposed model.

The minimum values of the quality metrics (RMSE,

8.949; MAPE, 4.096%) calculated using the coefficients

presented in formulas(4) and(5) also indicated a high

quality of the proposed model.

The changes in the relative content of IgG sub-

classes against the N and S proteins are shown in Fig.2.

Interesting, the N protein induced production of all

four IgG subclasses (although IgG2 and IgG4 were mi-

nor), whereas only the IgG1 and IgG3 subclasses of

anti- S  protein antibodies were detected, while anti-

S  protein IgG2 and IgG4 antibodies were absent. The hu-

moral response to both proteins showed the same trend:

IgG3 antibodies were gradually replaced by IgG1 an-

tibodies, indicating antibody response maturation, al-

though the rate and completeness of such replacement

differed. Thus, IgG1 antibodies represented 54.6  ±  2.7%

of all anti-S  protein antibodies already one month after

the disease onset vs. 35.2  ±  1.1% of anti-N protein IgG1

antibodies at the same time point. Within a year, the

fraction of anti-N protein IgG1 antibodies increased

to 71.38  ±  3.2% and remained at this level for another

6  months. In the case of the S protein, the content of

anti-S  protein IgG1 antibodies rapidly increased to

97.4  ±  0.5% 6 months after the disease onset, reached

100% by 12 months in all tested individuals, and re-

mained at this level for at least another 6 months.

Figure  3 shows a comparison between the levels

of anti-N  protein and anti-S  protein IgG antibodies for

the four studied groups 3 months after the disease on-

set (group  1), 3 months after immunization with the

second dose of vaccine (groups2 and3), and 3  months

after the recurrent disease onset (omicron strain)

(group 4). The content of antibodies varied in differ-

ent groups. Thus, in patients who had recovered from

COVID-19 once (group 1), the level of anti-N  protein

IMMUNE RESPONSE TO SARS-CoV-2 S AND N PROTEINS 877

BIOCHEMISTRY (Moscow) Vol. 89 No. 5 2024

Fig. 2. Changes in the content of different subclasses of IgG antibodies against SARS-CoV-2 antigens. a)N protein. b)S protein.

antibodies varied greatly, but did not significantly

exceed the level of anti-S  protein antibodies. No anti-

N  protein antibodies were detected in individuals vac-

cinated with Sputnik  V (group  2), which can be ex-

plained by the absence of N protein in this vaccine.

Ingroup3 with the hybrid immunity (vaccination 6-12

months after COVID-19), the content of anti-N protein

IgG antibodies was significantly lower (p  =  0.018) than

the level of antibodies to the S-protein. The content of

anti-S  protein antibodies in group  3 was significantly

higher (p  =  0.004) than in group  2. Group  4 with the

breakthrough immunity (patients who had COVID-19

twice) demonstrated a significant increase in the levels

of both anti-S  protein (p  =  0.0006) and anti-N  protein

(p  = 0.042) antibodies compared to group1.

Figure 4 shows a contribution of IgG1 antibodies

to the overall IgG response to the N and S proteins.

Interestingly, all anti-S  protein IgGs in all 4 groups

were of the IgG1 subclass. At the same time, the frac-

tion of anti-N  protein IgG1 antibodies was 72.8  ±  3.5%

in group 1. This subclass of antibodies was complete-

ly absent in group 2, since the vaccinated individuals

did not form a response to this protein. In group3, the

relative content of IgG1 antibodies did not differ sig-

nificantly from that in group1 (75.9  ±  3.8%), which is

understandable, since the immune response in these

patients had formed during the primary COVID-19

event, while the vaccine used for the following vacci-

nation later did not contain the N protein. In group4

(patients who had COVID-19 twice), the percentage of

IgG1 antibodies reached 99.1  ±  0.3%, which was signifi-

cantly different from groups1 and3 (p  <  0.01), indicat-

ing that maturation of anti-N protein antibodies contin-

ued with the secondary response to the N protein.

AFRIDONOVA et al.878

BIOCHEMISTRY (Moscow) Vol. 89 No. 5 2024

Fig. 3. Comparison of levels of IgG antibodies against N and S proteins in patients recovered from COVID-19 (group1), individu-

als vaccinated twice with SputnikV (group2), patients who had been ill with COVID-19 and then were vaccinated with SputnikV

(group3), and patients who had recovered from COVID-19 twice (group4).

Fig. 4. Contribution of IgG1 antibodies to the overall IgG response to the SARS-CoV-2 N and S proteins.

The results on the cellular immune response to

the N and S antigens are presented in Fig.  5. The cellu-

lar response to the S protein did not differ significant-

ly between the studied groups, although it was slightly

higher in the group with the breakthrough immunity.

The cellular response to the N protein in group3 was

significantly lower than in groups 1 and 4 (p  <  0.05).

We expected no cellular response to the N protein in

group  2 (individuals vaccinated with Sputnik V), as

in the case of humoral immune response. Indeed, we

saw no cellular response in 17 people in this group;

however, 6 people, who did not have the antibodies

IMMUNE RESPONSE TO SARS-CoV-2 S AND N PROTEINS 879

BIOCHEMISTRY (Moscow) Vol. 89 No. 5 2024

Fig. 5. Cellular immune response to the N and S proteins of SARS-CoV-2.

against the N  protein, as well as lacked antibodies to

the S  protein, prior to the vaccination demonstrated

a significant cellular response to the N protein, which

resulted in the average response level of 5.94± 2.3%

in group2. There was a strong positive correlation be-

tween the levels of cellular responses to the N and S

proteins (r  =  0.937). We also revealed a weak positive

correlation between the humoral and cellular respons-

es to the S protein (r=0.358) and the absence of such

correlation for the N protein.

DISCUSSION

We found that in the individuals who had recov-

ered from COVID-19, the concentration of IgG anti-

bodies to the N protein in the blood increased faster

than the concentration of anti S-protein antibodies.

The content of anti-N  antibodies showed a higher and

sharper peak and a more rapid decline. Similar results

were obtained by other researchers [16]. The curve

describing changes in the concentration of anti-N  pro-

tein antibodies was well approximated by the Fisher

distribution (Fig.  1), which is a special case of the Pear-

son distribution. In the case under consideration, the

antibody concentration in the blood was influenced

by two independent events: formation of early, short-

lived plasma cells that synthesized these antibodies,

which was followed by the death of these cells over

time, resulting in the decrease in the antibody con-

centration. The changes in the level of anti-S protein

antibodies had a different pattern. The concentration

of these antibodies increased more slowly, and instead

of a peak, reached a plateau that lasted up to a year,

after which the content of the antibodies gradually de-

creased. Our attempts to approximate this curve with a

single function were unsuccessful. We believe that the

curve representing changes in the anti-S  antibody con-

centration is a sum of two independent processes. The

first process, which is formation of early, short-lived

plasma cells synthesizing anti-S  protein antibodies, is

similar to that observed during formation of anti-N

protein antibodies and could be well approximated

by the Fisher distribution. The peak formation of the

antibodies by the short-lived plasma cells occurred at

3 months after the disease onset for both anti-N  pro-

tein and anti-S proteins antibodies. The second parts

of the curves ran almost parallel to each other, indi-

cating that the respective processes were identical (see

Fig. 1). The second event was the formation of long-

lived plasma cells that also synthesized anti-S  protein

antibodies. This second process was more prolonged

in time and could be approximated by the Erlang dis-

tribution (γ-distribution). This function is also a spe-

cial case of the Pearson distribution and is applicable

to describe the results of two continuous independent

events over time, in our case, formation and death of

long-lived plasma cells producing anti-S  protein anti-

bodies, which was reflected in the concentration of

these antibodies in the blood.

It is known that short-lived plasma cells synthe-

size predominantly IgG3 antibodies, while long-lived

AFRIDONOVA et al.880

BIOCHEMISTRY (Moscow) Vol. 89 No. 5 2024

plasma cells produce mainly IgG1 antibodies [17]. In-

terestingly, the initial response to the N protein repre-

sented formation of mostly IgG3 antibodies had been

formed in, while the switch to the IgG1 subclass oc-

curred slowly, with the content of IgG1 reaching ~  70%

within 6 months after the disease onset and remaining

at this level even after 1.5 years. Similar results were

obtained by other researchers [18]. At the same time,

in response to the S protein, this switch had occurred

much faster; the fraction of IgG1 exceeded 90% with-

in 3 months after the disease onset and then reached

100%. Presumably, these differences are due to the

fact that the short-lived plasmacytes dominated in the

response to the N protein, while the response to the

S protein was accompanied by the formation of long-

lived plasmacytes as well. It is also very likely that the

differences in the response to the two highly immu-

nogenic proteins of SARS-CoV-2 are associated with the

functions of these proteins. Thus, the N protein is lo-

cated inside the virion and is active at the stage of vi-

rus replication; anti-N  protein antibodies are not pro-

tective [9]. On the contrary, the S protein is located on

the virion surface and is responsible for the virus at-

tachment and fusion with the infected host cell; there-

fore, the antibodies against this protein can block the

SARS-CoV-2 infection [19].

If our approximations of changes in the antibody

levels are correct, then according to the Erlang func-

tion, the level of anti-S  protein antibodies should fall

slightly below 300  BAU/ml 2 years after the disease on-

set, decrease to ~40  BAU/ml after 3  years, and drop

below 10  BAU/ml (a cut off between the negative and

positive levels of these antibodies) after 4  years. Sim-

ilar dynamics in the antibody levels was observed in

patients who suffered from SARS-CoV-1 and MERS vi-

ruses [20]. Perhaps. this would have been the case if

SARS-CoV-2 had not mutated so often and had been

eliminated from the human population. Unfortunate-

ly, the reality presents a different picture. SARS-CoV-2

actively mutates; most mutations occur in the S pro-

tein, while the N protein remains the most conserved

one [21]. Such mutations allow the virus to evade the

antibody defense, leading to recurrent illnesses. Also,

active vaccination of the population has adjusted the

duration of antibody protection against SARS-CoV-2.

We studied 4 groups of people who had different histo-

ries of contact with SARS-CoV-2. People vaccinated twice

with Sputnik  V did not differ from those who had re-

covered from COVID-19 in the anti-S protein antibody

level and relative content of IgG subclasses. However,

the concentrations of both anti-N protein and anti-

S  protein antibodies in the individuals who had been ill

with COVID-19 twice (at the beginning of the pandemic

and again with the Omicron variant) were significantly

higher than in people who had been ill once. Interest-

ingly, not only the level of anti-N  protein anti bodies in-

creased, but these antibodies have become represented

almost completely by the IgG1 subclass. This suggests

that although the level of anti-N protein antibodies,

and therefore the content of plasma cells that synthe-

size them, were already very low at the time of relapse,

memory B cells responded with the secondary immune

response to the repeated recognition of the N  protein,

resulting in additional maturation of anti-N protein an-

tibodies. Anti-S  protein antibodies demonstrated a high

booster effect in the recurrent disease.

Both S and N proteins induced formation of cel-

lular immune response of CD8

cytotoxic lymphocytes.

It has been shown that T  cell respond not only to the

structural, but also to the accessory proteins of SARS-

CoV-2 [22]. The levels of response to the S protein in

the four studied groups did not differ significantly. This

indicates that CD8

lymphocytes are actively involved

in the immune response to both the disease and vac-

cination against COVID-19. Thus, it was shown that

pre-existing T  cells specific to the SARS-CoV-2 proteins

are able to prevent the development of the COVID clin-

ical picture [23]. The level of cellular response to the

N-protein in the group with the hybrid immunity (peo-

ple who had recovered from COVID-19 and were lat-

er vaccinated with Sputnik  V) was significantly lower

than in the group of patients who had recovered from

the disease, which could be explained by the absence

of N  protein in the composition of this vaccine. The

discovery of a high cellular immune response to the

N  protein in 6 people in the vaccinated group was

unexpected. However, they did not have anti-N  pro-

tein antibodies, and before vaccination, no antibodies

against the SARS-CoV-2 S  protein were detected. We be-

lieve that his may be due to a heterologous immune

response. It is likely that these people had previously

suffered from one of the common cold coronaviruses,

which had circulated freely in the human population

even before 2019. The N protein is extremely conserved

and contains epitopes that can cause the cross-reactiv-

ity of the T  cell-mediated immunity response among

different coronaviruses [13]. Any viral protein can be

an antigen for a T  cell response and trigger an attack

of cytotoxic cells. It is possible that such heterologous

immune response to the N protein of the common cold

coronaviruses has provided protection in people who

had mild or asymptomatic COVID-19. On the other

hand, it cannot be excluded that these 6 people suf-

fered from SARS-CoV-2 asymptomatically and without

IgG formation, but responded by forming the T cell re-

sponse, as it has been described before [24].

CONCLUSION

In conclusion, we demonstrated that the humor-

al immune responses to the S and N proteins of SARS-

IMMUNE RESPONSE TO SARS-CoV-2 S AND N PROTEINS 881

BIOCHEMISTRY (Moscow) Vol. 89 No. 5 2024

CoV-2 could form independently of each other. In our

case, the N protein induced formation of predominant-

ly short-lived plasma cells synthesizing IgG antibod-

ies of all four subclasses with a gradual switch from

IgG3 to IgG1 (about 70%). The response to the S pro-

tein included formation of both short-lived plasma-

cytes (which formed at the beginning of the response)

and long-lived plasmacytes. Short-lived plasma cells

respond to the S-protein with the synthesis of IgG1

and IgG3 subclasses, while long-lived plasma cells pro-

duced IgG1 antibodies. The changes in the content of

antibodies synthesized by the short-lived plasma cells

were described by the Fisher distribution, while the

Erlang distribution was more suitable to describe the

levels of antibodies synthesized by the long-lived plas-

ma cells. The content of antibodies in the groups with

the hybrid immunity exceeded the antibody levels in

people with the post-vaccination immunity. The con-

tent of antibodies in the group with the breakthrough

immunity it exceeded that in groups with the post-in-

fectious and post-vaccination immunity. The cellular

immunity to the S and N proteins varied depending

on the method of immune response induction (vacci-

nation or disease). Importantly, heterologous immune

response of CD8

T cells to the N protein of other coro-

naviruses may be involved in the immune protection

against SARS-CoV-2.

Acknowledgments. The authors express their grat-

itude to the Pasteur Research Institute of Epidemiology

and Microbiology (St.Petersburg, Russia) for providing

N-CoV-2-IgG PS test kits.

Contributions. Z.E.A. conducted experiments;

A.P.T. developed the study concept, supervised the

study, conducted experiments, discussed results, wrote

and edited the manuscript; I.S.M. carried out mathe-

matical modeling and discussed the results.

Funding. The work was carried out within the

framework of R&D project 121021100125-4 (02/10/2021).

Ethics declarations. All studies were conducted

in accordance with the principles of biomedical ethics

as outlined in the 1964 Declaration of Helsinki and its

later amendments. Each participant in the study pro-

vided a voluntary written informed consent after re-

ceiving an explanation of the potential risks and ben-

efits, as well as the nature of the upcoming study. The

authors of this work declare that they have no con-

flicts of interest.

REFERENCES

1. Markov, P. V., Ghafari, M., Beer, M., Lythgoe, K., Sim-

monds,P., Stilianakis, N.I., and Katzourakis,A. (2023)

The evolution of SARS-CoV-2, Nat. Rev. Microbiol., 21,

361-379, doi:10.1038/s41579-023-00878-2.

2. Meyer, B., Drosten, C., and Müller, M. A. (2014) Sero-

logical assays for emerging coronaviruses: challeng-

es and pitfalls, Virus Res., 194, 175-183, doi: 10.1016/

j.virusres.2014.03.018.

3. Sun, B., Feng, Y., Mo, X., Zheng, P., Wang, Q., Li, P.,

Peng,P., Liu,X., Chen,Z., Huang,H., Zhang,F., Luo,W.,

Niu,X., Hu,P., Wang,L., Peng,H., Huang,Z., Feng,L.,

Li,F., Zhang,F., Li,F., Zhong,N., and Chen,L. (2020) Ki-

netics of SARS-CoV-2 specific IgM and IgG responses in

COVID-19 patients, Emerg. Microbes Infect., 9, 940-948,

doi:10.1080/22221751.2020.1762515.

4. Qu,J., Wu,C., Li,X., Zhang,G., Jiang,Z., Li,X., Zhu,Q.,

and Liu, L. (2020) Profile of immunoglobulin G and

IgM antibodies against severe acute respiratory syn-

drome coronavirus 2 (SARS-CoV-2), Clin. Infect. Dis.,

71, 2255-2258, doi:10.1093/cid/ciaa489.

5. Rayati Damavandi, A., Dowran, R., Al Sharif, S.,

Kashanchi,F., and Jafari,R. (2022) Molecular variants

of SARS-CoV-2: antigenic properties and current vac-

cine efficacy, Med. Microbiol. Immunol., 211, 79-103,

doi:10.1007/s00430-022-00729-6.

6. Wang,Q., Ye, S.B., Zhou, Z.J., Song, A.L., Zhu,X., Peng,

J. M., Liang, R. M., Yang, C. H., Yu, X. W., Huang, X.,

Yu, J., Qiu, Y., and Ge, X. Y. (2023) Key mutations in

the spike protein of SARS-CoV-2 affecting neutraliza-

tion resistance and viral internalization, J.Med. Virol.,

1, e28407, doi:10.1002/jmv.28407.

7. He,Y., Zhou, Y., Wu, H., Kou,Z., Liu,S., and Jiang, S.

(2004) Mapping of antigenic sites on the nucleocapsid

protein of the severe acute respiratory syndrome coro-

navirus, J.Clin. Microbiol., 42, 5309-5314, doi:10.1128/

JCM.42.11.5309-5314.2004.

8. Masters, P. S., and Sturman, L. S. (1990) Background

Paper Functions of the Coronavirus Nucleocap-

sid Protein, in Coronaviruses and Their Diseases,

Springer, Berlin/Heidelberg, Germany, pp. 235-238,

doi:10.1007/978-1-4684-5823-7_32.

9. Batra, M., Tian, R., Zhang, C., Clarence, E., Sacher,

C.S., Miranda, J.N., De La Fuente, J.R.O., Mathew,M.,

Green, D., Patel, S., Bastidas, M. V. P., Haddadi, S.,

Murthi, M., Gonzalez, M. S., Kambali, S., Santos,

K.H. M., Asif,H., Modarresi,F., Faghihi,M., and Mir-

saeidi,M. (2021) Role of IgG against N-protein of SARS-

CoV2 in COVID19 clinical outcomes, Sci. Rep., 11, 3455,

doi:10.1038/s41598-021-83108-0.

10. Zhou,R., To, K.K., Wong, Y.C., Liu,L., Zhou,B., Li,X.,

Huang,H., Mo,Y., Luk, T.Y., Lau, T.T., Yeung,P., Chan,

W.M., Wu, A.K., Lung, K.C., Tsang, O.T., Leung, W.S.,

Hung, I.F., et al. (2020) Acute SARS-CoV-2 infection im-

pairs dendritic cell and Tcell responses, Immunity, 53,

864-877.e865, doi:10.1016/j.immuni.2020.07.026.

11. Ng, O.W., Chia,A., Tan, A.T., Jadi, R.S., Leong, H.N.,

Bertoletti, A., and Tan, Y. J. (2016) Memory T cell re-

sponses targeting the SARS coronavirus persist up

to 11 years post-infection, Vaccine, 34, 2008-2014,

doi:10.1016/j.vaccine.2016.02.063.

AFRIDONOVA et al.882

BIOCHEMISTRY (Moscow) Vol. 89 No. 5 2024

12. Le Bert,N., Tan, A.T., Kunasegaran,K., Tham, C.Y.L.,

Hafezi,M., Chia,A., Chng, M.H.Y., Lin,M., Tan,N., Lin-

ster,M., Chia, W.N., Chen, M.C., Wang, L.-F., Ooi, E.E.,

Kalimuddin,S., Tambyah, P.A., Low, J.G.-H., Tan, Y.-J.,

and Bertoletti,A. (2020) SARS-CoV-2-specific T cell im-

munity in cases of COVID-19 and SARS, and uninfected

controls, Nature, 584, 457-462, doi: 10.1038/s41586-

020-2550-z.

13. Oliveira, S.C., de Magalhães, M.T.Q., and Homan, E.J.

(2020) Immunoinformatic analysis of SARS-CoV-2 nu-

cleocapsid protein and identification of COVID-19 vac-

cine targets, Front. Immunol., 11, 587615, doi:10.3389/

fimmu.2020.587615.

14. Toptygina, A. P., Mamaeva, T. A., Alioshkin, V. A.,

and Pukhalskii, A. L. (2004) Spectrum of anti-mea-

sles immunoglobulin G subclasses in convalescents

after measles, Bull. Exp. Biol. Med., 137, 259-261,

doi:10.1023/B:BEBM.0000031564.27747.b4.

15. Toptygina, A. P., Semikina, E. L., Zakirov, R. Sh., and

Afridonova, Z. E. (2022) Comparison of the humoral

and cellular immunity in COVID-19 convalescents,

Russ. J. Infect. Immun., 12, 495-504, doi: 10.15789/

2220-7619-COT-1809.

16. Ren,L., Zhang,L., Chang,D., Wang,J., Hu,Y., Chen,H.,

Guo,L., Wu,C., Wang,C., Wang,Y., Wang,Y., Wang,G.,

Yang, S., de la Cruz, C. S., Sharma, L., Wang, L.,

Zhang,D., and Wang,J. (2020) The kinetics of humoral

response and its relationship with the disease sever-

ity in COVID-19, Commun. Biol., 3, 780, doi: 10.1038/

s42003-020-01526-8.

17. Toptygina, A. P., and Alioshkin, V. A. (2013) Compari-

son of primary and secondary humoral immune re-

sponse to vaccination “Priorix”, Russ. J.Infect. Immun.,

3, 359-364.

18. Korobova, Z.R., Zueva, E.V., Arsentieva, N.A., Batsun-

ov, O.K., Liubimova, N.E., Khamitova, I.V., Kuznetso-

va, R.N., Savin, T.V., Totolian, A.A., Rubinstein, A.A.,

Stanevich, O.V., Kulikov, A.N., and Pevtsov, D.E. (2022)

Changes in anti-SARS-COV-2 IgG subclasses over time

and in association with diseases severity, Viruses, 14,

941, doi:10.3390/v14050941.

19. Wang,X., Guo,X., Xin,Q., Pan,Y., Hu,Y., Li,J., Chu,Y.,

Feng, Y., and Wang, Q. (2020) Neutralizing antibody

responses to severe acute respiratory syndrome coro-

navirus 2 in coronavirus disease 2019 in patients and

convalescent patients, Clin. Infect. Dis., 71, 2688-2694,

doi:10.1093/cid/ciaa721.

20. Sariol,A., and Perlman,S. (2020) Lessons for COVID-19

immunity from other coronavirus infections, Immuni-

ty, 53, 248-263, doi:10.1016/j.immuni.2020.07.005.

21. Shrock, E., Fujimura, E., Kula, T., Timms, R. T., Lee,

I. H., Leng, Y., Robinson, M. L., Sie, B. M., Li, M. Z.,

Chen, Y., Logue, J., Zuiani, A., McCulloch, D., Lelis,

F.J.N., Clarke, W. A., Caturegli, P., Laeyendecker,O.,

Piechocka-Trocha, A., Li, J. Z., Khatri, A., Chu, H. Y.,

Villani, A.-C., Kays,K., Goldberg, M.B., Hacochen,N.,

Filbin, M. R., Yu, X. G., Walker, B. D., Wesemann,

D.R., Larman, H. B., Lederer, J. A., and Elledge, S. J.

(2020) Viral epitope profiling of COVID-19 patients

reveals cross-reactivity and correlates of sever-

ity, Science, 370, eabd4250, doi: 10.1126/science.

abd4250.

22. Peng,Y., Mentzer, A.J., Liu,G., Yao,X., Yin,Z., Dong,D.,

Dejnirattisai, W., Rostron, T., Supasa, P., Liu, C.,

López-Camacho, C., Slon-Campos, J., Zhao, Y., Stuart,

D.I., Paesen, G.C., Grimes, J.M., Antson, A.A., Bayfield,

O. W., Hawkins, D.E., Ker, D. S., Turtle, L., Subrama-

niam, K., Thomson, P., Zhang, P., Dold, C., Ratcliff, J.,

Simmonds, P., de Silva, T., Sopp, P., Wellington, D.,

Rajapaksa, U., Chen, Y. L., Salio, M., Napolitani, G.,

Paes, W., Borrow, P., Kessler, B., Fry, J. W., Schwabe,

N.F., Semple, M.G., Baillie, K.J., Moore,S., Openshaw,

P. J., Ansari, A., Dunachie, S., Barnes, E., Frater, J.,

Kerr,G., Goulder,P., Lockett,T., Levin,R., Cornall, R.J.,

Conlon,C., Klenerman,P., McMichael,A., Screaton,G.,

Mongkolsapaya, J., Knight, J.C., Ogg, G., and Dong,T.

(2020) Broad and strong memory CD4

and CD8

Tcells

induced by SARS-CoV-2 in UK convalescent individu-

als following COVID-19, Nat. Immunol., 21, 1336-1345,

doi:10.1038/s41590-020-0782-6.

23. Swadling,L., Diniz, M.O., Schmidt, N.M., Amin, O.E.,

Chandran, A., Shaw, E., Pade, C., Gibbons, J. M., le

Bert,N., Tan, A.T., Jeffery-Smith,A., Tan, C.C.S., Tham,

C.Y.L., Kucykowicz,S., Aidoo-Micah,G., Rosenheim,J.,

Davies, J., Johnson, M., Jensen, M. P., Joy, G., McCoy,

L.E., Valdes, A.M., Chain, B.M., Goldblatt,D., Altman,

D. M., Boyton, R. J., Manisty, C., Treibel, T. A., Moon,

J.C., van Dorp,L., Balloux,F., McKnight,A., Noursade-

ghi,M., Bertoletti,A., and Maini, M.K. (2022) Pre-ex-

isting polymerase-specific T cells expand in abor-

tive seronegative SARS-CoV-2, Nature, 601, 110-117,

doi:10.1038/s41586-021-04186-8.

24. Schwarzkopf, S., Krawczyk, A., Knop, D., Klump, H.,

Heinold, A., Heinemann, F. M., Thümmler, L.,

Temme,C., Breyer,M., Witzke,O., Dittmer,U., Lenz,V.,

Horn, P. A., and Lindemann, M. (2021) Cellular im-

munity in COVID-19 convalescents with PCR-con-

firmed infection but with undetectable SARS-

CoV-2-specific IgG, Emerg. Infect. Dis., 27, 122-129,

doi:10.3201/2701.203772.

Publisher’s Note. Pleiades Publishing remains

neutral with regard to jurisdictional claims in pub-

lished maps and institutional affiliations.