SARS-CoV-2 cellular immune response in uninfected health care workers with prolonged and close exposure to COVID-19 patients

Health care workers (HCW) are at an increased risk since they are directly exposed to SARS-CoV-2 infected patients, nevertheless, some remained without the development of anti-SARS-CoV-2 antibodies, suggesting lesser susceptibility to infection 1-5 . This study aimed to ascertain a potential specic cellular immune response to SARS-CoV-2 in these largely exposed HCWs. In this cross-sectional, case-control study, we analyzed 39 exposed uninfected HCWs and 17 convalescent HCWs. Cellular immune response was evaluated after SARS-CoV-2 stimulation with peptide pools (proteins S, M, and N), using bead-based multiplex assay (12 cytokines). Overall, 94.8% of uninfected HCWs had some degree of specic cellular response to SARS-CoV-2 structural proteins that could be classied, according to the number of cytokine production, as strong (61.5%), partial (33.3%), and weak/no response (5.1%). Strong responders showed a higher anti-inammatory cytokine production (IL5 and IL10, p<0.001 and 0.002, respectively), and similar (IFN-γ and TNF-α, p=0.435 and 0.532, respectively) or higher (IL12, p=0.021) pro-inammatory production compared to convalescents, resulted in a predominantly Th2 response. This study demonstrated a consistent and polyfunctional immune cellular response after stimulation with SARS-CoV-2 peptides in extensively exposed individuals that should be considered to establish the infection susceptibility, the impact in herd immunity, and the risk of relapses.

Introduction SARS-CoV-2 epidemic started in December 2019 in Wuhan (China) and has spread rapidly worldwide becoming pandemic threatening public health [1][2][3][4][5][6] . Health care workers (HCW) are at increased risk of SARS-CoV-2 infection since they are continuous and directly exposed to infected individuals 7,8 . Data about the seroprevalence of SARS-CoV-2 infection among health care workers are still scarce. It is estimated to be up to 38.9% (23.7% in an intramural survey including 4968 health professionals of our hospital, data not published) [9][10][11] , while it is up to 5.7% in the general population 12 . This higher prevalence among HCW is closely related to risk factors such as exposure to aerosol-generating procedures, suboptimal handwashing after patient contact, longer work hours, and suboptimal protective personal equipment use [13][14] .
Although many HCWs referred poor access to protective equipment while having very close contact with SARS-CoV-2 infected patients, they have not developed positive serology, suggesting less susceptibility to the infection. We hypothesized that these HCWs could have, at least in part, a cellular immune response that could prevent infection or antibody development. Hence, this study aimed to investigate the potential cellular immune reactivity to SARS-CoV-2 among uninfected HCWs despite long-term direct exposure to infected patients.

Materials And Methods
This was a cross-sectional, case-control study (2:1 ratio), performed in a tertiary University Hospital. The exposure level among HCWs included the continuous care to COVID-19 patients, shortage of complete personal protective equipment, the exposure to aerosol-generating procedures 7,8 , and additional close contact with infected households. Cases and controls were identi ed through informal interviews with hospital staff and by self-identi cation, and were included from May 6 to June 1, 2020.
Uninfected HCWs with proved direct and continued COVID-19 patient care for more than two weeks with no diagnosis of current or past SARS-CoV-2 infection, ascertain by a negative serology for anti-SARS-CoV-2 antibodies in an intramural survey (IgM/IgA and IgG antibodies, Novatec Immunodiagnostica, Germany), were included as cases. Convalescent HCWs, with similar exposure to COVID-19 patients, who had been diagnosed by RT-PCR or/and speci c serology, were included as controls.
This study was approved by our Institutional Review Board (EC162/20) and performed according to the principles of the Declaration of Helsinki. Written informed consent was obtained from all the participants.
The global study was registered at the clinicaltrials.gov (NCT04402827).

Variables and laboratories measurements
Age, sex, COVID-19 symptoms, exposure to SARS-CoV-2-infected patients, handling of aerosol-generating procedures, and additional close contacts with an infected household were collected at the inclusion visit.

Statistical analysis
Estimation of sample size was not done since no data about immune response to SARS-CoV-2 was available in this population, hence, HCWs were speci cally selected and not randomly assigned. To categorize the immune response as the outcome for this study, since no other categorization has been described so far, we de ned strong responders to those individuals with a total of the cytokines produced to the three viral proteins (a maximum of 30 positive cytokine production per sample) within the three highest quartile range, while partial responders were de ned as those within the lowest interquartile range. Characteristics of both groups were compared using two-tailed statistic tests, chi-square or Fisher's exact tests for categorical variables and Student's t-test or Mann-Whitney U-tests for continuous variables. Categorical variables are shown as frequencies and proportions where continuous variables are shown as mean and standard deviation or median and interquartile ranges (IQR). A p-value below 0.05 was considered to be statistically signi cant.

Results
A total of 60 patients, 40 cases, and 20 controls, were included in the study. Four individuals were nally excluded, one case with positive serology, and three controls with negative serology. Hence, a total of 39 cases and 17 controls were nally included. The baseline characteristics of the participants are shown in table 1. The mean age was 38 years, and 55.4% were female. The handling of aerosol-generating procedures and/or additional risk contact with an infected household were similar in both groups (p=0.224 and p=0.440, respectively).
The level of the soluble plasma cytokines was similar in both groups of participants, as shown in supplementary gure 1, with a slightly lower level of IL2 found in uninfected HCWs (p=0.063).

High cellular response in uninfected HCWs
Although almost all the uninfected HCWs showed speci c response by the production of at least one cytokine after SARS-CoV-2 protein S, M, and N stimulation of the PBMCs, they had lower total number of cytokines compared to convalescents (p<0.001, gure 1A, left). Nonetheless and strikingly, uninfected HCWs had a similar total number of cytokine production to viral peptide S (p=0.289) compared to that found in convalescents ( gure 1A, right). None of the samples tested negative for the positive control.
The median number of positive cytokine production in all the participants was 12 (IQR [8][9][10][11][12][13][14][15][16][17], and therefore the production of at least eight cytokines was considered as a strong response. Convalescent HCWs had a median number of cytokine production of 18 (minimum 10 and maximum 25), ful lling our de nition of strong responders in all cases. Thus, a total of 24 uninfected HCWs (61.5%) were classi ed as strong immune responders, 13 HCWs as partial responders with less than eight cytokine production, (33.3%), whereas 5.1% were classi ed as weak/no responders (two participants with two or one cytokine production, only IL6). No differences in age, sex, time of exposure, exposure to aerosol-generating procedures, or additional contact with an infected household (p=0.460, p=0.420, p=0.404, and p=0.509 respectively) were found between strong and partial responders. Uninfected HCWs with strong response had lower global cytokine production compared to convalescents ( gure 1B, left), but they produced higher number of cytokines in response to viral protein S (p=0.010), that was lower to proteins M and N (p<0.001 in both cases) ( gure 1B, right). As expected, the number of total cytokines in those classi ed as partial responders was lower compared to convalescents for the response to the three structural viral peptides (p<0.001 in all cases).
The level of each cytokine among uninfected and convalescent HCWs are shown in supplementary gure 2. Strikingly, uninfected strong responders had higher levels of Th2 cytokines IL5 and IL10 (p<0.001 and 0.002, respectively), and similar levels of IL4 (p=0.342) to viral protein S compared to convalescents ( gure 2). To protein M, they had similar levels of IL5 and IL10 (p=0.284 and p=0.115, respectively), and lower levels of IL4 (p=0.001), To protein N, only the levels of IL5 were similar to that found in convalescents, while they were lower for IL10 and IL4 (p=0.020 and p<0.001, respectively). Also, in comparison with convalescents, the level of Th1 cytokine IL12 to protein S was higher (p=0.021), while IFN-γ and TNF-α were similar (p=0.435 and 0.532, respectively), and they were lower to proteins M and N ( gure 2). Finally, the level of IL17A in strong responders was similar to that found in convalescents to protein S and M (p=0.278 and p=0.080), but lower to protein N (p=0.018). In contrast, HCWs with partial response showed lower levels of cytokine production in almost all cytokines, except for IL5, IL10, and IL12 with similar levels to protein S compared to convalescents, and mostly lower to proteins M and N, as shown in gure 2. To graphically show the changes observed in the 10 cytokines according to strong or partial response to the different structural proteins, and in comparison with convalescent, statistically signi cant differences are detailed in gure 3. As it could be observed, there was different cytokine response to protein S among uninfected HCWs with a strong or partial response (p=0.004), especially in Th2 cytokines. We also observed a higher Th1/Th2 ratio higher in convalescents in response to protein S compared to that found in uninfected strong and partial responders (supplementary gure 3). The IFNγ/IL10 ratio in convalescents was 6

Discussion
This study showed that the presence of SARS-CoV-2 cellular immune response in the absence of speci c antibodies could be more important than previously considered for immune protection. In 39 uninfected HCWs largely exposed to COVID-19 patients, we demonstrated a consistent and polyfunctional production of cytokines in response to SARS-CoV-2 peptides, covering viral structural proteins S, M and N, that was similar to that observed in 17 convalescent HCWs, adjusted by age and sex.
These ndings are in agreement with other works focused on SARS-CoV-2-speci c T cells that started to be characterized in COVID-19 patients and uninfected individuals [15][16][17][18][19][20] . Le Bert et al found speci c cellular responses in recovered patients and uninfected individuals mainly to structural protein (NP) and nonstructural proteins (NSP), although NPS response was rarely detected in SARS-COV-2 recovered individuals, measured by the production of IFN-γ and TNF-α by CD4 and CD8 T cells 17  According to the total number of cytokine production after protein S, M, and N stimulation of the PBMCs, we described different pro les. Notably, 61.5% of the uninfected HCWs were classi ed as having a strong response, and 33.3% a partial response. This high proportion of cellular response have to be taken in the context of subjects having been exposed to more than eight weeks, when most of the infectionsusceptible individuals have already got infected. Furthermore, even a partial response could have been enough to prevent infection since a response to SARS-CoV-2 protein S was observed in most of them.
Unfortunately, a predictive clinical or soluble factor to identify individuals with speci c immune response was not found.
The pre-existing cellular responses could have been induced by cross-reaction to seasonal endemic coronaviruses, such as OC43, HKU1, NL63, and 229E that present different degree of amino acid homology [21][22][23][24] . Immunity to these coronaviruses appears to be short-lived as antibody titers decay at 4-12 months after infection 25 . Nevertheless, re-infection with these coronaviruses can occur repeatedly within a single year, eventually activating the cross-reactive immunity. The role of preexisting SARS-CoV-2 reactive cells as a correlate of protection is somehow unclear and needs to be addressed in prospective and larger studies including individuals not exposed to SARS-CoV-2 infection. On the other hand, our data could suggest the development of a cellular response after contact with SARS-CoV-2 in this so high exposed population.
In the recently published ENE survey, among those who reported COVID-19 symptoms before the study only 16.9% had speci c antibodies 26 . The presence of cellular response to structural and/or accessory proteins suggesting a past SARS-CoV-2 infection was described in six out of eight household contacts, in the absence of antibodies production 27 . The lack of association of strong response with the time of exposure or handling aerosol-generating procedures does not invalidate this hypothesis, since the probability of infection is multifactorial. Moreover, the different immune response observed in uninfected HCWs, with predominance of Th2 response, could suggest a response to an aborted infection.
Interestingly, among uninfected strong responders, a higher anti-in ammatory cytokine production (IL5 and IL10) to protein S, and a similar pro-in ammatory response (IFN-γ and TNF-α) or even higher (IL12) was observed, compared to convalescents.
Of great interest, we demonstrate that the immune response is mostly driven against protein S, both in convalescent and uninfected HCW, especially in those with a strong response. These data support the role of this structural protein as the basis for vaccine development, as even subjects with partial response in our study, having a low response to protein M and N, are probably protected because of the response to this protein S.
Our study has limitations, including the cross-sectional design that limits the association between the exposure and the immune response. Also, there was a selection bias towards the inclusion of subjects, by including highly and prolonged exposed HCWs with negative speci c serology as the population of the study. Finally, speci c CD4 and CD8 T cell responses was not determined yet in this study. Instead, a wide spectrum of cytokines produced by many different immune cells was quanti ed to better understand the cytokine changes found in many SARS-CoV-2 infected patients, and how they may play a role in the immune system.
In summary, this study con rms the presence of SARS-CoV-2-speci c cellular immunity in seronegative highly-exposed individuals, and this can be e ciently protective. This is an important issue in terms of recommendations of the general population since herd immunity is crucial for estimating the risk of reinfections. Nonetheless, it is of great importance to investigate whether this cellular immunity is longlasting.    Cytokine level increments compared to convalescents according to strong (SR), and partial (PR) cellular responders to speci c SAR-CoV-2 peptides S, M and N. Each cytokine level in each of the groups has been labeled as higher, similar lower or undetectable compared to the levels found in convalescents. Cytokine

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.