Analysis of Co-inhibitory Receptor Expression in COVID-19 Infection Compared to Acute Plasmodium falciparum Malaria: LAG-3 and TIM-3 Correlate With T Cell Activation and Course of Disease

Coronavirus disease 2019 (COVID-19) which is caused by the novel SARS-CoV-2 virus is a severe flu-like illness which is associated with hyperinflammation and immune dysfunction. The virus induces a strong T and B cell response but little is known about the immune pathology of this viral infection. Acute Plasmodium falciparum malaria also causes acute clinical illness and is characterized by hyperinflammation due to the strong production of pro-inflammatory cytokines and a massive activation of T cells. In malaria, T cells express a variety of co-inhibitory receptors which might be a consequence of their activation but also might limit their overwhelming function. Thus, T cells are implicated in protection as well as in pathology. The outcome of malaria is thought to be a consequence of the balance between co-activation and co-inhibition of T cells. Following the hypothesis that T cells in COVID-19 might have a similar, dual function, we comprehensively characterized the differentiation (CCR7, CD45RO) and activation status (HLA-DR, CD38, CD69, CD226), the co-expression of co-inhibitory molecules (PD1, TIM-3, LAG-3, BTLA, TIGIT), as well as the expression pattern of the transcription factors T-bet and eomes of CD8+ and CD4+ T cells of PBMC of n = 20 SARS-CoV-2 patients compared to n = 10 P. falciparum infected patients and n = 13 healthy controls. Overall, acute COVID-19 and malaria infection resulted in a comparably elevated activation and altered differentiation status of the CD8+ and CD4+ T cell populations. T effector cells of COVID-19 and malaria patients showed higher frequencies of the inhibitory receptors T-cell immunoglobulin mucin-3 (TIM-3) and Lymphocyte-activation gene-3 (LAG-3) which was linked to increased activation levels and an upregulation of the transcription factors T-bet and eomes. COVID-19 patients with a more severe disease course showed higher levels of LAG-3 and TIM-3 than patients with a mild disease course. During recovery, a rapid normalization of these inhibitory receptors could be observed. In summary, comparing the expression of different co-inhibitory molecules in CD8+ and CD4+ T cells in COVID-19 vs. malaria, there is a transient increase of the expression of certain inhibitory receptors like LAG-3 and TIM-3 in COVID-19 in the overall context of acute immune activation.

Coronavirus disease 2019  which is caused by the novel SARS-CoV-2 virus is a severe flu-like illness which is associated with hyperinflammation and immune dysfunction. The virus induces a strong T and B cell response but little is known about the immune pathology of this viral infection. Acute Plasmodium falciparum malaria also causes acute clinical illness and is characterized by hyperinflammation due to the strong production of pro-inflammatory cytokines and a massive activation of T cells. In malaria, T cells express a variety of co-inhibitory receptors which might be a consequence of their activation but also might limit their overwhelming function. Thus, T cells are implicated in protection as well as in pathology. The outcome of malaria is thought to be a consequence of the balance between co-activation and co-inhibition of T cells. Following the hypothesis that T cells in COVID-19 might have a similar, dual function, we comprehensively characterized the differentiation (CCR7, CD45RO) and activation status (HLA-DR, CD38, CD69, CD226), the co-expression of co-inhibitory molecules (PD1, TIM-3, LAG-3, BTLA, TIGIT), as well as the expression pattern of the transcription factors T-bet and eomes of CD8 + and CD4 + T cells of PBMC of n = 20 SARS-CoV-2 patients compared to n = 10 P. falciparum infected patients and n = 13 healthy controls. Overall, acute COVID-19 and malaria infection resulted in a comparably elevated activation and altered differentiation status of the CD8 + and CD4 + T cell populations. T effector cells of COVID-19 and malaria patients showed higher frequencies of the inhibitory receptors T-cell immunoglobulin mucin-3 (TIM-3) and Lymphocyte-activation gene-3 (LAG-3) which was linked to increased activation levels and an upregulation of the transcription factors T-bet and eomes. COVID-19 patients with a more severe disease course showed higher levels of LAG-3 and TIM-3 than patients with a mild disease course. During recovery, a rapid normalization of these inhibitory receptors could be observed. In summary, comparing the expression of different co-inhibitory molecules in CD8 + and CD4 + T

INTRODUCTION
Since December 2019, when the emerging Coronavirus SARS-CoV-2 was first described in Wuhan, China, a rapid increase in cases and deaths worldwide could be recorded. The virus, responsible for Coronavirus disease 2019 (COVID- 19), belongs to the coronavirus family and causes respiratory tract infections. SARS-CoV-2 is a 29.903 nucleotide large, single-strand RNA virus coding for four structural proteins: the spike protein, the membrane protein, the envelope protein, and the nucleocapsid protein (1,2). Previous studies have shown that recovered COVID-19 patients developed anti-spike-protein immunity (3).
While most patients are asymptomatic or display only mild symptoms such as fatigue, fever, and dry coughs, some individuals develop pneumonia, severe acute respiratory distress syndrome, sepsis, and septic shock with an overall case fatality rate of 5% (4)(5)(6). In a subset of patients, COVID-19 is associated with a so-called cytokine storm, lymphopenia, and dysregulation of the immune system, yet the underlying mechanisms and cellular sources of these immunological complications remain to be fully understood (7,8). First studies have shown that reduced T cell counts correlated with disease severity in COVID-19 patients (9). Broadly directed, antigen-specific T cell responses could be detected in the effector and central memory subsets of CD8 + and CD4 + T cells of COVID-19 patients (1,10). CD8 + T cells predominantly produced IFN-γ, while CD4 + T cells produced Th1 and Th2 cytokines (1,10). The exact role and function of T cells in COVID-19 patients with regards to immunological complications like cytokine storm, however, needs to be elucidated in further studies (11).
From other viral infections, we know that cytotoxic T cells play an important role in killing virus-infected cells and thus contribute to the clearance of the infection (12,13). Detailed studies of the phenotype and function of virus-specific T cells are important to understand their potential role in the pathophysiology of the disease and to develop future treatment strategies (14)(15)(16). In order to induce effector T cells with high proliferative and cytotoxic activity and to prevent excessive host immune responses, a balanced T cell response is shaped by the simultaneous upregulation of co-stimulatory and coinhibitory receptors (17,18). Inhibitory receptors and their role in T cell exhaustion have been extensively studied in cancer as well as chronic infections such as HCV and HIV (14,(19)(20)(21)(22)(23). However, in the early stages of an acute infection, the kinetics and role of co-inhibitory molecules are not well-understood. It is not clear how early T cell exhaustion and loss of effector function set in and whether exhaustion plays a role in the pathophysiology of acute infections (9,17,24,25). Previous studies of T cells and their role in COVID-19 patients have shown an upregulation of inhibitory receptors like PD1. Therefore, it has been suggested that T cell exhaustion might play a role in the pathophysiology of COVID-19 infection (9). However, an upregulation of inhibitory receptors in acute infections does not necessarily correlate with terminal exhaustion of these cells but can be regarded rather as a characteristic of the overall immune activation to counterbalance excessive immune responses (17). A similar, strong up-regulation of a combination of several coinhibitory receptors in the context of massive T cell activation and secretion of pro-inflammatory cytokines has been observed in acute malaria (26)(27)(28). It has been shown that the expression of these co-inhibitory receptors has a double-edged role with potentially detrimental, but also beneficial effects (26,(29)(30)(31)(32): on the one hand, over-expression of inhibitory receptors can hinder the hosts ability to clear the infection. On the other hand, this inhibition of immune cells can also dampen hyperinflammation by down-regulating T cell effector functions (29,31). These studies highlight the importance of intricate modulation of the T cell response in malaria rather than supporting the concept of development of terminal T cell exhaustion at early stages of infection. We hypothesize that T cells are modulated in a similar way in COVID-19 and thus compared the pattern of coinhibitory receptor expression in malaria vs. COVID-19. In the case of immunological analysis of patients with COVID-19, it will be important to assess the level of co-expression and the ranges of the different co-inhibitory receptors to understand their adverse effects and interpret their role in the overall context of acute infection (33,34). Comparing the T cell response in COVID-19 patients not only with samples of healthy individuals but also with samples of patients with other acute infections will help to identify the specific immune signature of COVID-19 infection.
In this current study we comprehensively examined the T cell expression profile of inhibitory and stimulatory receptors as well as the expression of the transcription factors T-bet and eomes in a cohort of n = 20 COVID-19 and n = 10 malaria patients in order to obtain a more detailed understanding of the nature of T cells in the context of acute infection and disease severity. Both, acute infection with SARS-CoV-2 and Plasmodium falciparum, can cause excessive host immune activation that can in terms cause severe damage and even limit the hosts ability to clear the disease (8,28).

Patient Cohort
PBMC (peripheral blood mononuclear cells) of SARS-CoV-2 infected patients (n = 20) and P. falciparum infected patients Data are expressed as absolute numbers n (n/N) or n (range) or mean with standard deviation.
(n = 10) as well as uninfected healthy individuals (n = 13) were collected at the University Medical Center Hamburg Eppendorf. Blood samples of P. falciparum infected patients and healthy controls were collected prior to the COVID-19 outbreak. The study was approved by the local ethics board of the Ärztekammer Hamburg (PV4238, PV4780, PV7298) and written consent was obtained by all study participants. SARS-CoV-2 infection was verified by RT-PCR of nasopharyngeal swabs as previously described (35). Plasmodium falciparum infection was confirmed microscopically at the Bernhard-Nocht-Institute for Tropical Medicine. Thick and thin blood smears were stained with 4% Giemsa and examined under oil immersion (original magnification × 100). Clinical and laboratory data were obtained by structured chart review.

Intracellular Staining and Flow Cytometry
Intracellular as well as surface staining was performed as previously described (36). Cryopreserved PBMC were thawed and stained with the LIVE/DEAD TM Fixable Near-IR dye (Thermo Fisher, Schwerte, Germany). Cells were then stained with the following surface antibodies

Statistical Analysis
Flow cytometric data was analyzed with FlowJo version 10 (Treestar, Ashland, OR, USA). Statistical analysis was performed using the GraphPad Prism 8 software (GraphPad Software, San Diego, CA). Unpaired groups were analyzed using the Mann-Whitney test while paired groups were analyzed with the Wilcoxon test. For bivariate correlation analysis the Spearman correlation was applied. Data are expressed as mean with standard deviation. P < 0.05 were considered to be significant.

Clinical Characteristics of the COVID-19 and Malaria Cohort
The clinical cohorts consisted of 20 SARS-CoV-2 and 10 P. falciparum infected patients and 13 healthy individuals. All patients were admitted to the University Medical Center Hamburg Eppendorf due to the severity of their symptoms. None of the patients needed to be admitted to the intensive care unit at the time of blood collection. During the course of the hospital stay, a total of three COVID-19 patients were transferred to the intensive care unit, one of whom unfortunately later died due to a pulmonary embolism. On average, the COVID-19 patients spent 9.9 days and the malaria patients spent 4.9 days in the hospital. The mean age of the COVID-19 patients was 54 years compared to 55 years of patients with malaria. Of note, one COVID-19 patient had received treatment with Rituximab 3 months prior to blood sample collection and another patient had received treatment with a PD-1 antibody several weeks prior to the infection with SARS-CoV-2. We enrolled PBMC of 13 healthy volunteers into our study, eight female and five male with a mean age of 35 ± 8. The analysis of lymphocyte subsets in the clinical charts showed a significant generalized decrease of the T lymphocyte count affecting both CD8 + and CD4 + T cells with a resulting normal CD4/CD8 ratio in COVID-19 patients ( Table 1). In accordance with these results we found a significantly lower T lymphocyte frequency as measured by CD3 + frequency in our FACS analysis in samples of COVID-19 patients but not in malaria patients ( Figure 1A). A more detailed description of the cohort is given in Table 1.

Acute COVID-19 and Malaria Infection Result in a Comparably Elevated Activation Status of the CD8 + and CD4 + T Cell Populations
For a description of the activation status, we assessed the frequencies of CD69 and HLA-DR/CD38 on CD8 + and CD4 + T cells in COVID-19 and malaria patients as well as in healthy controls. There were increased levels of CD69 and HLA-DR/CD38 co-expression on CD8 + and CD4 + T cells in both COVID-19 and malaria patients compared to healthy controls which is characteristic for acute infections ( Figure 1B) (18,37,38). The gating strategy for CD8 + and CD4 + T cells is illustrated in Supplementary Figure 3.
Based on the established differentiation markers CCR7, CD45RO, and CD27 (39), we defined naïve (CCR7 + CD45RO − ), central memory CM (CCR7 + CD45RO + ), transitional memory TM (CCR7 − CD45RO + CD27 + ), effector memory EM (CCR7 − CD45RO + CD27 − ) and terminal effector memory EMRA (CCR7 − CD45RO − ) CD8 + and CD4 + T cells and assessed the frequency of the activation markers on each subset (39). Overall, we saw an increase of CD8 + EM T cells and CD8 + and CD4 + CM T cells in both COVID-19 and malaria compared to healthy individuals ( Figure 1C). Representative dot plots for the gating of naive and memory subsets of CD8 + and CD4 + T cells can be seen in Figure 1D. CD69 was more frequently expressed on CM, TM, and EMRA CD8 + and CD4 + T cells in COVID-19 and malaria patients compared to the respective subsets of T cells in healthy individuals. The same statistical trend toward higher CD69 frequencies was evident in the EM subset of CD8 + and CD4 + T cells in COVID-19 and malaria patients compared to healthy controls (Figure 2A). Looking at the activation as measured by HLA-DR and CD38 expression, we observed an increase of HLA-DR + CD38 + cells in the TM and EM subset of CD8 + T cells and in the CM and TM subset of CD4 + T cells in both COVID-19 and malaria patients compared to healthy controls. Of note, CD8 + CM, TM, and EM T cells showed particularly high frequencies of HLA-DR + CD38 + cells in COVID-19 and patients infected with P. falciparum (Figure 2A). Together with the overall increase of CM and EM CD8 + T cells in COVID-19, this could imply an important role for these cells during the acute infection.
In addition to the analysis of the traditional activation markers, we also looked at the expression of the co-stimulatory receptor CD226 on CD8 + and CD4 + T cells in COVID-19 and malaria patients compared to healthy individuals. The CD226 molecule enhances cytotoxic effector functions when upregulated on CD8 + T cells (14,40,41). Indeed, a significantly higher frequency of CD226 was detectable on bulk CD8 + T cells in both COVID-19 and malaria compared to healthy individuals. In COVID-19 patients, the TM, EM, and EMRA CD8 + T cell subsets showed a higher frequency of CD226 compared to healthy controls, while in malaria patients the upregulation of CD226 was significant across all differentiation subsets compared to healthy individuals which implies a highly activated and functional phenotype of these cells ( Figure 2B and Supplementary Figure 4A). Representative dot plots for the gating of CD69, HLA-DR/CD38 and CD226 on CD8 + and CD4 + T cells are illustrated in Figure 2C.
Overall, we observed a strong activation of T cells in COVID-19 and malaria with an increased expression of both early (CD69 + ) and late (HLA-DR + CD38 + ) activation markers as well as the co-stimulatory receptor CD226.

COVID-19 and Acute Malaria Infection Are
Accompanied by Elevated Frequencies of LAG-3 and TIM-3 on CD8 + and CD4 + T Cells Next, we examined the expression of the inhibitory receptors PD1, TIGIT, LAG-3, TIM-3, and BTLA of T cells in COVID-19 and malaria compared to healthy controls. Previous research has established inhibitory receptors to play a critical role in T cell exhaustion in chronic infections after prolonged antigen exposure (14,15,42,43). However, inhibitory receptors also have important functions and are transiently expressed during acute infections (17,24,(44)(45)(46). Many studies have described an increased expression of inhibitory receptors on both CD8 + and CD4 + T cells during malaria infection (26,27,47,48). Whether this increase has a beneficial or damaging effect, however, is not yet fully understood. To see whether similar effects can be observed in COVID-19, we assessed the frequencies of the coinhibitory receptors PD1, TIGIT, BTLA, LAG-3, and TIM-3 on CD8 + and CD4 + T cells in COVID-19 and malaria patients and compared them with healthy individuals (Figure 3A). We also assessed the expression of these inhibitory receptors on the different memory subsets to detect potential influences on the expression due to a change in subset distribution ( Figure 3B and Supplementary Figure 4).
A trend toward higher PD1 expression on CD8 + and CD4 + T cells in COVID-19 could be observed although with no statistical significance ( Figure 3A). However, malaria patients showed a strong upregulation of PD1 on both CD8 + and CD4 + T cells ( Figure 3A). This increase of PD1 on T cells in malaria was significant across all memory subsets of CD4 + T cells and on naïve and EMRA CD8 + T cells (Supplementary Figure 4B). The mean fluorescence intensity (MFI) of PD1 showed an increased expression of PD1 on bulk CD8 + T cells in COVID-19 and malaria patients compared to healthy individuals. In contrast, while we could not see any difference in the PD1 MFI on CD4 + T cells between COVID-19 patients and healthy donors, malaria patients showed a significantly increased PD1 MFI on bulk, naïve, and all CD4 + memory T cell subsets compared to healthy donors. Interestingly, the EM subset of CD8 + T cells, which presumably includes a high proportion of SARS-CoV-2-specific T cells (10), showed a comparable PD1 MFI in healthy donors and COVID-19 and malaria patients (Supplementary Figure 4C).
We did not detect increased TIGIT expression on bulk T cells in COVID-19 patients compared to healthy controls ( Figure 3A). The EM CD4 + T cell subset in COVID-19 even showed a significant decrease of TIGIT expression. The same trend was evident on CD8 + EM T cells although this trend did not reach statistical significance. TIGIT expression in malaria patients was significantly increased on bulk CD8 + and CD4 + T cell populations. However, when looking at the expression pattern of TIGIT on different memory subsets, only TM CD4 + T cells showed a significantly higher TIGIT frequency (Supplementary Figure 4D). Overall it seems that in acute COVID-19 infection there is a trend toward a decreased TIGIT frequency on T cells compared to healthy donors while T cells in acute malaria patients generally showed a trend toward higher TIGIT frequencies compared to healthy individuals.
BTLA inhibits lymphocyte effector functions during immune response, similar to PD1 (49). It is well-established that BTLA expression is downregulated during CD8 + and CD4 + T cell differentiation, with the highest expression on naïve T cells (49). Indeed, a lower BTLA frequency in memory compared to naive subsets of CD8 + and CD4 + T cells could be detected in COVID-19 and malaria patients as well as healthy donors ( Figure 3A). However, the decrease of BTLA on CD8 + T cell subsets detected in both, COVID-19 and malaria, was not as strong as in healthy subjects. We observed a significantly higher expression of BTLA on TM and EM CD8 + T cells in COVID-19 and malaria patients compared to healthy donors ( Figure 3A).
Lastly, we looked at the expression of LAG-3 and TIM-3. Both inhibitory receptors have previously been described to be upregulated during acute malaria infection with complex effects on T cell function (26). Accordingly, we observed a significant increase of LAG-3 and TIM-3 on CD8 + and CD4 + T cells in malaria patients when compared to healthy donors. In COVID-19 patients we saw a similar upregulation of LAG-3 and TIM-3 on T cells of even higher frequencies on CD8 + T cells (Figure 3A). A significant increase of LAG-3 and TIM-3 on both CD8 + and CD4 + T cells across all subpopulations in COVID-19 patients could be detected (Figure 3B).
To further characterize the expression pattern of LAG-3 and TIM-3 we examined the co-expression of PD1/LAG-3 and PD1/TIM-3 on CD8 + and CD4 + T cells of COVID-19 and malaria patients compared to healthy controls ( Figure 3C). Interestingly, we found differences in the distribution of these receptors, not only when comparing T cells of COVID-19 and malaria patients with T cells of healthy donors, but also when comparing COVID-19 with malaria. CD8 + and CD4 + T cells of COVID-19 patients showed high frequencies of single-positive PD1 − LAG-3 + and PD1 − TIM-3 + CD8 + and CD4 + T cells with a significant increase compared to healthy controls. However, in PBMC of malaria patients we only observed an increase of PD1 − LAG-3 + CD8 + T cells when compared to the frequency of PD1 − LAG-3 + CD8 + and CD4 + T cells in healthy controls. Double-positive PD1 + LAG-3 + and PD1 + TIM-3 + cells were increased in both COVID-19 and malaria when compared with healthy donors. Looking at the single-positive PD1 + LAG-3 − and PD1 + TIM-3 − cells we found no difference between COVID-19 patients and healthy donors in the CD8 + T cell compartment. Only CD4 + T cells in COVID-19 showed a higher frequency of PD1 + LAG-3 − and PD1 + TIM-3 − T cells compared to healthy donors although the frequency of these cells was highest in malaria with a significant increase of both PD1 + LAG-3 − and PD1 + TIM-3 − on CD8 + and CD4 + T cells compared to healthy controls. Overall, we saw a shift toward double-positive PD1 + LAG-3 + and PD1 + TIM-3 + and singlepositive PD1 − LAG-3 + and PD1 − TIM-3 + cells in the CD8 + T cell subset of COVID-19 patients whereas in malaria, the distribution changed more toward double-positive PD1 + LAG-3 + and PD1 + TIM-3 + and single-positive PD1 + LAG-3 − and PD1 + TIM-3 − T cells. Lastly, we looked for PD1 + LAG-3 + TIM-3 + T cells. However, the frequency of these PD1 + LAG-3 + TIM-3 + T cells was comparably low ranging from only 0.013-0.76% of CD8 + T cells in COVID-19 (data not shown).
Overall, we saw a strong increase of LAG-3 and TIM-3 expression on T cells in COVID-19 and malaria. PD1 was only slightly increased on T cells in COVID-19 patients while the frequency of TIGIT on T cells did not differ between COVID-19 patients and healthy controls. In malaria we observed an upregulation of both PD1 and TIGIT on CD8 + and CD4 + T cells compared to healthy donors.

Co-expression of PD1, LAG-3, and TIM-3 on T Cells Depends on the Degree of T Cell Activation
T cells of patients with COVID-19 and acute malaria showed high frequencies of the activation markers CD69 and HLA-DR/CD38 as well as high frequencies of the inhibitory receptors LAG-3 and TIM-3. To characterize whether stimulatory and inhibitory receptors are co-expressed on, or limited to distinct T cells we analyzed the expression of CD69 and the co-expression of HLA-DR/CD38 on PD1 − LAG-3 − and PD1 − TIM-3 − T cells and compared it to the expression on double-positive PD1 + LAG-3 + and PD1 + TIM-3 + T cells in COVID-19 and malaria patients. Double-negative T cells did not co-express HLA-DR and CD38 while double-positive T cells showed high frequencies of HLA-DR + CD38 + CD8 + and CD4 + T cells (Figure 4A). This upregulation of HLA-DR/CD38 on T cells co-expressing PD1 and LAG-3 and PD1 and TIM-3 could be observed on both CD8 + and CD4 + T cells in COVID-19 and malaria patients. Similarly, the frequency of CD69 was lower on PD1 − LAG-3 − and PD1 − TIM-3 − CD8 + and CD4 + T cells compared to the expression on double-positive T cells in COVID-19 and malaria patients ( Figure 4B). In summary, the expression of activation markers (CD69, HLA-DR, CD38) on T cells was closely linked to the co-expression of inhibitory receptors (PD1, LAG-3, TIM-3) on T cells in COVID-19 and malaria patients.

Upregulation of T-bet and Eomes in T Cells That Co-express Inhibitory Receptors in COVID-19 and Acute Malaria Infection
In a next step, we analyzed the transcription factors T-bet and eomes in COVID-19 and malaria patients. For this, we stained PBMC with the transcription factors T-bet and eomes and analyzed their expression in T cells in relation to the co-inhibitory receptors PD1 and LAG-3. The transcription factors T-bet and eomes have important functions in the immune response to acute infections and are upregulated to preserve cytotoxic effector functions in T cells (50)(51)(52)(53)(54). In chronic settings it has been shown that exhaustion and functional impairment of T cells is accompanied by an upregulation of eomes and downregulation of T-bet in exhausted T cells (22,55,56). Looking at the expression of T-bet and eomes in CD8 + and CD4 + T cells in COVID-19 and malaria patients we saw an upregulation of T-bet in COVID-19 and an upregulation of both T-bet and eomes in malaria ( Figure 4C). Next, we aimed to look at T-bet and eomes in the context of the co-inhibitory receptors PD1 and LAG-3. For this, we compared the expression of T-bet and eomes in PD1 + LAG-3 + T cells to that in PD1 − LAG-3 − T cells in COVID-19 and malaria ( Figure 4D). Similar to the expression of activation markers on these cells, T-bet and eomes were both more frequently expressed in PD1 + LAG-3 + T cells than in their double-negative counterparts. Altogether, the simultaneous upregulation of T-bet and eomes in T cells co-expressing PD1 and LAG-3 further supports the notion that an upregulation of coinhibitory receptors can be seen as a result of the overall immune activation during COVID-19 and acute malaria infection.
Additionally, we examined the frequency of CD127 expression on CD8 + T cells in COVID-19 and malaria patients in relation to the expression of the co-inhibitory receptors PD1 and LAG-3. CD127 as part of the IL-7 receptor is hugely important for the survival and differentiation of T cells and is involved in memory development (57,58). CD127 low T cells are considered short-lived effector cells that undergo apoptosis after antigen clearance (59). In some viral infections such as HIV and HCV the downregulation of CD127 on bulk CD8 + T cells has been demonstrated to have damaging effects on viral control and promote T cell exhaustion (60,61). Bulk CD8 + T cells of COVID-19 and malaria patients, however, showed no difference in CD127 expression compared with healthy individuals (Supplementary Figure 5A). CD127 on PD1 + LAG-3 + CD8 + Tells in COVID-19 and malaria was downregulated compared to their double-negative PD1 − LAG-3 − counterparts (Supplementary Figure 5B).

Detection of High Expression of LAG-3 and TIM-3 on T Cells in Patients With Severe COVID-19 and Rapid Normalization After Clinical Recovery
To further evaluate whether the expression of LAG-3 and TIM-3 on T cells was related to the severity and disease course in COVID-19, we sub-stratified patients into a mild and a severe  Table 1). Patients with mild disease course were defined to have a maximum hospital stay of five days and absence of any complications. Severe disease course was defined as hospital stay longer than 20 days, oxygen saturation lower than 85%, transfer to the ICU or death. According to this definition, six patients fitted into the mild group and five patients fitted into the severe group. Next, we compared the activation and inhibitory marker expression on T cells between these two groups. Interestingly, we saw a higher expression of LAG-3 and TIM-3 on CD8 + and CD4 + T cells in patients with a severe disease course compared to patients with a mild disease course ( Figure 5A). However, there was no significantly higher expression of other co-inhibitory receptors on T cells in the severe group compared to the mild group (Supplementary Figure 6A). Additionally, we also compared the peak levels of the clinical inflammation parameters C-reactive protein (CRP), IL-6, or ferritin with the expression of activation and inhibitory receptors on CD8 + and CD4 + T cells in COVID-19. CD69 on CD8 + T cells and TIM-3 on CD4 + T cells correlated with CRP levels (Figure 5B). A similar trend of correlation between CRP levels and the frequency of TIM-3 on CD8 + T cells could be observed. IL-6 and ferritin levels did not correlate with the expression of co-inhibitory molecules (data not shown).
Finally, we investigated the expression kinetics of inhibitory receptors during the course and recovery of COVID-19 infection. For this, PBMC of three COVID-19 patients were collected at a second time point on day 13, 15, or 20 after the first sample date. All three patients were discharged in good health from the hospital shortly after the second sample collection. Frequencies of inhibitory receptor expression on CD8 + and CD4 + T cells were analyzed and compared to those measured at the first time point. A rapid decrease and normalization in all three patients for LAG3, TIM3, and PD1 expression on CD8 + and CD4 + T cells was observed, independently of the individual expression level measured at the first time point (Figure 5C).

DISCUSSION
One of the main results of this cross-sectional side-by-side comparison of co-inhibitory and co-stimulatory receptors on T cells isolated from PBMC of patients with acute COVID-19 and malaria infection was the comparable T cell activation and phenotype observed, following SARS-CoV-2 and P. falciparum infection. Of interest, a strong upregulation of LAG-3 and TIM-3 on T cells in both COVID-19 and malaria patients was linked to the co-expression of activation markers and correlated with disease course (Figures 3-5) (26). LAG-3 and TIM-3 are coinhibitory receptors with multiple functions in modulating proinflammatory T cell responses (62).
Elevation and sustained co-expression of inhibitory receptors such as PD1, LAG-3, TIM-3, CTLA4, and BTLA is the hallmark of exhausted T cells (14,20,42). This unique T cell population is the result of persisting antigen stimulation found in cancer and chronic infections. Previous studies of T cells in malaria have shown that the upregulation of co-inhibitory molecules such as LAG-3 and TIM-3 in response to acute infection can have both beneficial and detrimental effects (29)(30)(31)(32). It is still highly debated whether T cell exhaustion can occur during early malaria infection or other acute viral infections since the expression of inhibitory receptors is also induced through T cell activation and differentiation (17,24,27,63). Legat et al. showed that stimulating human CD8 + T cells induced a strong and rapid upregulation of inhibitory receptors after 24 h while cytokine production was not necessarily affected (17). Additionally, in a murine LCMV model it was demonstrated that irreversible T cell exhaustion occurred only after about 30 days of infection (25).
Our data demonstrate a strong upregulation of LAG-3 and TIM-3 on T cells in COVID-19 and malaria patients across all differentiation stages. In COVID-19 patients we found high frequencies of single-positive PD1 − LAG-3 + and PD1 − TIM-3 + CD8 + and CD4 + T cells and fewer doublepositive PD1 + LAG-3 + and PD1 + TIM-3 + T cells. However, in malaria patients we found high levels of single-positive PD1 + LAG-3 − and PD1 + TIM-3 − as well as high levels of double-positive PD1 + LAG-3 + and PD1 + TIM-3 + CD8 + and CD4 + T cells. Triple-positive PD1 + LAG-3 + TIM-3 + T cells were nearly absent in both COVID-19 and malaria patients. Double-positive cells were highly activated compared to their double-negative counterparts and showed increased levels of the transcription factors T-bet and eomes in COVID-19 and malaria. A previous study by Niu et al. demonstrated that the expression of PD1 and LAG-3 on T cells in patients with acute sepsis had a regulatory impact on cytokine production (64). The degree of inhibition was dependent on the expression pattern of PD1 and LAG-3 on T cells. In comparison, single-positive PD1 − LAG-3 + T cells showed the highest cytokine production followed by single-positive PD1 + LAG-3 − T cells while doublepositive PD1 + LAG-3 + T cells produced the least amount of cytokines (64).
Overall, when comparing the activation and expression signatures of different co-inhibitory receptors in COVID-19 and malaria patients we found relatively subtle differences between T cells of patients with either infection. The activation markers CD69 and HLA-DR/CD38 were elevated on T cells in COVID-19 and malaria patients. The upregulation of the co-inhibitory receptors LAG-3 and TIM-3 was more pronounced on CD8 + T cells in COVID-19 patients while malaria patients showed especially high expression of LAG-3 and TIM-3 on CD4 + T cells. In line with this, acute malaria infection has been shown to result in a strong polarization of CD4 + T cells (26) while it can be speculated that there seems to be a polarization toward CD8 + T cells in COVID-19. Furthermore, T cells of malaria patients showed even higher frequencies of PD1 and TIGIT than COVID-19 patients and healthy donors ( Figure 3A). Overall, our data show a rather similar T cell signature of co-inhibitory molecules in COVID-19 and malaria, an infection in which the upregulation of co-inhibitory receptors is a result of their strong activation (26,27). However, acute malaria generally seemed to result in an upregulation and co-expression of a broader range of co-inhibitory receptors on T cells ( Figure 3A).
Importantly, we also observed a strong upregulation of the transcription factors T-bet and eomes in T cells co-expressing inhibitory receptors. T-bet and eomes are important mediators during acute infections and promote effector functions and cytotoxicity in CD8 + and CD4 + T cells (50)(51)(52)(53). T-bet drives the expansion of cytokine producing Th1 cells and inhibits the formation of Th2 and Th17 cells (50,51). In this context T-bet has also been shown to repress the expression of PD1 and counterregulate the Th1 response by inducing TIM-3 on CD4 + T cells. In CD8 + T cells T-bet promotes cytotoxicity by inducing the production of granzyme B and perforin (50,52). Eomes is also found in both CD8 + and CD4 + T cells during acute infection. In CD4 + T cells eomes can either induce the production of IFN-γ by Th1 cells or promote Tr1 cells by driving IL-10 production although the exact mechanisms underlying this are not yet fully understood (65,66). In CD8 + T cells eomes is needed for a T-bet independent induction of IFN-γ and later during infection drives the formation of memory cells by aiding the responsiveness to IL-15 (52). Regarding T cell exhaustion, T-bet and eomes have distinct roles compared to those in effector and memory T cells. Exhausted T cells express high levels of eomes and decreased expression of T-bet (22,56,67). We observed an increase of T-bet in bulk CD8 + and CD4 + T cells and high frequencies of both T-bet and eomes in CD8 + and CD4 + T cells coexpressing PD1 and LAG-3 indicating strong effector functions in these cells. Further and more comprehensive studies of the transcription factor profile of T cells in COVID-19 and malaria are needed to understand and interpret the expression pattern of inhibitory receptors.
Recently, the CD226/TIGIT axis has been identified as an important regulator in anti-tumor and anti-viral T cell responses (14). TIGIT expression on CD8 + and CD4 + T cells in COVID-19 patients did not differ greatly from healthy individuals. We even observed a downregulation of TIGIT on some T cell subsets (Supplementary Figure 4B). Considering that TIGIT and CD226 compete for the same ligand CD155 on APCs and binding of CD226 to this ligand promotes cytotoxicity in CD8 + T cells the simultaneous increase of CD226 and decrease of TIGIT expression could benefit CD8 + T cells to maintain cytotoxic effector functions in COVID-19 (40,68). In malaria patients we observed an increase of TIGIT on bulk CD8 + and CD4 + T cells. However, looking at the different T cell subsets no significant changes could be seen which indicates that the increase in TIGIT expression might be linked to the differentiation status and change in distribution away from naïve toward effector cells. However, the overall frequency of TIGIT + T cells in malaria and COVID-19 patients was lower than the frequency observed in chronic settings such as HCV (14).
Expression of LAG-3 and TIM-3 on T cells in COVID-19 patients was closely related to disease course with higher levels detectable in severe patients compared to the mild patient group. Toward recovery of COVID-19 infection, the expression of inhibitory receptors on T cells decreased rapidly. Taken together, our results do not hint toward the development of terminal T cell exhaustion during acute COVID-19 infection. However, whether the upregulation of inhibitory receptors on T cells in COVID-19 has detrimental or beneficial effects needs to be further investigated. Moreover, we only analyzed the bulk expression levels of co-inhibitory molecules of T cells and pathogen-specific T cells might show a different pattern.
On the same note, blockade of the inhibitory receptors LAG-3 and TIM-3 as a therapeutic strategy in COVID-19 or malaria must be critically evaluated (69). On the one hand, it could enhance the effector T cell response to SARS-CoV-2 infection, on the other hand, such an intervention could also increase hyperinflammation. Interestingly, Robilotti et al. describe that therapy with immune checkpoint inhibitors correlated with a more severe outcome in COVID-19 patients (69). Interestingly, two patients of our COVID-19 cohort were previously treated with biologicals (Rituximab and Pembrolizumab) but neither patient showed a severe disease course, increased hyperinflammation or differing T cell pattern in terms of activation, differentiation, or expression of coinhibitory molecules (data not shown). Nonetheless, study of the immune response following SARS-CoV-2 infection in patients receiving immunomodulatory therapy can give us further insight into the role and function of different immune cells.
Disease progression and adverse outcome in COVID-19 seemed to be associated with the degree of hyperactivation leading to cytokine storm and lymphopenia (4). The question remains, whether and in how far T effector cells are involved in the pathophysiology and hyperinflammation of COVID-19. The so-called cytokine storm is known from other infections such as severe acute respiratory syndrome (SARS) and ebola (70,71). High levels of cytokines like IL-6, or Interferon-y could be measured in the serum of COVID-19 patients, hinting toward a dysbalanced, overreacting immune system (4). The underlying mechanisms and driving forces of this cytokine storm in COVID-19 remain to be fully understood. However, from our results and findings in initial COVID-19 studies, it appears that T cells do not play a direct role in the development of hyperinflammation (10,72). The main source of IL-6 does not appear to be T cells but rather monocytes and macrophages (10,73). It can even be speculated whether the upregulation of LAG-3 and TIM-3 is a reaction of the immune system to counteract systemic hyperinflammation. Interestingly, a study on ebola has demonstrated that fatal ebola infections were associated with higher expression of CTLA4 and PD1 on CD8 + and CD4 + T cells and was accompanied by strong inflammation while ebola patients who survived showed less inflammation markers and lower CTLA4 and PD1 expression on T cells (74). In accordance, our data show a correlation between disease severity of COVID-19 and expression of inhibitory receptors like LAG-3 and TIM-3 on T cells.
Our study was limited by the relatively small and heterogeneous cohort and we assessed only few patients longitudinally. In order to further investigate the exhaustive state of the SARS-CoV-2-specific T cell response in COVID-19, intracellular cytokine assays (ICS) after antigen-specific short term stimulation, MHC class I and class II multimer-staining and further functional assays in larger prospective cohorts with broader variations of the clinical course and worse outcome are needed to understand the role and function of antigenspecific cells. Additionally, functional T cell subsets such as follicular helper cells, Th17, and regulatory T cells need to be characterized in parallel. Of note, the frequency of peripheral Tregs as measured by clinical immunology did not differ between patients with COVID-19 and healthy individuals ( Table 1). In our study, we examined the phenotype of peripheral CD8 + and CD4 + T cells. It has been shown that the expression pattern of co-inhibitory receptors on T cells depends on their localization (75). T cells in the peripheral blood can greatly differ in their expression pattern to those in the tissue (76). Considering this, further studies that examine T cells in the lung, liver, and other organs in COVID-19 and malaria are needed.
The direct comparison of the T cell response between an acute viral and an acute parasitic infection does not seem selfevident. However, there are only few acute infections which are accompanied by hyperinflammation with high levels of proinflammatory cytokines and in which the T cell response is regulated by a strong upregulation of a broad spectrum of coinhibitory receptors, such as ebola, SARS, MERS, and malaria (28,71). Interestingly, we found relatively minor differences of the T cell signature and co-expression of co-inhibitory molecules of T cells in the direct comparison of patients with COVID-19 and acute malaria supporting the notion that co-inhibitory molecules have a dual function, with potentially beneficial but also detrimental effects on T cell function (26,29,31,32). Our data suggest that there is a common pattern of T cell regulation in acute infections and that the up-regulation and co-expression of co-inhibitory molecules is as a consequence of their strong activation. Nonetheless, comparisons of the T cell phenotype and functional immune response with other acute viral respiratory infections like influenza are needed in future studies.
In summary, comparing the expression of different coinhibitory molecules of CD8 + and CD4 + T cells in COVID-19 vs. malaria there is a transient increase of the expression of certain inhibitory receptors like LAG-3 and TIM-3 in COVID-19 in the overall context of acute immune activation. During recovery, a rapid normalization of these inhibitory receptors could be observed. The results of this study will be the basis for further analysis regarding T cells and their potential role for immune pathogenesis during COVID-19 infection.

DATA AVAILABILITY STATEMENT
The datasets generated for this study are available on request to the corresponding author.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Ärztekammer Hamburg (PV4238, PV4780, and PV7298). The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
MH, SS, TJ, and JS designed the study. MH, SS, NW, MW, TJ, and JS designed the panels. MH, SS, NW, RW, and MW conducted experiments. MH and SS analyzed the data. JS, TB, and MR recruited the patients. JS supervised the study at all stages. MH, SS, TJ, and JS wrote the first draft. JS and AL gave institutional support. All authors reviewed the manuscript and gave important input.

FUNDING
This project has been funded by the Deutsche Forschungsgemeinschaft SFB841 (SS, JS, TJ, NW, and AL) and SFB1328 (JS) and Deutsches Zentrum für Infektionsforschung DZIF (MH, JS, TJ, TB, and MW). The funders had no role in study design, data collection, and analysis, decision to publish, or preparation of the manuscript.