key: cord-0722769-5nxyso3l
authors: Capelle, Christophe M.; Ciré, Séverine; Domingues, Olivia; Ernens, Isabelle; Hedin, Fanny; Fischer, Aurélie; Snoeck, Chantal J.; Ammerlaan, Wim; Konstantinou, Maria; Grzyb, Kamil; Skupin, Alexander; Carty, Cara L.; Hilger, Christiane; Gilson, Georges; Celebic, Aljosa; Wilmes, Paul; Del Sol, Antonio; Kaplan, Ian M.; Betsou, Fay; Abdelrahman, Tamir; Cosma, Antonio; Vaillant, Michel; Fagherazzi, Guy; Ollert, Markus; Hefeng, Feng Q.
title: Combinatorial analysis reveals highly coordinated early-stage immune reactions that predict later antiviral immune responses in mild COVID-19 patients
date: 2022-03-29
journal: Cell Rep Med
DOI: 10.1016/j.xcrm.2022.100600
sha: 560403c898b52a59effb1ad76fe2af7261739179
doc_id: 722769
cord_uid: 5nxyso3l

While immunopathology has been widely studied in severe COVID-19 patients, immune responses in non-hospitalized patients have remained largely elusive. We systematically analyze 484 peripheral cellular or soluble immune features in a longitudinal cohort of 63 mild and 15 hospitalized patients versus 14 asymptomatic and 26 household controls. We observe a transient increase of IP10/CXCL10 and IFN-β levels, coordinated responses of dominant SARS-CoV-2-specific CD4 and less CD8 T cells, various antigen presenting and antibody-secreting cells in mild patients within three days of PCR diagnosis. Both the frequency and expression of key functional markers of innate immunity are impaired in hospitalized patients at day 1 of inclusion. T-cell and dendritic-cell responses at day 1 are highly predictive for SARS-CoV-2-specific antibody responses after three weeks in mild but not hospitalized patients. Our systemic analysis reveals a combinatorial picture and trajectory of various arms of the highly-coordinated early-stage immune responses in mild COVID-19 patients.

The immunopathology underlying severe COVID-19 has been thoroughly studied over the last two 90 years, including antibody (Ab) responses, cellular immune subsets, cytokines and chemokines that 91 were linked to characteristics and outcome of the disease 2-7 . However, with few exceptions 8 , 92 relatively little is known about the details of the immune response in mild patients (MP) and included in the patient subgroups (between 57% and 69%), while only around 30% of HC were male 144 (Table S1). No comorbidity information was available for HP and HC. For the other patient groups, 145 the prevalence of comorbidities (asthma, chronic hematologic disease, obesity and uncomplicated 146 diabetes) was higher among MP than ASP (5% to ~8% in MP vs. none in ASP) ( Table S1 ).

A whole blood count analysis was used to further characterize COVID-19 patients on D1. We found 149 no significant difference between ASP and MP in any of the tested 17 general blood count 150 parameters (Table S1, Figure S1 ). However, HP showed a remarkable difference compared to MP with other groups, HP showed a significant, but modest increase in the number or frequency of white 155 blood cells (WBC), monocytes, granulocytes and platelets, whereas a significant but modest 156 decrease in red blood count and hematocrit was observed (Figure S1d-k) . Considering the fact that 157 the time from the onset of first symptoms to diagnosis might take longer in severe cases, i.e., in HP, 158 we further analyzed only those HP with a shorter prodromal phase that was comparable to MP. To 159 this end, we performed a subcohort analysis by selecting HP (n=4) with at most seven-day delay 160 from the appearance of the first symptoms to inclusion to match the average time interval of all the 161 MP (for details, refer to STAR Methods). Encouragingly, both the CRP levels and the percentages 162 of lymphocytes were still significantly higher or lower, respectively, among HP vs. MP on D1 ( Figure   163 S1l, m).

As expected, on D1 we did not observe a significant increase in IgG Ab titers to SARS-CoV-2 S, 166 RBD, N and NTD in the non-hospitalized groups vs. HC (Fig. 1c-f ). In contrast, while only 48% and 167 43% of MP showed slightly increased IgG levels, 93% and 73% of HP already displayed significantly 168 enhanced IgG titers to CoV-2 S and N antigens respectively. Three weeks later, we observed a 169 significant increase in IgG levels to all the four antigens also in MP compared to HC together with a 170 J o u r n a l P r e -p r o o f Page 6 of 38 further enhancement of IgG levels in HP. The positivity rate for IgG Abs to RBD, NTD or 171 N reached up to 89% among MP and 100% among HP at D21. ASP had a lower positivity rate for 172 IgG against S/RBD (67%) and N (75%) than MP at D21 (Fig. 1c-e) . IgG Abs against NTD were in 173 general much lower in ASP and MP, except for HP at D21 (Fig. 1f ). Next, we tested the functional 174 capacity of the induced Abs in a surrogate virus neutralization assay. In line with the serology findings 175 on D1 (Fig. 1c, d) , HP already showed blocking Abs that interfered with ACE2 binding to or RBD at this early stage (Fig. 1g, h) . On D21, also MP had developed a significant inhibitory 177 capacity to block ACE2 binding to CoV-2 S or RBD vs. HC. Similar to the correlation reported by 178 others 19 , IgG Ab titers against both CoV-2 S and RBD, were highly correlated (Spearman r=0.89 and 179 0.82 for S and RBD, respectively) with the inhibitory capacity of sera from all the patients (Fig. 1i, j) . 180 

182 183 As shown above, routine laboratory data as well as deep serologic profiling of SARS-CoV-2-specific 184 Ab responses already distinguished HP from MP or ASP and HC. However, these analyses were 185 not sufficient to further differentiate non-hospitalized clinical phenotypes. A meta-analysis has found 186 that only 3% of the mean convalescent neutralizing Ab levels are necessary to predict protection 187 from severe COVID-19 20 , but that work has also indicated that those Abs are not sufficient to protect 188 from severe cases. 189 190 Thus, we aimed to explore the full complexity of innate and adaptive cellular immune signatures that 191 orchestrate the response to SARS-CoV-2 infection across the full spectrum of phenotypes. We systematically investigated 484 cellular immune subsets or combinations of various 193 lineage and functional markers (Fig. 1a) by three staining panels using 18-color flow cytometry (for 194 general gating strategy, see Figure S2 ; for cellular markers analyzed, refer to Key Resources Table   195 S2) in our longitudinal cohort. When compared with HC, ASP displayed no obvious change in all the We next used PCA to show that 484 immune features were only able to partition HP at D1 from all 201 other groups, but not MP at D1 from any HC (both D1 and D14) (Fig. 2a) . Then we asked whether 202 specific immune subsets were differentially present in MP vs. HC at D1 (Fig. 2b) . We observed 203 differences in the frequency of several CD8 T cells subsets, such as KI67 + , CD38 + , and 204 HLADR + CD38 + , representing proliferating, activated and antigen-specific responsive CD8 T cells, 205 respectively that were significantly enhanced in MP (Fig. 2c-e) . The profile included an increase of 206 both TBET-dependent (TBET + KI67 + ) and -independent (EOMES + KI67 + ) responsive CD8 T cells 207 ( Fig. 2f-h) . Also the fraction of proliferating CD4 T cells, especially, Th1-responsive (TBET + KI67 + ) 208 cells was already enhanced early on in MP on D1 (Fig. 2i, j, Figure S4a ). In parallel, the frequency 209 of antigen presenting cells (APCs) and Ab secreting cells, such as mature dendritic cells 210 (HLADR + CD38 high DCs) and short-lived plasmablasts (CD27 + CD38 high ), was increased in MP ( Fig.   211 2k, l, m, Figure S4b, 4c) . Notably, the frequency of activated CD38 + CD8 T cells, of HLADR + CD38 +

CD4 T cells and of mature DCs measured at D1 was highly predictive for the degree of the 213 serological titers of anti-SARS-CoV-2 N IgG and other CoV-2 antigens at D21 among ASP and MP 214 ( Fig. 2n-p) . Since the magnitude of the responses of CD4 T cells, CD8 T cells and mature DCs at 215 D1 was all highly correlated with the same parameter (i.e., Ab levels) at D21, the response levels of 216 these three subsets at D1 should correlate with each other. Such early responses of those immune 217 subsets were highly coordinated only in ASP and MP. On the contrary, it is noteworthy that neither 218 the frequency of activated subsets among CD4 and CD8 T cells, nor of mature DCs at D1 was 219 significantly correlated to the even higher titers of CoV-2 N IgG ( Fig. 2q-s) and other CoV-2 antigens 220 in HP at D21. This finding indicates that the progression and deterioration of COVID-19 is averted 221 only in the presence of a highly coordinated interplay of early innate and adaptive immune 222 responses, which are strongly correlated to the subsequent production of Ab titers.

On D21, MP were characterized by an enhanced cytotoxic CD8 T cell (GZMB + ) response, especially 225 of terminally differentiated responsive CD8 T cells (CD45RO -KI67 + ) (Figure S3d-f ). CD4 T cells also 226 showed similar changes. The frequencies of CD4 T cells expressing CD57 or GZMB as well as of 227 CD45RO and CD57 double-positive cells were also significantly enhanced in MP vs. HC ( Figure   228 S3g-i). Notably, the frequency of GZMB + CD4 cytotoxic T cells showed a trend to be elevated 229 (p=0.053, Kruskal-Wallis test including multiple-group correction) already on D1 ( Figure S3h) . The 230 CD57 expressing CD4 T cells detected on D21 appeared to be mainly cytotoxic effector cells since 231 the percentage of GZMB + CD57 + cells was also significantly enhanced among CD4 T cells ( Figure   232 S3j). Although CD57 is known as a T-cell senescence marker, CD57 + T cells, similar to the scenario 233 of PD-1 + T cells 21 , were apparently still functional during the acute phase of COVID-19, thus likely 234 contributing to a sufficient control of the infection in MP.

Early-stage impaired innate immunity in hospitalized, but not mild patients 236 237 Until now, we parsed primarily early immune signatures in MP in relation to HC on D1 and D21. Yet, 238 the analysis of early cellular responses and later Ab responses showed a positive correlation only in 239 MP and ASP, but not HP (Fig. 2n-s) . This finding prompted us to further compare between HP and 240 other groups. Thus, we asked whether any additional early immune signatures observed in MP were 241 significantly different from HP and determined the immune signatures that were significantly 242 upregulated or downregulated in MP vs. HP on D1 (Fig. 3a) . As shown in the volcano plot, major 243 differences were present primarily among innate immune cells, such as monocytes, DC and natural 244 killer (NK) cells, and to a lesser extent also among B and T cells. Compared with HP, MP showed a 245 much higher frequency (~40% in MP vs. ~10% in HP) of non-classical monocytes (ncMono, 246 HLADR + CD38 -) 22 . The ncMono were not only higher in frequency among MP, but expressed also 247 higher levels of critical functional markers, such as CD86/CD80 double-positivity (Fig 3b-d) , and CD13 ( Figure S4d, Figure S5a , b). Similar to monocytes, the frequency of APC such as 249 plasmacytoid DC (pDC) and myeloid DC (mDC) was significantly higher in MP vs HP (Fig. 3a, d-g) . 250 It is noteworthy that the frequency of ncMono and mDC (Fig.3b, f) was also slightly lower in MP vs.

J o u r n a l P r e -p r o o f Page 9 of 38 HC, indicating a disease severity-related effect and further supporting the involvement of both cell 252 types in early protective immune responses of COVID-19. Although mature DC were higher in both 253 MP and HP (Fig. 2l) , the frequency of CD86 -CD80 + cells (Fig. 3g, h) and of CD13 + cells ( Figure   254 S5c) among total DC was decreased in HP only, thus indicating a reduction in phagocytic and 255 antigen-presenting capacity of individual DC 23 . In line with the notion of reduced APC functions, the 256 downstream events of APC activation, the frequency of activated CD4 T cells (CD27 + ICOS + ) and 257 the ICOS MFI among CD8 T cells were decreased only in HP, but not MP (Figure S5d, e) . 258 Furthermore, the frequency of NK cells was also significantly decreased only in HP, but not in MP 259 vs. HC (Fig 3i, j) . In line with the overall compromised innate immune cell profile, critical senescence 260 and exhaustion markers such as KLRG1 and PD-1 were enhanced among several NK subsets in 261 HP only (Fig. 3k, Figure S4e , Figure S5f , g).

In contrast to the compromised innate immune compartment, the expression levels of CD86 and of 264 PD-L1 among class-switched memory B cells were even substantially enhanced in HP vs. both HC 265 and MP at D1 (Figure S5h , i). Considering these results together with the high SARS-CoV-2-specific 266 IgG titers, the ACE2 blocking capacity of patient serum and the high frequency of plasmablasts (even 267 higher than MP at both D1 and D21, Fig. 2m ), we concluded that Ab-secreting cells were not 268 impaired in both MP and HP at D1, thus leading to a robust Ab response after three weeks.

We next sought to understand whether the impaired innate immune response in HP was paralleled 271 by early deviated CD8 T cell profiles. Although the intensity (MFI) of ICOS on total CD8 T cells was 272 decreased (Figure S5e) , the frequency of ICOS + cells was unchanged ( Figure S3c ) and the 273 frequency of CD40L + and PD-1 + GZMB + cells among CD8 T cells was even significantly enhanced in 274 HP on D1 (Figure S5j, k) . Furthermore, since the frequency of CD8 T cells expressing other key 275 functional markers was not decreased in HP ( Fig. 2c-g) , the functional antiviral capacity of CD8 T Figure S5 ) in HP at D1, except for the frequency of one NK subset (Figure 3j ), 293 remained significantly higher (or lower for inhibitory markers) in MP even at D21 vs. HP at D1. In 294 summary, our data firmly support that it is the early-stage response of innate immunity (including NK 295 cells) that differentiates MP and HP during natural SARS-CoV-2 infection.

To gain further insight into the coordinated early immune response of COVID-19, we analyzed 24 299 different cytokines (refer to STAR Methods) at both D1 and D21 in sera of all groups. Interestingly, 300 at D1, we observed increased levels of interferon gamma-inducible protein 10 (IP10/CXCL10) in HP 301 and MP (Fig. 4a, b) . Unexpectedly, a similar regulation was found for the type I-interferon 302 which was previously reported to be undetectable in severe COVID-19 patients at around 10 days 303 after symptom onset 24 . Both MP and HP showed a significant increase of IFN-β compared to HC at 304 D1 (Fig. 4c) . While the levels of IP10 and IFN-β showed only a temporary increase among MP, 305 J o u r n a l P r e -p r o o f Page 11 of 38 declining to normal levels at D21, both IP10 and IFN-β levels remained elevated in HP after three 306 weeks (Fig. 4c) . These results point to a crucial and dynamic role of IP10 and IFN-β, which is tightly 307 regulated during the early stage of protective immune responses in COVID-19 patients. This notion 308 is also supported by the fact that levels of IP10 and IFN-β were significantly correlated with the 309 frequency of mature DCs among all the analyzed patients at D1 (Fig. 4d, e) . Interestingly, the 310 serological IP10 and IFN-β levels were significantly correlated with the magnitude of the CD8 T-cell 311 response among ASP and MP at D1 (Fig. 4f, g) . IFN-β levels were also correlated with the frequency 312 of HLA-DR+CD38+ cells among CD4 T cells (Fig. 4h) , indicating that the temporary IFN-β surge preventing from severe illness. The CD8 responses and IFN-β levels in HP even showed a trend to 316 be negatively correlated (R=-0.48, p-value=0.08, Figure S7b ). Furthermore, both IP10 and IFN-β 317 levels among mild patients significantly correlated with the SARS-CoV-2 viral load at D1 (Fig. 4i, j) , 318 as demonstrated by a negative correlation with the PCR Cq values. Regarding other circulating soluble factors, we found a significant and substantial increase in plasma 327 levels of eosinophil chemotactic protein (eotaxin-1/CCL11) and vascular endothelial growth factor A 328 (VEGFA) only in HP at D1 (Fig. 4l-m) . The enhanced levels of IL-6 and the regulatory cytokine IL-329 10 in HP vs. MP on D1 were still mostly seen within the normal range ( Figure S7h To identify the T-cell response on a broader scale, we performed TCRb sequencing analysis among 339 45 MP vs. eight ASP and 21 HC on D1 and D21. Aging has a strong impact on the TCR repertoire 25 .

As expected, sample clonality, the inverted normalized diversity index, was significantly correlated 341 with age ( Fig. 5a) . Since a decrease in TCR diversity was previously associated with aging and 342 impaired immunity against influenza virus infection and other diseases 26,27 , we sought to compare 343 the TCR diversity between different groups. The productive clonality of the sequenced TCR-β 344 repertoire was increased in MP at D21 vs. HC (Fig. 5b) . At D21, only the usage of one specific V 345 gene (TCRBV06-07) was significantly underrepresented in MP compared to HC (Fig. 5c) . Notably, 346 the SARS-CoV-2-specific T-cell clonotypes were substantially expanded (~six times higher than in 347 HC) among MP already at D1, as reflected by both clonal breadth and depth 28 , and maintained at 348 D21 ( Fig. 5d, e) . These results indicate a key functional role of early-responsive SARS-CoV-2-349 specific T cells in MP. Unexpectedly, inferred expanded CD4 SARS-CoV-2-specific T cells showed 350 an average frequency, ~ six times higher than that of inferred CD8 SARS-CoV-2-specific TCR 351 clonotypes among MP at D1 ( Fig. 5f-h) . This finding was in line with a trend for increased frequency 352 of GZMB + cells among CD4 Tconv cells, but not among CD8 T cells in MP vs. HC on D1 ( Figure   353 S3d, h). At D21, inferred CD4 SARS-CoV-2 specific TCR clonotypes continued to dominate over 354 CD8 clonotypes in MP (Fig. 5g, h) . The expansion of inferred CD4 or CD8 SARS-CoV-2-specific T 355 cells was highly correlated with the frequency of responsive ICOS + KI67 + cells among total CD4 or 356 CD8 T cells in both ASP and MP at D1 (Fig. 5i, j) . These data highlight a crucial role of early- To further consolidate our observations on inferred SARS-CoV-2-specific T cells based on 365 sequencing approaches 28 , we used another more direct experimental approach to analyze virus-366 specific T cells in a cytokine-independent but viral-peptide-specific way. To this end, the activation 367 induced marker (AIM) assay was selected to independently quantify SARS-CoV-2-specific T cell 368 responses. The AIM assay has been successfully used by others to quantify SARS-CoV-2-specific 377 378 SARS-CoV-2-specific CD8 responses were detected by calculating the portion of AIM (CD69 + 4-379 1BB/CD137 + ) 30,33 among total CD8 T cells. In line with our TCR sequencing data ( Fig. 5f-h) , the 380 overall re-stimulation response (SI) for CD8 T cells was much lower than for CD4 T cells in both 381 groups ( Fig. 6c-e) . Nevertheless, the SI for CD8 responses was still significantly higher in HP vs.

MP at D1 (Fig. 6e, f) . Contrary to CD4 T cells, we could not identify a significant difference between 383 MP and HC for CD8 T cell responses (Fig. 6e, f) . As shown above using the sequencing method, a 384 significant difference was observed between MP and HC ( Fig. 5h ) and the CD8 T-cell 385 activation/proliferation response was highly correlated with the frequency of sequencing-derived 386 SARS-CoV-2-specific T cells (Fig. 5j) . Notably, the frequency (~5E-5) of SARS-CoV-2-specific cells 387 J o u r n a l P r e -p r o o f Page 14 of 38 among total CD8 T cells detected using the robust tetramer approach 13 was in a similar range to that 388 inferred by the TCR-sequencing approach here. This indicates that the ex-vivo TCR sequencing-389 based approach might be even more sensitive than the in-vitro AIM assay, at least to detect antigen-390 specific CD8 T cells.

Excitingly, the SARS-CoV-2-specific CD4 T cells quantified by the AIM assays were highly correlated 393 with the clonal breadth determined by the TCR-β repertoire sequencing method (Fig. 6g, h) . This 394 correlation did not exist for the quantified SARS-CoV-2-specific CD8 T cells (Fig. 6i) . It is worthy to In May 2020, a mass PCR screening program was implemented on a population-wide level in 407 Luxembourg 37 , which allowed us to get unique access to PCR-positive ASP and MP and to 408 prospectively recruit them into our longitudinal study that was initiated simultaneously 17 . This 409 endorsed us to access a rich resource to fully explore and understand all essential facets of the 410 early-stage and dynamic immunological changes following recent SARS-CoV-2 infection in MP, 411 using an unbiased, combinatorial and prospective approach. HP. In addition, similar findings were also made in HP regarding the reduced frequency of several 499 NK subsets, paralleled by a substantial enhancement in the expression of the inhibitory and terminal 500 differentiation marker KLRG1 on NK cells. Thus, our data of impaired innate immune signatures in 501 HP confirm previous findings of impaired innate immunity in severe or critically ill patients 41,58,59 . 502 However, without including MP into the analyses, none of those previous studies has highlighted 503 that critical differences in early-stage innate immune responses actually exist between non-HP (i.e., 504 ASP and MP) and HP. Thus, our results provide strong support that an early-stage impairment of 505 innate immunity is a unique clinical immunological feature in HP.

Since our analysis was only performed in peripheral blood, our observations might be affected by a 508 potential redistribution of immune cell types between blood and inflammatory tissues. According to 509 the work deposited by Kedzierska and colleagues 60 , the respiratory tract, relative to blood, was 510 dominated by infiltrating neutrophils, a higher frequency of intermediate monocytes/macrophages 511 and effector T cells in severe patients. In our study, we observed a lower frequency of ncMono, pDC, 512 mDC and NK subsets and substantial changes in their critical functional markers in blood of HP vs. 513 MP on D1. Therefore, most of our results on the frequency of these immune cell subsets should not 514 be caused by a potential re-distribution between blood and infected tissues. Our observation on the 515 expression levels of functional markers is obviously independent from a potential re-distribution of 516 immune cells. Our findings about the frequency of ncMono in blood might have to be evaluated with 517 more caution, as monocyte infiltration into the airways has been identified as an important driver of 518 severe COVID-19 61 . In summary, most of the differences we observed between HP and MP are 519 likely not due to a potential re-distribution between blood and infected tissues. would be on a group of selected people who are PCR tested regularly for professional reasons (e.g. 551 sports athletes), which was not within the scope of our current study. Another limitation is the partial 552 comparability of time from symptom onset to inclusion between MP and HP.

In our cohort, we could not observe any distinguishable immune signature in the peripheral blood of 555 ASP throughout all adaptive and innate immunity analyses. Although the sample size was relatively 556 small in our ASP, the most plausible explanation is a more prominent role of a tissue-resident rather We worked to ensure gender balance in the recruitment of human subjects. We worked to ensure 588 ethnic or other types of diversity in the recruitment of human subjects. We worked to ensure that the 589 study questionnaires were prepared in an inclusive way. The author list of this article includes 590 contributors from the location where the research was conducted who participated in the data 591 collection, design, analysis, and/or interpretation of the work. Due to its highly-significant P value, IFN-β was also marked although its change fold was slightly All data reported in this work will be shared by the lead contact upon request.

This work does not report original code.

Any additional information required to re-analyze the data reported in this work is available from 771 the lead contact upon request. To quantify the viral load in swabs, RNA extracts were tested in duplicates with the rRT-PCR protocol 935 described above, together with a 3-fold dilution curve of the EDX SARS-CoV-2 standard (COV019, CoV-2 S inhibition % 

In a parallel, longitudinal and prospective cohort based on a general population, using a systems-immunology analysis approach, Capelle et al. provide a comprehensive resource of multilayered early immunological responses in mild COVID-19 patients. They reveal that a combination of highly coordinated early-stage immune responses is unique to mild patients.

A global database of COVID-19 vaccinations

Comprehensive mapping of immune perturbations 986 associated with severe COVID-19

Deep immune profiling of COVID-19 patients reveals distinct 989 immunotypes with therapeutic implications

Systems biological assessment of immunity to mild versus 992 severe COVID-19 infection in humans

Marked T cell activation, senescence, exhaustion and skewing towards 995 TH17 in patients with COVID-19 pneumonia

Functional exhaustion 997 of antiviral lymphocytes in COVID-19 patients

Low-Avidity CD4(+) T Cell Responses to SARS-CoV-2 in 1001 Unexposed Individuals and Humans with Severe COVID-19

Longitudinal analysis reveals that delayed bystander CD8+ T cell 1005 activation and early immune pathology distinguish severe COVID-19 from mild disease

A single-cell atlas of the peripheral immune response in 1009 patients with severe COVID-19

Single-cell RNA sequencing of blood antigen-presenting cells 1012 in severe COVID-19 reveals multi-process defects in antiviral immunity

COVID-1015 19 immune features revealed by a large-scale single-cell transcriptome atlas

Robust T Cell Immunity in Convalescent Individuals 1022 with Asymptomatic or Mild COVID-19

SARS-CoV-2-specific T cell immunity in cases of COVID-19 and SARS, and 1025 uninfected controls

A roadmap towards personalized 1027 immunology. npj Systems Biology and Applications

Systems immunology: just getting started

Protocol for a prospective, longitudinal cohort of people with COVID-19 and 1032 their household members to study factors associated with disease severity: the Predi-COVID study

Lymphopenia 1035 predicts disease severity of COVID-19: a descriptive and predictive study

Robust neutralizing antibodies to SARS-CoV-2 infection persist for 1039 months

Neutralizing antibody levels are highly predictive of immune 1042 protection from symptomatic SARS-CoV-2 infection

PD-1-Expressing SARS-CoV-2-Specific CD8(+) T Cells Are Not Exhausted, but Functional in 1046 Patients with COVID-19

Mononuclear phagocytes and marker modulation: when 1048 CD16 disappears, CD38 takes the stage

Aminopeptidase N (CD13) is 1050 involved in phagocytic processes in human dendritic cells and macrophages

Impaired type I interferon activity and inflammatory responses in severe 1054 COVID-19 patients

The influence of age on T cell generation and TCR diversity

Age-1059 associated decline in T cell repertoire diversity leads to holes in the repertoire and impaired immunity to 1060 influenza virus

T cell responses in patients with COVID-19

Diagnosis and Tracking of SARS-CoV-2 Infection By T-Cell Receptor 1065

Magnitude and Dynamics of the T-Cell Response to SARS-1068

CoV-2 Infection at Both Individual and Population Levels

Targets of T Cell Responses to SARS-CoV

Coronavirus in Humans with COVID-19 Disease and Unexposed Individuals

Antigen-Specific Adaptive Immunity to SARS-CoV-2 in Acute 1075 COVID-19 and Associations with Age and Disease Severity

SARS-CoV-2-reactive T cells in healthy donors and patients with COVID-19

SARS-CoV-2-specific T cell memory is sustained in COVID-19 convalescent 1082 patients for 10 months with successful development of stem cell-like memory T cells

Cytomegalovirus (CMV) seroprevalence in the adult 1086 population of Germany

Selective and cross-reactive SARS-CoV-2 T cell epitopes in unexposed 1089 humans

Suboptimal SARS-CoV-2-specific CD8(+) T cell response associated with 1092 the prominent HLA-A*02:01 phenotype

SARS-CoV-2 transmission risk from asymptomatic carriers: Results from a mass 1096 screening programme in Luxembourg. The Lancet Regional Health -Europe 4, 100056

Longitudinal high-throughput TCR repertoire profiling 1100 reveals the dynamics of T-cell memory formation after mild COVID-19 infection

Longitudinal profiling of respiratory and systemic immune 1104 responses reveals myeloid cell-driven lung inflammation in severe COVID-19

Longitudinal proteomic analysis of severe COVID-19 reveals 1108 survival-associated signatures, tissue-specific cell death, and cell-cell interactions

Time-resolved systems immunology reveals a late juncture linked to 1112 fatal COVID-19

Central Memory and Effector Memory T Cell Subsets: 1114 Function, Generation, and Maintenance

Production of type I interferons: plasmacytoid dendritic cells 1117 and beyond

Untuned antiviral immunity in COVID-19 revealed by 1120 temporal type I/III interferon patterns and flu comparison

Autoantibodies against type I IFNs in patients with life-threatening 1124 COVID-19

Dysregulation of type I interferon responses in COVID-19

Harnessing Type I IFN Immunity Against SARS-CoV-2 with 1129 Early Administration of IFN-beta

The chemokine receptor D6 limits the inflammatory response in vivo

CXCL10 Is an Agonist of the CC Family Chemokine Scavenger 1135 Receptor ACKR2/D6

CXCL10/IP-10 in 1137 infectious diseases pathogenesis and potential therapeutic implications

Plasma inflammatory cytokines and chemokines in severe acute respiratory 1141 syndrome

Plasma 1143 IP-10 and MCP-3 levels are highly associated with disease severity and predict the progression of COVID-19

A dynamic COVID-19 immune signature includes 1147 associations with poor prognosis

Elevated plasma levels of selective cytokines in COVID-19 patients reflect viral load and lung injury

Time-resolved systems immunology reveals a late juncture linked to 1153 fatal COVID-19

Immunology of COVID-19: Current State of the Science

Complexity of immune responses in 1158 COVID-19

Imbalanced Host Response to SARS-CoV-2 Drives Development of 1161 COVID-19

Innate immune deficiencies in patients with 1164 COVID-19. medRxiv

Immune responses in COVID-19 respiratory tract and blood reveal 1167 mechanisms of disease severity. medRxiv

Longitudinal profiling of respiratory and systemic immune 1170 responses reveals myeloid cell-driven lung inflammation in severe COVID-19

Rapid induction of antigen-specific CD4+ T cells is associated with 1174 coordinated humoral and cellular immune responses to SARS-CoV-2 mRNA vaccination

Pre-activated antiviral innate immunity in the upper 1178 airways controls early SARS-CoV-2 infection in children

Standard Peripheral Blood Mononuclear Cell Cryopreservation Selectively Decreases 1182 Detection of Nine Clinically Relevant T

Comprehensive assessment of T-cell receptor beta-chain diversity in alphabeta 1186 T cells

Multiplex Identification of Antigen-Specific T Cell Receptors Using a Combination of Immune Assays and 1189 Immune Receptor Sequencing