key: cord-0333735-cz8co138
authors: Garcia, M.; Carrasco Garcia, A.; Tauriainen, J.; Maleki, K.; Vaheri, A.; Makela, S.; Mustonen, J.; Smed-Sorensen, A.; Strandin, T.; Mjosberg, J.; Klingstrom, J.
title: Innate lymphoid cells are activated and their levels correlate with viral load in patients with Puumala hantavirus caused hemorrhagic fever with renal syndrome.
date: 2022-05-14
journal: nan
DOI: 10.1101/2022.05.10.22274837
sha: af6d07c0d6da880e4bf4379b147f0c5e5397cbdd
doc_id: 333735
cord_uid: cz8co138

Background. Innate lymphoid cells (ILCs) are involved in immunity and homeostasis but, except for natural killer (NK) cells, their role in human viral infections is not well known. Puumala virus (PUUV) is a hantavirus that causes the acute zoonotic disease hemorrhagic fever with renal syndrome (HFRS). HFRS is characterized by strong systemic inflammation and NK cells are highly activated in HFRS, suggesting that also other ILCs might be responding to infection. Methods. Here we phenotypically analyzed peripheral ILCs in acute and convalescent PUUV-infected HFRS patients. Additionally, plasma levels of soluble factors and viral load were analyzed. Findings. Overall, the frequencies of NK cells and naive ILCs were reduced while the frequency of ILC2, in particular the ILC2-lineage committed c-Kitlo ILC2 subset, was increased during acute HFRS. Interestingly, we observed a negative correlation between viral load and frequencies of both NK and non-NK ILCs in acute HFRS. Phenotypically, ILCs displayed an activated profile with increased proliferation, and showed altered expression of several homing markers during acute HFRS. In line with the observation of activated ILCs, plasma levels of inflammatory proteins, including the ILC-associated cytokines interleukin (IL)-13, IL-23, IL-25, IL-33, and thymic stromal lymphopoietin (TSLP), were elevated during acute HFRS. Interpretation. These findings indicate a general involvement of ILCs in response to human hantavirus infection. Further, this constitutes the first comprehensive study of ILCs in a hantavirus-caused disease, aiding in further understanding the role of these cells in disease pathogenesis and in human viral infections in general. Funding. A full list of funding bodies that contributed to this study can be found in the Acknowledgements section.

Our research shows that all circulating ILCs, including non-NK ILCs, are activated, 78 proliferating, and correlate with viral load in acute HFRS patients, indicating a general 79 involvement in human hantavirus infection and disease. Being the first comprehensive study 80 on ILCs in hantavirus infections, further research will give relevant insight into the roles of 81 ILCs in disease pathogenesis and protection. 82

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 14, 2022. ; https://doi.org/10.1101/2022.05.10.22274837 doi: medRxiv preprint INTRODUCTION 83 Innate lymphoid cells (ILCs) are a group of innate immune cells that play important roles in 84 the modulation of immune and inflammatory responses. 1 Naïve ILCs (nILC) constitute an 85 immature subset 2,3 that can home from peripheral blood to tissues where they give rise to the 86 mature ILC subsets 4 . Mature ILCs are classified in five main subsets based on the 87 transcription factors they express and the cytokines they produce: natural killer (NK) cells, 88 ILC1, ILC2, ILC3, and lymphoid tissue inducer (LTi) cells. 5,6 NK cells share the same 89 features as ILC1, but in addition can, as opposed to the other ILCs, kill virus-infected cells. 6 90 ILC1 and ILC3 are mainly found in mucosal tissues while NK cells and ILC2 are found both 91 in tissues and in peripheral blood. 2,4 Furthermore, two functionally distinct subsets of ILC2 92 can be found in peripheral blood: c-Kit lo and c-Kit hi ILC2, the first being more committed to 93 the ILC2 lineage. 7 94 NK cells have been extensively described in different viral infections. 8-10 On the other 95 hand, non-NK ILCs (hereinafter referred to as ILCs) were more recently discovered, 11,12 and 96 thus less is known regarding their role in viral infections. Due to their location in mucosal 97 tissue, ILCs are on the first line of defense and hence potentially essential in the early phases 98 of viral infections. 13 Their enrichment in lungs suggests an important role for ILCs in 99 respiratory viral infections. 13, 14 There are conflicting data regarding ILC2 and virus infections: 100 they have been suggested to promote tissue repair and protect the lungs of influenza-infected 101 mice, 15, 16 while other studies reported that ILC2 induce airway hyperreactivity in influenza-102 infected 17-19 and respiratory syncytial virus-infected 20 mice. Furthermore, ILCs have 103 recently been investigated in the context of a few human viral diseases. In HIV-1 infected 104 individuals, levels of ILCs were found to be reduced in ileum and colon, 21 as well as in 105 circulation, and to negatively correlate with viral load. 22 Total peripheral ILC levels were also 106 found to be decreased in SARS-CoV-2-infected coronavirus disease-19 (COVID-19) patients, 107 6 with ILC2 decreased in severe but not in moderate patients. [23] [24] [25] Moreover, in infants with 108 respiratory syncytial virus bronchiolitis, elevated levels of ILC2 in the airways were found to 109 associate with disease severity. 26 Here we performed a detailed characterization of peripheral blood ILCs and NK cells, 137 as well as of their cytokine and chemokine milieu, in PUUV-infected HFRS patients. We 138 observed increased plasma levels of inflammatory proteins, including ILC-associated 139 cytokines. We showed that NK cell frequencies are reduced during acute HFRS but recover 140 during convalescence. While total ILC frequencies did not change, we report increased 141 frequency of ILC2 and a concomitant decreased frequency of nILC during acute HFRS. In 142 particular, the ILC2-lineage committed c-Kit lo ILC2 subset was increased during acute HFRS. 143

Furthermore, NK cells and ILCs displayed an activated phenotype and ongoing proliferation 144 during acute HFRS. Interestingly, we observed a negative correlation between viral load and 145 severe HFRS based on a scoring system adapted from the sequential organ failure assessment 161 scoring system, where the maximum levels of creatinine (4 = > 440, 3 = 300-440, 2 = 171-162 299, 1 = 110-170, and 0 = < 110 µmol/l), minimum level of platelets (4 = < 20, 3 = 20-49, 2 = 163 50-99, 1 = 100-150, and 0 = > 150 x 10 3 /µl), and minimum mean arterial blood pressure (1 = 164 < 70 and 0 = ≥ 70 mmHg) were ranked. A total score of ≥ 5 was considered severe and < 5 165 mild. 61 Plasma levels of IL-5, IL-6, IL-7, IL-10, IL-13, IL-15, IL-17A, IL-18, IL-23, IL-25, IL-33,  203   IFN-γ, TNF, CCL20, CCL27, CCL28, GM-CSF, TSLP, 

The funders of this study had no role in the study design, data collection, data analysis, data 226 interpretation, or writing of the report. 227

A total of 17 PUUV-infected hospitalized HFRS patients and 10 healthy controls were 231 included in the study ( Table 1) . Peripheral blood samples were obtained at the acute (5-8 days 232

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 14, 2022.

after onset of symptoms), early convalescent (20-27 days after onset of symptoms), and late 233 convalescent (180 or 360 days after onset of symptoms) phase of disease (Fig 1a) . 234

Hospitalized HFRS patients showed a typical clinical presentation during the acute phase, 235 with thrombocytopenia, elevated C-reactive protein (CRP) and creatinine plasma levels, and 236 viral load (Table 1, Fig 1b, and Suppl Table 2 ). For most of the patients all parameters 237 normalized to levels within the normal range during the convalescent phase of disease (Fig  238   1b) . The severity of the patients was assessed with a scoring system based on platelet counts, 239 creatinine values, and mean arterial blood pressure values, as previously describe. 61,62 Two 240 patients scored as severe, while all other scored as mild (Suppl Table 2 Fig 1a) , we observed differences in levels of all proteins between acute 248 and later stages of HFRS (Fig 2a and b) . As previously reported, 37,40-43,64-67 the plasma levels 249 of tumor necrosis factor (TNF), IL-6, granulocyte-macrophage colony-stimulating factor 250 (GM-CSF), IL-10, interferon gamma (IFN-γ), IL-15, IL-18, and granzyme A (GrzA) were all 251 significantly higher in acute HFRS as compared to the convalescent phases (Fig 2b) . Further, 252 we observed significantly higher levels of the type 2-associated cytokines IL-13, IL-25 (also 253 called IL-17E), IL-33, and thymic stromal lymphopoietin (TSLP), related to ILC2 and T 254 helper 2 cell activity and involved in functions such as tissue repair. 68, 69 We also observed 255 significantly higher levels of the type 3-associated cytokine IL-23 in the acute phase of 256 HFRS, involved in the activation of immune cells such as ILC3 and T helper 17 cells which 257 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 14, 2022. ;

12 have a role in protection from tissue damage (Fig 2b) . 5,70 Moreover, we observed that levels 258 of the chemokines CCL20 and CCL27 were significantly higher in acute samples, while the 259 level of CCL28 was significantly decreased (Fig 2b) . 260

Principal component analysis (PCA) showed that samples from the acute phase 261 separated from samples from the early and late convalescent phase. This separation was 262 mainly driven by IL-10, GrzA, IFN-γ, TNF, IL-18, CCL27 and IL-33 (Fig 2c) . Interestingly, 263 patient 3, one of the two most severely ill patients in the cohort, deviated from the rest of the 264 patients, both in the acute and early convalescent phase (Fig 2c) . This patient showed the 265 highest levels of several soluble factors and presented higher levels of many of them in the 266 early convalescent phase than in the acute phase (Fig 2a) , suggesting a longer than usual 267 acute phase. 268

Additionally, we observed significant positive correlations between the type 2-269 associated cytokines IL-25 and IL-13 and between IL-25 and TSLP during the acute phase of 270 HFRS ( Suppl Fig 1b and Fig 2d and e) . Further, out of the 15 acute HFRS patients, 13 were 271 positive for PUUV S RNA in blood and a positive correlation was seen as well between viral 272 load and IFN-γ ( Suppl Fig 1b and Fig 2f) . 273

Altogether, these results showed that PUUV-infected HFRS patients display a strong 274 inflammatory response, including elevated levels of 16 cytokines many of which are known to 275 be produced by or involved in the activation of ILCs and NK cells. 276

We next characterized the ILC and NK cell compartments in PBMCs from HFRS patients. 279

For the identification and analysis of ILCs and NK cells, we used 18-parameter flow 280 cytometry and a modification of a well-established gating strategy (Suppl Fig 2) . 71 281 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 14, 2022. ;

We observed decreased frequencies of total CD56 + NK cells in peripheral blood in the 282 acute and early convalescent phase of disease, which normalized in late convalescence (Fig  283   3a) . There was a decreased frequency of CD56 dim NK cells, with a concomitant increase of 284 the smaller population of CD56 bright NK cells, during the acute phase of HFRS (Fig 3b and c) . 285

High frequencies of CD69 + and HLA-DR + total CD56 + NK cells were detected in the acute 286 phase of HFRS, indicating NK cell activation (Fig 3d) . As expected from the general NK cell 287 activation, frequencies of NKp44 + and NKG2A + NK cells were also increased in the acute 288 phase of HFRS (Fig 3d) . Furthermore, the frequency of Ki-67 + NK cells was increased, 289

showing that the NK cells proliferated during acute HFRS (Fig 3d) . Additionally, when 290 analyzing for homing receptors, we observed a decreased frequency of α4β7 + NK cells and an 291 increased frequency of CCR6 + and CCR10 + NK cells in the acute phase of HFRS (Fig 3d) , 292

suggesting an effect on NK cell migration. The frequency of CD45RA + and CD161 + NK cells 293 was decreased in HFRS patients as compared to healthy controls, with a tendency to recovery 294 in the convalescent phase of HFRS (Fig 3d) . Analysis of surface markers in the CD56 bright and 295 CD56 dim NK cell subsets showed similar results as observed for total NK cells (Suppl Fig 3) . 296

A PCA revealed a separation of acute HFRS from convalescent HFRS patients and healthy 297 controls based on the level of expression of the different analyzed surface markers in total NK 298 cells (Fig 3e) . 299

Next, we examined correlations between frequencies of NK cells and both soluble 300 plasma proteins and clinical parameters (Suppl Fig 4a-d) . IL-10 plasma levels positively 301 correlated with the frequencies of both activated (CD69 + ) and proliferating (Ki-67 + ) NK cells 302 (Fig 3f and g) . Granzyme A levels correlated positively with the frequency of CD69 + NK 303 cells (Fig 3h) , and negatively with the frequency of CD161 + NK cells (Fig 3i) . Interestingly, 304 the frequency of activated (CD69 + ) NK cells correlated positively with viral load while the 305 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 14, 2022. ;

14 frequency of total NK cells showed a negative correlation with viral load during the acute 306 phase of HFRS (Fig 3j and k) . 307 308 Peripheral ILCs are activated and proliferate during acute HFRS 309

We next characterized the peripheral ILC responses in the HFRS patients. No significant 310 difference in total ILC frequency was observed between the patients and the healthy controls 311 (Fig 4a) . Interestingly, as for NK cells, we observed a negative correlation between viral load 312 and the frequency of ILCs in acute HFRS, showing reduced frequencies of peripheral ILCs in 313 patients with higher viral loads (Fig 4b) . Furthermore, we observed increased frequencies of 314 activated (CD69 + ) and proliferating (Ki-67 + ) ILCs during the acute phase of HFRS (Suppl 315 Fig 5a) , while no differences were observed in the frequencies of NKp44, HLA-DR, and 316 CD45RA-expressing ILCs (Suppl Fig 5a) . When assessing expression of homing markers, 317

we found a decreased frequency of α4β7 + ILCs in acute HFRS, but no significant difference 318 in frequencies of ILCs expressing the chemokine receptors CCR6 and CCR10 (Suppl Fig 5a) . 319

Next, we explored specific ILC subsets. CD117 neg ILCs in peripheral blood have been 320

shown to make up a heterogenous population with yet undefined functions. 2 We therefore 321 decided to focus our analysis on the more well-defined nILC and ILC2. The composition of 322 these ILC subsets changed over time in the HFRS patients (Fig 4c) . We observed an increase 323 in ILC2 frequency and a decreased frequency of naïve ILC (nILC) in the acute phase of 324 HFRS (Fig 4c and d) . 325 326 Peripheral c-Kit lo ILC2 are increased in frequency during HFRS 327

Next, we characterized the phenotype of the ILC subsets. Increased frequencies of activated 328 (CD69 + ) and proliferating (Ki-67 + ) nILC were observed in acute HFRS (Fig 4e) . Moreover, 329 similar to NK cells (Fig 3d) , a decreased frequency of α4β7 + nILC was observed in acute 330 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 14, 2022. ; https://doi.org/10.1101/2022.05.10.22274837 doi: medRxiv preprint 15 HFRS (Fig 4e) . No differences were observed in the frequencies of nILCs expressing HLA-331 DR, NKp44, CCR6, and CCR10 in HFRS patients as compared to healthy controls (Fig 4e) . 332 PCA based on the frequency of expression of surface markers in nILC showed a separation 333 between the acute HFRS patients and healthy controls (Fig 4f) . 334

The ILC2 population showed a similar phenotypic pattern as the nILC, with increased 335 frequency of activated (CD69 + ) and proliferating (Ki-67 + ) cells during acute HFRS (Fig 4g) . 336

Additionally, a significantly decreased frequency of CCR6 + ILC2 was observed in the acute 337 phase of disease (Fig 4g) , while no differences were observed in the frequencies of ILC2 338 expressing NKp44, HLA-DR, CCR10, and 47 in HFRS as compared to healthy controls 339 (Fig 4g) . In line with these findings, a PCA based on the frequency of expression of surface 340 markers on ILC2s showed no clear separation between the acute and convalescent HFRS, but 341 a separation of the acute and control samples was observed (Fig 4h) . Further, when analyzing 342 for possible correlations to soluble proteins (Suppl Fig 6) , we observed a positive correlation 343 between the plasma levels of IL-10 and the frequency of CD69 + ILC2 (Fig 4i) and between 344 plasma levels of the CCR10 ligand CCL27 and CCR10 + ILC2 (Fig 4j) , as well as a negative 345 correlation between plasma levels of TSLP and frequency of CCR10 + ILC2 in acute HFRS 346 patients (Fig 4k) . The first two correlations were also observed for total ILCs, as well as a 347 positive correlation between plasma levels of IL-7 and Ki-67 + ILC (Suppl Fig 5b) . 348

Having observed an increased frequency of ILC2 (Fig 4d) , a decreased frequency of 349 CCR6 + ILC2 (Fig 4g) , and increased plasma levels of type 2 cytokines in acute HFRS (Fig 2) , 350 we next assessed whether there were changes in the ILC2 subsets in HFRS patients. Indeed, 351

we observed increased frequencies of c-Kit lo ILC2 and, concomitantly, decreased frequencies 352 of c-Kit hi ILC2 during the acute phase of HFRS, as compared to convalescent HFRS patients 353 and healthy controls (Fig 5a and b) . Moreover, c-Kit hi ILC2 showed higher frequency of 354 CCR6 expression than c-Kit lo ILC2 both in patients and healthy controls (Fig 5c) and, 355 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 14, 2022.

aligning with the relative depletion of CCR6 + ILC2 in acute HFRS, (Fig 4g) we observed a 356 lower frequency of CCR6 + c-Kit hi ILC2 in the acute phase of HFRS as compared to the 357 convalescent phase (Fig 5c) . We observed a decreased frequency of CD161 + NK cells in acute HFRS patients. The 407 role of CD161 in NK cells is not fully understood, but it has been suggested to mark pro-408 inflammatory NK cells with a high ability to respond to innate cytokines. 96 In conclusion, this study provides the first comprehensive characterization of total 449 circulating ILCs in hantavirus-infected patients. We report an overall activated and 450 proliferating ILC profile in these patients, with a particular increased frequency of the ILC2 451 subset, and a skewing towards the ILC2-lineage committed c-Kit lo ILC2 in acute HRFS. 452

Additionally, we show that NK cells are reduced in frequencies and confirm that remaining 453 circulating NK cells are highly activated and proliferating in acute HFRS. Moreover, we 454 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 14, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 14, 2022.

Bar graphs are shown as mean and lines connect paired samples from the same 908 patient. Statistical significance was assessed using the Wilcoxon signed-rank test to compare 909 groups of HFRS patients, and the Kruskal-Wallis test followed by Dunn's multiple 910 comparisons test to compare healthy controls with groups of HFRS patients. Severe patients 911 are indicated by a black circle. Patients with low cell numbers (fewer than 20 events) in the 912 corresponding gate were removed from the analysis. *p < 0.05; **p < 0.01; ***p < 0.001. 913

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 14, 2022. ; https://doi.org/10.1101/2022.05.10.22274837 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. IL-5  IFN-IL-6  IL-7  IL-10  IL-13  IL-15  IL-17A  IL-18  IL-23  IL-25  IL-33  GM-CSF  Granzyme A  TNF  TSLP  CCL20  CCL27  CCL28   Acute Early Convalescence Late Convalescence a

Severe . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Ly . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 14, 2022. ; https://doi.org/10.1101/2022.05.10.22274837 doi: medRxiv preprint

Cytokine response to Hantaan virus 706 infection in patients with hemorrhagic fever with renal syndrome

TH2 cell development and function

Group 2 Innate Lymphoid Cells in Airway Diseases

A 713 cytokine network involving IL-36γ, IL-23, and IL-22 promotes antimicrobial defense 714 and recovery from intestinal barrier damage

Spatial and Temporal Mapping of Human Innate Lymphoid Cells Reveals Elements of 718

Hantavirus-infection Confers Resistance to Cytotoxic Lymphocyte-Mediated 721

Cell Activation in Human Hantavirus Infection Explained by Virus-Induced IL-725

Pulmonary innate lymphoid cell responses during rhinovirus-induced asthma

A clinical trial

Upregulation of IFN-γ and IL-12 is associated with a milder form 733 of hantavirus hemorrhagic fever with renal syndrome

Spectrum of hantavirus infection: hemorrhagic fever 736 with renal syndrome and hantavirus pulmonary syndrome

The 739 pathogenesis of nephropathia epidemica: New knowledge and unanswered questions

Hantavirus infection: a global 742 zoonotic challenge

Avšič-Županc T. Viral load and 744 immune response dynamics in patients with haemorrhagic fever with renal syndrome

Viral load 747 and humoral immune response in association with disease severity in Puumala 748 hantavirus-infected patients-implications for treatment

IL-25-and IL-33-responsive type 2 innate lymphoid cells are defined by expression of 752 CRTH2 and CD161

Group 2 innate lymphoid cells in pulmonary immunity 754 32 and tissue homeostasis

Role of ILC2 in Viral-Induced 756

Group 2 Innate Lymphoid Cells

Team Players in Regulating Asthma

Severity of Haemorrhagic Fever with Renal Syndrome and Regulates the Inflammatory 762

Response in Hantaan Virus-Infected Endothelial Cells

Leukocyte Trafficking to the Small 765 Intestine and Colon

Transcriptional 768 programs of lymphoid tissue capillary and high endothelium reveal control 769 mechanisms for lymphocyte homing

CCR6 as a mediator of 771 immunity in the lung and gut

Distinct roles of L-selectin and integrins α4β7

LFA-1 in lymphocyte homing to Peyer's patch-HEV in situ: The multistep model 775 confirmed and refined

Hantavirus Inhibits TRAIL-Mediated Killing of Infected Cells by Downregulating 778

Death Receptor 5

Hantavirus inhibits apoptosis by 780 preventing mitochondrial membrane potential loss through up-regulation of the pro-781 survival factor BCL-2

Viruses and interferon: A fight for supremacy

Antiviral actions of interferons

Direct antiviral mechanisms of interferon-gamma

Interferon-gamma (IFN-γ): Exploring its implications in 790 infectious diseases

CD161 defines a functionally distinct subset of pro-inflammatory natural killer cells

Reduced 795 frequencies of NKp30+NKp46+, CD161+, and NKG2D+ NK cells in acute HCV 796 infection may predict viral clearance

Unconventional repertoire profile is imprinted during acute chikungunya infection for 800 natural killer cells polarization toward cytotoxicity

Activation, exhaustion, and persistent decline of the antimicrobial MR1-restricted 804

Abbreviations: IFN-γ: interferon gamma; IL: interleukin; TNF: tumor necrosis factor 835 alpha; TSLP: Thymic stromal lymphopoietin; GM-CSF: granulocyte-macrophage colony-836 stimulating factor; CCL: chemokine ligand. Graphs show data of individual subjects (circles) 837 and the median (bars) ± interquartile range. Statistical significance was assessed using the 838Wilcoxon signed-rank test. Severe patients are indicated by a black circle and patient 3 is 839 labeled as P03. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. 840 Kit lo and CCR6 + c-Kit hi ILC2 in control donors (n=10) and HFRS patients during the acute 906 (n=15), early convalescence (n=16), and late convalescence (n=17) phase. 907