Innate lymphoid cells are activated in HFRS, and their function can be modulated by hantavirus-induced type I interferons

Hantaviruses cause the acute zoonotic diseases hemorrhagic fever with renal syndrome (HFRS) and hantavirus pulmonary syndrome (HPS). Infected patients show strong systemic inflammation and immune cell activation. NK cells are highly activated in HFRS, suggesting that also other innate lymphoid cells (ILCs) might be responding to infection. Here, we characterized peripheral ILC responses, and measured plasma levels of soluble factors and plasma viral load, in 17 Puumala virus (PUUV)-infected HFRS patients. This revealed an increased frequency of ILC2 in patients, in particular the ILC2 lineage-committed c-Kitlo ILC2 subset. Patients’ ILCs showed an activated profile with increased proliferation and displayed altered expression of several homing markers. How ILCs are activated during viral infection is largely unknown. When analyzing PUUV-mediated activation of ILCs in vitro we observed that this was dependent on type I interferons, suggesting a role for type I interferons—produced in response to virus infection–in the activation of ILCs. Further, stimulation of naïve ILC2s with IFN-β affected ILC2 cytokine responses in vitro, causing decreased IL-5 and IL-13, and increased IL-10, CXCL10, and GM-CSF secretion. These results show that ILCs are activated in HFRS patients and suggest that the classical antiviral type I IFNs are involved in shaping ILC functions.


Introduction
Innate lymphoid cells (ILCs) play important roles in the modulation of immune and inflammatory responses [1].Naïve ILCs (nILC) constitute an immature ILC subset [2,3] that can, at least in mice, home from peripheral blood to tissue, where they give rise to the mature ILC subsets [4].Mature ILCs are classified in five main subsets: natural killer (NK) cells, ILC1s, ILC2s, ILC3s, and lymphoid tissue inducer (LTi) cells [5,6].NK cells share the same features as ILC1s, but can in addition kill virus-infected cells [6].ILC1s and ILC3s are mainly found in mucosal tissues while NK cells and ILC2s are found both in tissues and in peripheral blood [2,4].Furthermore, two functionally distinct subsets of ILC2s can be found in peripheral blood: the more ILC2 lineage-committed c-Kit lo ILC2 and the plastic c-Kit hi ILC2s [7].
While NK cells have been extensively described in different viral infections [8][9][10], less is known regarding non-NK ILCs (hereinafter referred to as ILCs) in this context.In HIV-1 infected individuals, levels of ILCs were found to be reduced in ileum and colon [11] as well as in circulation, and to negatively correlate with viral load [12].Total peripheral ILC levels were also found to be decreased in SARS-CoV-2-infected coronavirus disease-19 (COVID- 19) patients, with ILC2s levels being decreased in severe but not in moderate disease [13][14][15].Moreover, in infants with respiratory syncytial virus-caused bronchiolitis, elevated levels of ILC2s in the airways were found to associate with disease severity [16].Overall, these studies suggest that viral infection can have an effect on the ILC landscape, but the mechanisms behind virus-induced activation of ILCs is largely unknown.
Rodent-borne hantaviruses can cause hemorrhagic fever with renal syndrome (HFRS) in Eurasia and hantavirus pulmonary syndrome (HPS) in the Americas, with case fatality rates of 0.4-10% depending on the specific HFRS-causing hantavirus, and 35%, respectively [17][18][19][20].Humans are normally infected via inhalation of hantavirus-infected rodent excreta, and the virus then spreads systemically, primarily infecting endothelial cells [17,21].Puumala virus (PUUV) is endemic in Europe, where it is the most common causative agent of HFRS [18,22].HFRS patients initially develop non-specific symptoms such as high fever and headache, often progressing to kidney dysfunction and gastrointestinal symptoms, including abdominal pain, diarrhea, vomiting, and gastrointestinal bleeding.Lung dysfunction is common in HPS but can also occur in severe HFRS [19,23].

PLOS PATHOGENS
Innate lymphoid cells in hemorrhagic fever with renal syndrome Hantaviruses are believed to trigger immunopathogenic responses that contribute to hyperinflammation [24][25][26][27][28].However, the exact mechanisms leading to these events remain to be understood [22,29,30].Several different types of immune cells have been shown to be altered in patients.Peripheral NK cells show signs of strong activation and proliferation in the acute phase of HFRS [31][32][33].B and T cells are highly expanded in the circulation of HFRS and HPS patients, with concomitant elevated levels of CD8 + T cells in the respiratory airways [24,[34][35][36][37]. Mucosa-associated invariant T (MAIT) cells were recently shown to be reduced, but highly activated, in circulation during PUUV-caused HFRS [27].Levels of neutrophils are highly increased and also strongly activated in hantavirus-infected patients [38][39][40].Mononuclear phagocytes are susceptible to hantavirus infection, and their activation and redistribution from circulation towards the airways and kidneys in HFRS patients have been reported [41][42][43].Combined, these reports show that hantaviruses trigger strong immune cell responses, indicating a possible effect also on ILCs.
Type I IFNs comprise a wide array of interferons, including IFN-β and several subtypes of IFN-α.Type I IFNs have several antiviral effects; both direct, by inducing an antiviral state in exposed cells, and indirect, by further activating antiviral responses in immune cells [44,45].
Here we performed a detailed characterization of peripheral blood ILCs and NK cells, as well as of their cytokine and chemokine milieu, in PUUV-infected HFRS patients.We observed increased plasma levels of inflammatory proteins, including ILC-associated cytokines.We showed that NK cell frequencies are reduced during acute HFRS but recover during convalescence.While total ILC frequencies did not change, we report an increased frequency of ILC2s, in particular of the ILC2 lineage-committed c-Kit lo ILC2 subset, and a concomitant decreased frequency of nILCs during acute HFRS.Furthermore, NK cells and ILCs displayed an activated and proliferating phenotype during acute HFRS.We observed a negative correlation between viral load and the frequencies of both NK cells and ILCs in acute HFRS, suggesting a potential direct or indirect influence of hantaviruses on the ILC landscape.Finally, in vitro studies showed that PUUV-mediated activation of ILC2s is dependent on type I IFNs, and that IFN-β per se impacts ILC2 functions by inducing specific cytokine responses.

Ethics statement
The study was approved by the Ethics Committee of Tampere University Hospital (ethical permit nr.R04180) and all subjects provided written informed consent.Control peripheral blood mononuclear cells (PBMCs) were obtained from the Blood Transfusion Clinic at the Karolinska University Hospital Huddinge, Stockholm, Sweden (ethical permit nr.2020-02604).

Patient samples
17 patients, 12 females and 5 males, with serologically confirmed acute HFRS were included in the study.The patients were diagnosed at the Tampere University Hospital in Finland during 2002-2007.Whole blood samples were collected, and PBMCs were isolated as previously described and stored in liquid nitrogen at -150˚C, while plasma was stored at -80˚C, until further use [43].
Control PBMCs were obtained from buffy coats of 10 blood donors.PBMCs were isolated from the buffy coats by density centrifugation using Lymphoprep (StemCell Technologies), according to the manufacturer's guidelines and stored at -150˚C until further use.

Viral load in plasma
RNA was isolated from 140 μl EDTA plasma using a column-based RNA isolation kit, according to the manufacturer's instructions (Viral RNA mini kit, Qiagen).Isolated RNA was subjected to PUUV S RNA RT-qPCR analysis based on a previously described protocol [47], with TaqMan fast virus 1-step master mix (ThermoFisher Scientific) and using AriaMx instrumentation (Agilent).

Flow cytometry
PBMC samples were thawed in RPMI (Cytiva) complete media [L-glutamine (ThermoFisher Scientific), 10% FCS (Sigma-Aldrich), penicillin/streptomycin (P/S) (Cytiva)] with 10 U/mL DNase (Roche) and counted.3-4 million cells per sample were stained.Briefly, cells were incubated with LIVE/DEAD Fixable Green Dead Cell Stain Kit (ThermoFischer Scientific) and fluorochrome-conjugated antibodies directed against surface markers (S1 Table ) for 20 min at room temperature in the dark followed by 2 washes with flow cytometry buffer (PBS with 2 mM EDTA).Cells were then fixed and permeabilized using FACS Lysing solution and FACS Permeabilizing solution (BD Biosciences), and next incubated with intracellular staining antibodies (S1 Table ) for 30 min at 4˚C in the dark.Cells were then washed and resuspended in flow cytometry buffer.Samples were acquired with a BD LSR Fortessa (BD Biosciences) flow cytometer.Flow cytometric analysis was performed using FlowJo version 10.7.2 (TreeStar, Ashland).To ensure unbiased manual gating, a blinded analysis was implemented, whereby all FCS3.0 files were renamed and coded by one person and blindly analyzed by another person.All samples were compensated electronically, and gatings were based on fluorescent-minusone (FMO) or negative controls.After all gatings were performed, samples were decoded and data analysis was performed.

ELISA
IFN-α levels in plasma and supernatants were analysed using human IFN-α pan ELISA development kit (Mabtech) according to the manufacturer's guidelines.Plasma was diluted 1:1 in ready-to-use ELISA diluent (Mabtech) prior to the ELISA.IFN-β levels in supernatants were analysed using IFN-β DuoSet ELISA development kit (R&D Systems) according to the manufacturer's guidelines.

ILC2 in vitro expansion
Sorted human ILC2s were seeded at 1,000 cells per well in U-bottom 96-well plates in expansion media [collection media supplemented with 50 ng/mL of IL-1β (BioTechne) and 100 U/ mL of IL-2 (Peprotech)] and incubated at 37˚C.At day 5-6 half of the media was renewed with fresh expansion media.At day 10-11 cells were counted and split at a 1:3-1:4 ratio.Finally, at day 13-14 cells were counted and seeded at a concentration of 250,000 cells per well in new 96-well plates in resting medium [collection media supplemented with 2 U/mL of IL-2 (Peprotech) and 5 ng/mL of IL-7 (BioTechne)] to bring them to a resting phase.Cells were incubated for 24 h at 37˚C and were then either directly used for subsequent experiments or stored at -150˚C until further use.

Cells and viruses
Primary human umbilical vein endothelial cells (HUVECs) (pooled donors, sex: mixed) (Lonza) were maintained in EGM-2 medium supplemented according to the manufacturer's guidelines (Lonza), with the exception that cortisone was only included in the medium until cells were split to plates used for infection experiments.Cells were cultured at 37˚C in 5% CO 2 .PUUV strain CG-1820 was propagated as previously described [27].

In vitro co-culture of ILC2s and HUVECs
HUVECs (0.2x10 6 cells per well in a 24-well plate) were infected with PUUV at multiplicity of infection (MOI) 1 with gentle shaking every 15 min, or left unstimulated in infection medium [HBSS (Gibco), 1% FCS (Sigma-Aldrich), 2% HEPES 1M (HyClone), 1% P/S (Cytiva)].After 1 h of incubation at 37˚C, the infection medium was removed and fresh EGM-2 medium added to the cells.After 48 h, media was renewed and 24 h later (day 3 after infection) expanded and rested ILC2s (0.1x10 6 cells) were added to uninfected and PUUV-infected HUVECs.In parallel, cell culture supernatants from uninfected and PUUV-infected HUVECs were collected, centrifuged to remove cell debris, and added to ILC2s (0.1x10 6 cells per well in a 24-well plate).
After 24 h of culture, ILC2s from the different conditions were washed, stained (panel described in S3 Table ), fixed, and acquired as described above in the flow cytometry section.
In a separate experiment, freshly sorted and expanded ILC2s were stimulated with 10-fold dilutions (0.1, 1, 10, 100, and 1000 ng/mL) of recombinant human IFN-β (Peprotech).Supernatants were collected at day 1, 2, and 3 post-stimulations.Cytokine levels were analysed by multiplex immunoassay as described above.Media alone was used as negative control.

In vitro hantavirus-infection of ILC2s
Sorted and expanded fresh human ILC2s were exposed to PUUV at MOI 0.3, 1, and 3, for 1h.The cells were then washed and fresh ILC2 resting medium was added to the cells.Medium was renewed at 5h post infection (hpi), 1 day post infection (dpi), and 3 dpi.Cells were sampled at 5 hpi and at 5 dpi, lyzed with Trizol and stored at -20˚C.RNA was then extracted using direct-zol RNA MiniPrep Kit (Zymo Research) according to the manufacturer´s instructions, with RNA eluted in 30 μl nuclease-free water.The extracted RNA was subsequently reverse transcribed using high-capacity Reverse Transcriptase kit (Thermo Fischer Scientfic) according to the manufacturer´s instructions and stored at -20˚C until further analysis.

Real-time PCR
Real-time PCR was performed on a CFX96 Real-Time PCR System (Biorad) using iTaq Universal Probes Supermix (Biorad) with the following cycling conditions: initial denaturation at 95˚C for 3 min, followed by 45 cycles of 95˚C for 5 s and 60˚C for 30 s.

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

Study design and patient characteristics
A total of 17 hospitalized PUUV-infected HFRS patients and 10 uninfected controls were included in the study (Table 1).
Peripheral blood samples were obtained from 15 patients at the acute (5-8 days after onset of symptoms), 16 patients at the early convalescent (20-27 days after onset of symptoms), and 17 patients at the late convalescent (180 or 360 days after onset of symptoms) phase of disease

PLOS PATHOGENS
Innate lymphoid cells in hemorrhagic fever with renal syndrome (Fig 1A).Samples were not available from two patients during the acute phase and from one patient during the early convalescence phase.Hospitalized HFRS patients showed a typical clinical presentation during the acute phase, with thrombocytopenia, elevated C-reactive protein (CRP) and plasma creatinine levels, and viremia (Fig 1B, and Tables 1 and S4).For most of the patients all parameters normalized to levels within the normal range during the convalescent phase of disease (Fig 1B).The severity of the patients was assessed with a scoring system based on platelet counts, creatinine values, and mean arterial blood pressure values, as previously described [43,46].Two patients scored as severe, while all other scored as non-severe (S4 Table ).

HFRS patients present a strong inflammatory response during the acute phase of disease
Hantavirus-caused disease is characterized by strong systemic inflammatory responses [25][26][27][28]49].Using a multiplex immunoassay, we assessed the plasma levels of 19 cytokines in HFRS patients during the acute and convalescent phase of disease (Figs 2A and 2B, and S2A).As previously reported [25][26][27][28]49], the plasma levels of tumor necrosis factor (TNF), interleukin (IL)-6, granulocyte-macrophage colony-stimulating factor (GM-CSF), IL-10, interferon gamma (IFN-γ), IL-15, IL-18, and granzyme A (GrzA) were all significantly higher in acute HFRS as compared to the convalescent phases (Fig 2B).Further, we observed significantly higher levels of the type 2-associated cytokines IL-13, IL-25 (also called IL-17E), IL-33, and thymic stromal lymphopoietin (TSLP), related to ILC2 and T helper 2 cell activity and involved in functions such as tissue repair [50,51].We also observed significantly higher levels of the type 3-associated cytokine IL-23 in the acute phase of HFRS (Fig 2B).IL-23 is involved in the activation of immune cells such as ILC3s and T helper 17 cells, which have a role in protection from tissue damage [5,52].Moreover, we observed that levels of the chemokines CCL20 and CCL27 were significantly higher in acute samples, while the level of CCL28 was significantly decreased (Fig 2B).No differences in levels of IL-5, IL-7, and IL-17A were observed during the acute compared to later stages of HFRS (S2A Fig).
Principal component analysis (PCA) showed that samples from the acute phase separated from samples from the early and late convalescent phase.This separation was mainly driven by IL-10, GrzA, IFN-γ, TNF, IL-18, CCL27, and IL-33 (Fig 2C).Interestingly, patient 12, one of the two most severely ill patients in the cohort, deviated from the rest of the patients, both in the acute and early convalescent phase (Fig 2C).This patient showed the highest levels of several soluble factors and presented higher levels of many of them in the early convalescent phase than in the acute phase (Fig 2A ), suggesting a longer than usual acute phase.We observed significant positive correlations between the type 2-associated cytokines IL-25 and IL-13 and between IL-25 and TSLP during the acute phase of HFRS (Figs 2D and 2E, and  S2B).
Out of the 15 acute HFRS patients, 13 were positive for PUUV S RNA in blood.During the acute phase, a strong positive correlation was observed between viral load and plasma levels of IFN-γ (Figs 2F and S2B).
Altogether, these results showed that PUUV-infected HFRS patients display a strong inflammatory response, including elevated levels of 16 cytokines, many of which are known to be produced by or involved in the activation of ILCs and NK cells.

Peripheral NK cells are activated but decreased in frequency during acute HFRS
Next, we characterized the ILC and NK cell compartments in PBMCs from HFRS patients.For the identification and analysis of ILCs and NK cells, we used 18-parameter flow cytometry and a modification of a well-established gating strategy (S1 Fig) [53].
We observed decreased frequencies of total CD56 + NK cells in peripheral blood in the acute and early convalescent phase of disease, which normalized in late convalescence (Fig 3A and  3B).There was a decreased frequency of CD56 dim NK cells, with a concomitant increase of the smaller population of CD56 bright NK cells, during the acute phase of HFRS (Fig 3C and 3D).High frequencies of CD69 + and HLA-DR + total CD56 + NK cells were detected in the acute phase of HFRS, indicating NK cell activation (Fig 3E).As expected from the general NK cell activation, frequencies of NKp44 + and NKG2A + NK cells were also increased in the acute phase of HFRS (Fig 3E).Furthermore, the frequency of Ki-67 + NK cells was increased, suggesting that the NK cells proliferated during acute HFRS (Fig 3E).Additionally, when analyzing for homing receptors, we observed a decreased frequency of α4β7 + NK cells and an increased

Peripheral non-NK cell ILCs are activated and proliferate during acute HFRS
We next characterized the peripheral non-NK cell ILC responses in the HFRS patients.No significant difference in total ILC frequency was observed between the patients and the controls (Fig 4A and 4B).Interestingly, as for NK cells, we observed a strong negative correlation between viral load and the frequency of ILCs in acute HFRS, showing reduced frequencies of peripheral ILCs in patients with higher viral loads (Fig 4I).Furthermore, we observed increased frequencies of activated (CD69 + ) and proliferating (Ki-67 + ) ILCs during the acute phase of HFRS (S5A-S5B Fig) .No differences were observed in the frequencies of HLA-DR, and CD45RA-expressing ILCs (S5C-S5D Fig) .When assessing expression of homing markers, we found a decreased frequency of α4β7 + ILCs in acute HFRS, but no significant difference in frequencies of ILCs expressing the chemokine receptors CCR6 or CCR10 (S5E-S5G Fig) .Next, we explored specific ILC subsets.CD117 neg ILCs in peripheral blood constitute a heterogenous population with yet undefined functions [2].We therefore decided to focus our analysis on the more well-defined nILC and ILC2 subsets.The composition of these ILC subsets changed over time in the HFRS patients (Fig 4C and 4D).We observed an increase in ILC2 frequency and a decreased frequency of nILC in the acute phase of HFRS (Fig 4C and  4D).

Peripheral c-Kit lo ILC2s are increased in frequency during HFRS
Next, we characterized more in detail the phenotype of these ILC subsets.Increased frequencies of activated (CD69 + ) and proliferating (Ki-67 + ) nILCs were observed in acute HFRS (Fig early convalescence (n = 16), and late convalescence (n = 17) phase.(e) Percentage of CD69 + , Ki-67 + , HLA-DR + , NKp44 + , NKG2A + , CCR6 + , CCR10 + , α4β7 + , CD45RA + , and CD161 + NK cells of total CD3 -CD56 + cells in control donors (n = 10) and HFRS patients during the acute (n = 15), early convalescence (n = 16), and late convalescence (n = 17) phase.(f) Principal component analysis of total CD3 -CD56 + NK cells in controls and HFRS patients displaying the contribution of NK cell surface markers indicated in (e).Each dot represents one donor.(g-i) Spearman rank correlation between (g) plasma IL-10 levels and the percentage of CD69 + NK cells, (h) plasma granzyme A (GrzA) levels and the percentage of CD161 + NK cells, and (i) plasma viral load (n = 13; PUUV S RNA copies/mL) and the percentage of CD69 + NK cells in acute HFRS patients.Bar graphs are shown as mean and lines connect paired samples from the same patient (circles).Statistical significance was assessed using the Wilcoxon signed-rank test to compare groups of HFRS patients, and the Kruskal-Wallis test followed by Dunn's multiple comparisons test to compare controls with groups of HFRS patients.

ILC2s are activated by soluble factors produced by PUUV-infected endothelial cells
To investigate how ILCs are activated in the context of hantavirus infection, we performed an in vitro co-culture assay.Given that ILCs are scarce in peripheral blood, we enriched ILCs from buffy coats, sorted ILC2s and expanded them ex vivo.ILC2s co-cultured with PUUVinfected HUVECs showed an increased frequency of CD69 expression, indicating activation (Fig 6A and 6B).To test if activation was cell contact-dependent, we next exposed ILCs to conditioned medium from uninfected and PUUV-infected HUVECs.This revealed a cell contactindependent activation of ILC2s, mediated by conditioned medium from PUUV-infected HUVECs (Fig 6A and 6B).ILC2 exposed to supernatants from PUUV-infected HUVEC also showed increased viability (Fig 6C and 6D).Taken together, these results suggest that soluble factor(s) secreted by hantavirus-infected HUVECs can activate ILC2s.

PUUV-mediated ILC2 activation is dependent on type I IFNs
To identify potential factors involved in the PUUV-mediated activation of ILC2s, we pretreated supernatants from infected and uninfected HUVECs with cytokine-blocking compounds before supernatants were added to the ILC2s.Blocking type I IFNs significantly decreased the activation and viability of ILC2s exposed to supernatants from PUUV-infected HUVECs (Fig 6E -6G).In contrast, blocking IL-6, IL-12, and IL-18 did not affect the activation of ILC2s (S7 Fig) .Endothelial cells normally respond to viral infections by producing IFN-β.[54,55] To verify that PUUV induced an IFN-β response in infected HUVECs, supernatants were collected daily for three days and subsequently analysed for IFN-β levels.IFN-β was detected in supernatants from PUUV-infected cells at days two and three post infection (Fig 6H ), showing that the conditioned supernatants contained type I IFNs.
To verify that type I IFNs per se could activate ILC2s, we next exposed ILC2s to increasing concentrations of IFN-β for 24 h and then analyzed their activation status.This revealed a dosedependent effect on the activation of ILC2s (Fig 6I ) and a dose-independent increase in viability (Fig 6J).The highest concentration of IFN-β tested caused similar levels of ILC2 activation as the canonical ILC2 activators (a combination of the alarmins TSLP, IL-25, and IL-33) (Fig 6I).We next assessed IFN-β-mediated effects on ILC2 cytokine responses.To this end, supernatants from ILC2s exposed to different concentrations of IFN-β were collected daily for three days and then cytokine levels were analysed.While IL-4, IL-6, and IL-15 were not detected in supernatants from untreated or IFN-β-treated cells, supernatants from IFN-β-treated ILC2s contained elevated levels of CXCL10, IL-10, and GM-CSF, and lower levels of IL-5 and IL-13 as compared to controls ( Fig 7).
We observed a small effect on activation (Fig 6B ) but not on viability (Fig 6D ) of PUUVexposed ILC2s, indicating a possible direct effect of PUUV on activation.Endothelial cells are the main target for hantaviruses, but also other types of cells, including monocytes, can be infected [42].To test if also ILC2s can be infected by hantavirus we exposed ILC2s to PUUV for 5 days and then analysed levels of PUUV RNA in the cells.Five days after infection, a 10-fold increase in PUUV RNA was detected in ILC2s from one of three donors, suggesting ILC2s may be a target for hantavirus infection (S8 Fig) .Finally, we assessed the levels of type I IFNs in HFRS patients.HFRS patients in the acute phase presented with higher levels of IFN-α in plasma as compared to late convalescent patients (Fig 6K ), confirming that the level of type I IFNs is increased in HFRS.
supernatants, or exposed to PUUV or media alone (n = 8, 4 independent experiments).(e-g) Type I IFN blocking reagent B18R was added to supernatants of HUVEC cultures (PUUV-infected and uninfected) for 5 h prior to addition to human expanded ILC2s (n = 6, same donors as in (a-d), 3 independent experiments) and incubation for 24 h.(e, g) Percentage of (e) activation (CD69 + ) and of (g) live human expanded ILC2s.(f) Percentage of inhibition of the increase in activation (CD69 + ) of ILC2s caused by PUUV-infected HUVEC, when pre-blocking supernatants with B18R.(h) Levels of IFN-β in supernatants from PUUV-infected and uninfected HUVECs over time (n = 3).ND: not detected.(i, j) Percentage of (i) activation (CD69 + ) and of (j) live human expanded ILC2s cultured for 24 h with increasing concentrations of recombinant human IFN-β (ng/mL), the canonical ILC2 activators (alarmins TSLP, IL-25, and IL-33) plus IL-2, or only media (n = 3, 2 independent experiments).(k) Levels of IFN-α in plasma of patients with HFRS during the acute (n = 15), early convalescence (n = 16), and late (n = 17) convalescent phase of disease.Black dots represent the severe HFRS patients.Bar graphs are shown as mean and lines connect paired samples from the same patient (circles).Statistical significance was assessed using the Wilcoxon signed-rank test to compare groups of HFRS patients (k), and paired t-test to compare conditions from in vitro assays.*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.https://doi.org/10.1371/journal.ppat.1012390.g006 HFRS is characterized by a strong general immune activation, including hyperinflammation [18,60].Viral load normally peaks between the first 3 to 5 days of disease, and hantaviruses are normally not detected in the convalescent phase [19,22,61,62].In line with this, the HFRS patients in our study presented with typical laboratory findings together with a strong inflammatory response in the acute phase of disease [25][26][27][28]49].From an ILC-perspective, we observed increased levels of cytokines such as IL-13, IL-25, IL-33, and TSLP in acute PUUVinfected HFRS patients.The alarmins IL-25, IL-33, and TSLP are known activators of ILC2s, which, upon activation, can secrete IL-13 [63,64].These alarmins are upregulated upon infection or damage of epithelial, endothelial, and stromal cells [65][66][67].Recently, IL-33 was reported to be elevated in plasma of Hantaan virus-infected HFRS patients and to positively correlate with disease severity, indicating involvement in pathogenesis [68].There are conflicting data regarding IL-5 in HFRS.While we did not observe an increase in IL-5, Shakirova and coworkers have reported elevated IL-5 levels in HFRS patients [69].
As recently shown by others [32], we observed a transient reduction in NK cell frequencies in peripheral blood of PUUV-infected HFRS patients.Moreover, as earlier shown [31], we observed that remaining circulatory NK cells were highly proliferating, with approximately half of them expressing Ki-67.During the acute phase of HFRS, NK cells also showed increased expression of several activation markers and altered expression of chemokine receptors such as CCR6 and α4β7, which are associated with migration to tissues (lung and intestines, respectively) [70][71][72].Hence, NK cell migration may explain the observed decreased frequencies of NK cells in peripheral blood during the acute phase of HFRS.Interestingly, in peripheral blood viral load correlated negatively with the frequency of NK cells but positively with the frequency of activated NK cells.This suggests that viral replication indirectly impacts the activation of NK cells during HFRS.In line with this, we have previously shown that hantaviruses have strong anti-apoptotic properties that potentially protect infected cells from cytotoxic lymphocyte-mediated killing and may lead to NK cell-mediated bystander killing of uninfected cells [57,58,73,74].We observed a decreased frequency of CD161 + NK cells in acute HFRS, mirroring our previous observation of downregulated CD161 expression on MAIT cells in HFRS [27].
Levels of IFN-γ positively correlated with viral load in HFRS patients.IFN-γ is a cytokine with strong antiviral effects produced by a wide array of innate and adaptive lymphocytes, including ILCs, NK cells, NKT cells, and T cells [44,[75][76][77].The main target of hantavirus infection, endothelial cells, and the potential monocytes , do not express IFN-γ.It can therefore be speculated that active viral replication in infected endothelial cells indirectly, for example via enhanced IL-15 production [58], activates the IFN-γ production by NK cells and other immune cells.
The frequency of ILCs has been shown to be reduced in the circulation of individuals with acute HIV infection, and to correlate negatively with the viral load [12].We recently reported a decrease in peripheral ILC frequencies and numbers in COVID-19 [13].In contrast, here, we did not observe a significant change in the frequency of ILCs in HFRS.Interestingly though, and in line with that observed in HIV-infected individuals [12], frequencies of peripheral ILCs in acute HFRS showed a strong negative correlation with viral load.Furthermore, as described for moderately ill COVID-19 patients [13], acute HFRS patients presented increased frequencies of ILC2s with a concomitant reduction of nILC frequencies.In contrast to our findings in COVID-19 patients [13], we observed increased frequencies of Ki-67-expressing ILC2s and nILCs in the acute phase of HFRS, showing increased proliferation.Moreover, the decreased levels of α4β7 + nILCs in acute HFRS suggest that the reduced frequency of peripheral blood nILCs could be due to their migration to tissues [4].Alternatively, nILCs might differentiate into mature ILC subsets, such as ILC2s, in circulation.However, it remains unclear if nILCs are recruited to tissues and if mature ILC subsets can be replenished by circulating nILCs in humans [4,78,79].Acute HFRS patients presented increased frequencies of c-Kit lo ILC2s.Two ILC2 subsets have been defined, differing in their surface expression of cKit and their functionality.c-Kit lo ILC2s are more mature and ILC2-lineage-committed, while c-Kit hi ILC2s show plasticity towards an ILC3 phenotype and functionality [7].Moreover, c-Kit lo ILC2s express lower levels of CCR6 compared to c-Kit hi ILC2s [7].In line with this, acute HFRS patients showed decreased frequency of CCR6 + c-Kit hi ILC2s as compared to convalescent patients and controls.This suggests a skewing of c-Kit hi ILC2s towards more ILC2-commited cells in HFRS, possibly explaining the increase in c-Kit lo ILC2 frequencies in acute HFRS.Alternatively, this decrease in CCR6 + c-Kit hi ILC2 levels could also be due to migration of these cells to lung, where ILC2s have been shown to play an important role in tissue repair during influenza infection in mice [64,80].
Several immune cell types have been described to be activated in HFRS and HPS patients [24,27,31,32,34,40,42].Nonetheless, the mechanisms of their activation remain largely elusive.We recently showed that hantavirus-mediated activation of MAIT cells is type I IFN-dependent [27].Here, we show that hantavirus can also activate ILCs via induction of type I IFNs produced by infected cells.Besides from type I IFNs, levels of a wide range of cytokines are elevated during HFRS, which likely contribute further to activation of ILCs and MAIT cells in patients.We observed that ILC2s may be infected with PUUV, suggesting ILCs are potential targets for hantavirus.Interestingly, Dengue virus, another zoonotic RNA virus, has been shown to infect ILC2s, with a majority of all circulating ILC2s being infected in some patients [81].It remains to be investigated if ILCs are infected also in HFRS patients and what potential effects this could have on the ILC phenotype and functions.
Type I IFNs are crucial for a successful immune defense against viruses.Besides from inducing an antiviral state that interferes with viral replication, type I IFNs also stimulate immune cells, like B, T, and NK cells, to take part in the antiviral defense [45,82].The effects of type I IFNs on immune cells are complex and context dependent; both the quantity and the timing of the IFN production impact their response, as well as other specific immune responses simultaneously taking place, for instance inflammatory responses [45].We observed that IFN-β had a clear effect on ILC2 cytokine responses by decreasing IL-5 and IL-13 and increasing IL-10, GM-CSF, and CXCL10 secretion.IFN-β has previously been shown to decrease human and mouse ILC2 production of IL-5 and IL-13 [83,84].Further, it has reported that type I IFNs inhibit ILC2 proliferation in mice [83], while others-in line with our observation of increased human ILC2 survival-reported decreased apoptosis in IFN-β stimulated mouse ILC2s [84].Of particular interest in the context of viral infections was the observation of increased IL-10 secretion by ILC2s upon IFN-β stimulation.Human ILC2s have been previously shown to acquire a regulatory phenotype when induced with retinoic acid, inhibiting the secretion of IL-5 and IL-13 while stimulating the secretion of IL-10 [85].Further, allergic patients treated with allergen immunotherapy showed increased levels of IL-10 + ILC2 which, in vitro, attenuated type 2 T helper responses and maintained epithelial cell integrity [86].Whether hantavirus infection could trigger a similar protective IFN-induced IL-10 + ILC2 phenotype in patients remains to be further investigated.In summary, the overall effect of type I IFNs on ILC2s will depend on the combined effect of the cytokine milieu induced by the infection.It remains to be studied if different viruses, which induce type I IFNs with different magnitudes and kinetics and different additional inflammatory cytokine responses, have different potential to activate or suppress ILC functions in patients.
Here we characterized peripheral ILCs in PUUV-infected HFRS patients.Future studies of ILCs in tissue samples, such as lung and intestines, can add important knowledge regarding possible ILC tissue infiltration and local ILC responses.
In conclusion, this study provides the first comprehensive characterization of total circulating ILCs in hantavirus-infected patients.We report an overall activated and proliferating ILC profile in these patients, with a particular increased frequency of the ILC2 subset, and a skewing towards the ILC2 lineage-committed c-Kit lo ILC2 in acute HRFS.Further, we show that, in vitro, ILC2 activation and functionality can be mediated by type I IFNs secreted by PUUVinfected endothelial cells, and that ILC2 might be a potential target for hantavirus infection.Additionally, we observe that NK cells are reduced in frequencies and confirm that remaining circulating NK cells are highly activated and proliferating in acute HFRS.Moreover, we report a negative correlation between viral load and the frequencies of both NK cells and ILCs in acute HFRS, suggesting a potential influence of viral replication on these cells during the acute phase of hantavirus-caused disease.

Fig 5 .Fig 6 .
Fig 5. Increased ILC2 c-Kit lo frequency in peripheral blood of acute HFRS patients.(a) Representative flow cytometry plots showing percentages of c-Kit lo and c-Kit hi ILC2s in a control donor and in an HFRS patient (gated on total ILC2s as in S1 Fig).(b) Percentage of c-Kit lo and c-Kit hi ILC2s in control donors (n = 10) and HFRS patients during the acute (n = 15), early convalescence (n = 16), and late convalescence (n = 17) phase.(c) Percentage of CCR6 + c-Kit lo and CCR6 + c-Kit hi ILC2s in control donors (n = 10) and HFRS patients during the acute (n = 15), early convalescence (n = 16), and late convalescence (n = 17) phase.Bar graphs are shown as mean and lines connect paired samples from the same patient (circles).Statistical significance was assessed using the Wilcoxon signed-rank test to compare groups of HFRS patients, and the Kruskal-Wallis test followed by Dunn's multiple comparisons test to compare controls with groups of HFRS patients.Severe patients are indicated by a black circle.Patients with low cell numbers (fewer than 20 events) in the corresponding gate were removed from the analysis.*p < 0.05; **p < 0.01; ***p < 0.001.https://doi.org/10.1371/journal.ppat.1012390.g005

Fig 7 .
Fig 7. IFN-β influences ILC2 cytokine responses in vitro.Levels of IL-5, IL-10, IL-13, CXCL10, and GM-CSF in supernatants from ILC2s exposed to IFN-β in vitro.Human ILC2s were expanded in vitro and exposed to IFN-β (0.1, 1, 10, 100, and 1000 ng/mL) for 3 days.Untreated cells were used as a negative control.Bar graphs are shown as mean with standard deviation (n = 3).https://doi.org/10.1371/journal.ppat.1012390.g007 the correlation coefficients.(PDF) S3 Fig. NK cell subsets are activated, proliferating, and present a migratory profile in peripheral blood of HFRS patients.(a-b) Percentage of CD69+, Ki-67+, HLA-DR+, NKp44 +, NKG2A+, CCR6+, CCR10+, α4β7+, CD45RA+, and CD161+ (a) CD56dim NK cells and (b) CD56bright NK cells in control donors (n = 10) and HFRS patients during the acute (n = 15), early convalescence (n = 16), and convalescence (n = 17) phase.(c-e) Spearman rank correlation between (c) plasma IL-10 levels and the percentage of Ki-67+ NK cells, (d) plasma granzyme A (GrzA) levels and the percentage of CD69+ NK cells, and (e) plasma viral load (n = 13; PUUV S RNA copies/mL) and the percentage of NK cells out of CD45+ lymphocytes in acute HFRS patients.Bar graphs are shown as mean and lines connect paired samples from the same patient.Statistical significance was assessed using the Wilcoxon signed-rank test to compare groups of HFRS patients, and the Kruskal-Wallis test followed by Dunn's multiple comparisons test to compare controls with groups of HFRS patients.Severe patients are indicated by a black circle.ρ: Spearman´s rank correlation coefficient.*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.(PDF) S4 Fig. Correlations of soluble factors and clinical parameters with NK cells in HFRS patients.Spearman correlation matrix of the level of soluble factors in plasma and the percentage of (a) NK cells, (b) CD56dim NK cells, and (c) CD56bright NK cells in acute HFRS patients.Spearman correlation matrix of the clinical parameters and the percentage of (d) NK cells, (e) CD56dim NK cells, and (f) CD56bright NK cells in acute HFRS patients.The colour of the circles indicates positive (red) and negative (blue) correlations that were statistically significant (p < 0.05) as measured by the Spearman's rank correlation coefficient test.The colour intensity and the size of the circle are proportional to the correlation coefficients.Ly: lymphocytes.Days a. symp.: days after symptoms onset.(PDF) S5 Fig. ILCs display an activated and proliferating profile in peripheral blood of HFRS patients.(a-g) Representative flow cytometry plots and graphs showing the percentage of CD69+, Ki-67+, HLA-DR+, CD45RA+, CCR6+, CCR10+, and α4β7+ ILCs in control donors (n = 10) and HFRS patients during the acute (n = 15), early convalescence (n = 16), and late convalescence (n = 17) phase.(h-j) Spearman rank correlation between (h) plasma IL-10 levels and the percentage of CD69+ ILCs, (i) plasma CCL27 levels and the percentage of CCR10 + ILCs, and (j) plasma IL-7 levels and the percentage of Ki-67+ ILCs in acute HFRS patients (n = 15).(k-m) Spearman rank correlation between (k) plasma IL-10 levels and the percentage of CD69+ ILC2, (l) plasma CCL27 levels and the percentage of CCR10+ ILC2s, and (m) plasma TSLP levels and the percentage of CCR10+ ILC2s in acute HFRS patients (n = 15).Bar graphs are shown as mean and lines connect paired samples from the same patient.Statistical significance was assessed using the Wilcoxon signed-rank test to compare groups of HFRS patients, and the Kruskal-Wallis test followed by Dunn's multiple comparisons test to compare controls with groups of HFRS patients.Severe patients are indicated by a black circle.ρ: Spear-man´s rank correlation coefficient.*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.(PDF) S6 Fig. Correlations of soluble factors and clinical parameters with ILCs in HFRS patients.Spearman correlation matrix of the level of soluble factors in plasma and the percentage of (a) ILCs, (b) ILC2s, and (c) nILCs in acute HFRS patients.Spearman correlation matrix of the clinical parameters and the percentage of (d) ILCs, (e) ILC2s, and (f) nILCs in acute HFRS Table for a full list of antibodies used.