Longitudinal blood cell transcriptomic proling and in vitro temporal proteomics provides novel insights into metabolic reprogramming and host-immune responses against Crimean-Congo Hemorrhagic Fever Virus

The pathogenesis and host-viral interactions of the Crimean–Congo hemorrhagic fever orthonairovirus (CCHFV) are convoluted and is not well evaluated. To understand the host immune responses against CCHFV, we have performed a global transcriptomic analysis of peripheral blood mononuclear cells from a longitudinal cohort of CCHF patients who survived the infection and temporal untargeted proteomics analysis of CCHFV infected Huh7 cells. Our results indicate that during the acute phase of CCHFV infection, the host's metabolic reprogramming towards central carbon metabolism including glycolysis and glutaminolysis occurs that favours the virus replication as blocking these pathways in vitro inhibits CCHFV cellular replication. Furthermore, CCHFV replication was inhibited by blocking Akt with MK-2206 suggesting a regulatory role of PI3K/Akt/mTOR pathways. We also show activation of key interferon stimulating genes during infection, suggesting a role for type I and II interferon-mediated antiviral mechanisms. Targeting immune-metabolic pathways could be attractive therapeutic intervention for CCHFV. of remaining samples using the R/Bioconductor package DESeq2 v1.26.0 45 . Genes with adjusted p-values of less than 0.05 were considered as signicantly regulated. Functional analysis of the signicantly regulated genes was performed using the enrichr module of the python package GSEAPY v 0.9.16 46, 47 . Gene ontology (GO) enrichment analysis was performed using the Enrichr web resource 47 , and the GO biological process 2018 category was selected. Redundant GO terms were removed using the online tool REVIGO 15 . Heatmaps were generated using the R/Bioconductor package ComplexHeatmap v2.2.0 48 . Bubble plots, MA plots, Volcanoplots, violin plots and bar plots were created using R package ggplot2 v3.3.2. Network visualization was performed using Cytoscape v3.6.1 49 . Venn diagrams were constructed using the online tool InteractiVenn 50 . All the codes are available in GitHub (https://neogilab.github.io/CCHF-Turkey/). Raw RNAseq data is available in Sequence Read Archive (SRA) with temporary id SUB8608640. . cells were with CCHFV (Ibar 10200 strain) at MOI of 0.1. After 1h, inoculum was replenished with fresh Leitbovitz medium containing 5% FBS and incubated 48h. After infection cells were xed 30 min using ice-cold acetone. Cells were permeabilized using 0.1% Triton x-100 for 10 min at RT prior probe incubation at 40°C of IFI27 and CCHFV probes (RNAscope® ISH ACD,US) for 2h, 30min, 15 min, 30 min and 15 min, respectively. Cells were further stained using DAPI and coverslip attached using ProLong Gold Antifade reagent (P10144, Thermosher). Images were acquired using Nikon Single Point scanning confocal with 60/1.4 oil objective. CGGAAAGCCTATAAAACCTACCTTC-3’, Reverse-2: 5’-CGGAAAGCCTATAAAACCTGCCYTC-3’ and Reverse-3: 5’-CGGAAAGCCTAAAAAATCTGCCTTC-3’ and Probe FAM-CTGACAAGYTCAGCAAC –MGB. was using a


Introduction
Crimean-Congo hemorrhagic fever orthonairovirus (CCHFV) is a negative-sense RNA virus belonging to the Nairoviridae virus family, which causes a mild to severe viral hemorrhagic fever and is a major emerging viral threat causing an increasing number of outbreaks in several parts of the globe. Crimean-Congo hemorrhagic fever (CCHF) poses a substantial threat to public health due to its high mortality rate in humans (3-40%), modes of transmission (tick-to-human/animal, animal-to-human and human-tohuman) and geographical distribution 1 . CCHF is endemic in almost 30 countries in sub-Saharan Africa, South-Eastern Europe, the Middle East, Central Asia 2, 3 . The Ixodid ticks, especially those of the genus, Hyalomma, are both a vector and a reservoir for CCHFV and are highly ubiquitous with its presence in more than 40 countries 4 . In recent years, CCHFV outbreaks have become more frequent and have spread to new geographical areas. This has been attributed to climate changes and the spread of infected ticks by birds and the livestock trade. The presence of the CCHFV vector in Portugal, Spain, Germany, and even Sweden 5 and England 6 highlights the need for stricter surveillance due to the possibility of a future epidemic 7 .
Viruses exploit the host metabolic machinery to meet their biosynthetic demands. This reliance is further highlighted by observed variations in the cell-speci c viral replications and production leading to changes in host metabolism 8 . In particular, virus infection leads to signi cant metabolic alterations in the host, in some cases resulting in changes in glycolysis rates and ATP production rates 9 . The changes in the energy metabolism can therefore be seen as an evolving property of the combined host-virus metabolic system and could be related to changes in host cellular demands arising from viral production 10 .
One of the primary antiviral defense mechanisms is type-I interferon (IFN-I) response. IFN-I are pleiotropic cytokines with varied cellular functions mediated by the transcriptional activations of several interferon-stimulated genes (ISGs). It is known that CCHFV replication is sensitive to IFN-I 11 . However, the virus is also able to delay the induction of IFN, and IFN treatment has been shown to be ineffective following establishment of infection, suggesting that CCHFV has developed mechanisms to block innate immune responses 12 . The existing knowledge from the in vitro studies, autopsies of CCHF patients and clinical ndings suggests a regulatory mechanism of the IFN-response during CCHFV infection and indicates that ISGs may offer protection by inhibiting CCHFV. Furthermore, It has been suggested that many of the ISGs can rewire cellular metabolism to support an antiviral response. 13 However, not much is known about the induction of ISGs during CCHFV infection and their relationship with metabolic pathways.
Because of the sporadic nature of CCHF outbreaks in humans in the endemic regions, a lack of infrastructure, and the absence of structured studies, little is known about the pathogenesis and hostvirus interactions during the acute phase of CCHF disease and associated sequelae after recovery. Understanding the host response to CCHFV is necessary to design better therapeutic and containment strategies for CCHF. Recently, systems biology studies using -omics approaches on patient material and infected cells have offered a route to elucidate potential mechanisms of the host immune response against invading viruses and disease pathogenesis, as recently reported by us in case of SARS-CoV-2 infection 14 .
In this study, we have applied global blood transcriptomics in longitudinal blood samples collected during both the acute phase of CCHFV infection and the convalescent-phase (nearly after a year of recovery) and temporal quantitative proteomics analysis and modulation of the key pathways by drugs in in vitro infection model to understand the host immune and cellular responses against CCHFV. Our study provides a comprehensive understanding of the regulation of cellular pathways during CCHFV infection.

Results
Samples and clinical data. In this study, 18 samples were collected during the acute phase of the disease with a median time of 4 days (range 1-6 days) after the onset of symptoms. By using severity grading scores (SGS), 33% (6/18) patients were grouped into severity group 1 (SG-1), 61% (11/18) patients into severity group 2 (SG-2) and, 6% (1/18) patients into severity group 3 (SG-3). The median age of the patients was 49 years (range: 18-79), and 12 (66.7%) of the patients were male. A 79-year-old male patient in SG-3 died on the third day of hospitalization. The case-fatality-rate (CFR) for the cohort was 5.6%. Follow-up samples were collected from 12 individuals after a median duration of 54 weeks (range: 46-57 weeks). The CCHF patient characteristics are summarized individually in Supplementary Table S1.
The impact of CCHF severity on peripheral blood transcriptome and cytokine pro le during acute infection. We rst considered the effect of the SGS disease severity group on the gene expression pro le at the mRNA level during the acute phase of infection. The UMAP using all of the genes (Fig. S1A) and hierarchical clustering analysis of the top 500 high variable genes (Fig. S1B) did not show distinct differences between samples of different severity groups. To speci cally investigate the genes that were signi cantly associated with disease severity, the samples were grouped into either SG-1 or severity groups 2 and 3. The differential gene expression (DGE) was plotted as an MA-plot, as a measure of the number of genes whose expression was consistently altered between these two groups (Fig. S1C). There were 12 genes (ERG, PROM1, HP, HBD, AHSP, CTSG, PPARG, TIMP4, SMIM10, RNASE1, VSIG4, CMBL, MT1G) that were signi cantly upregulated in patients in SG-2 and SG-3 compared to SG-1 (Fig. S1C). However, no obvious links between these genes were noted. Serum samples collected during the acute phase of the disease from SG-1 (n = 6) and SG-2 (n = 11) were analyzed using Luminex assay to assess the soluble cytokine and chemokine markers. Of the 22 markers used for analysis, only IL-8 and GM-CSF were signi cantly higher in SG-2, and a near signi cant upregulation was also observed with TNF-α and IL-10 compared to SG-1 (Fig. S2A). The levels of all of the rest 18 biomarkers are shown in Fig. S2B. Cumulatively, the transcriptomics and cytokine array data suggest that there were no signi cant in ammatory differences with severity in patients who survived CCHF infection.
Transcriptomics changes in the acute phase and post-CCHF recovery. Due to the natural heterogeneity in human cohorts, we used longitudinal samples from 12 patients (SG-1: n = 5; SG-2: n = 7) to perform differential expression analyses for each infected patient between the time of infection and approx. 1year post-recovery (Range: 46-57 weeks). We observed a clear separation between the acute phase and the recovered phase samples using UMAP clustering (Fig. S3). The DGE pro le for the acute phase However, the pathways related to the down-regulated genes were mainly related to antiviral defense mechanism-associated pathways including innate immune responses like Th1, Th2, and Th17, cell differentiation, the NF-kB pathways, chemokine signaling pathway, etc. (Fig. 1C). Interestingly, although the HIF-1 signaling pathway was downregulated, the expression of the HIF-1 targeted genes that play a role in glycolysis and coding for the glycolytic enzymes, Enolase 1 (ENO1) and Phosphoglucomutase 2 (PGM2) was signi cantly upregulated, while hexokinase 1 (HK1) were downregulated (Supplementary Data File 1). However, when we compared the acute phase with the recovered phase in SG-1, and SG-2 separately, there was a distinct differential expression of genes (DEG) pro le. In SG-1 the differentially expressed genes were signi cantly fewer (n = 1617, upregulated: 954 and downregulated: 663) compared to those in SG-2 (n = 4256, upregulated: 2182 and downregulated: 2074) ( Fig. 1D and 1E). This indicates that the disease severity had a signi cant effect on gene expression pro ling during the acute phase while comparing when they recovered. There were 1451 overlapping genes between SG-1 and SG-2 that were differentially up-regulated (n = 882) and downregulated (n = 569). Using gene ontology (GO) analysis after removal of the redundant terms using REVIGO 15 , the top two GO terms that were signi cantly upregulated were the type I interferon signaling pathway (GO:0060337) and the regulation of viral genome replication (GO:0045069) (Fig. 1F).
Interferon signaling related pathways in CCHFV infection. We further explored the alteration of gene expression in all the interferon signaling related pathways by performing supervised analysis with the genes that are part of 13 GO terms and are part of both type-I and type-II interferon signaling pathways ( Fig. 2A). Of the IFN-regulated genes, IFI27 (ISG12) showed the most robust upregulation (Fig. 2B). This was further supported by the RNAscope analysis targeting the IFI27 transcript in the SW13 cell line infected with CCHFV strain IbAr10200 (Fig. 2C). Among the pathways, the maximum signi cantly differentially expressed genes were part of cellular response to IFN-γ (GO:0071346) (n = 57) followed by the Type-I interferon (IFN-I) signaling pathway (GO:0060337) (n = 35) ( Fig. 2A) and the majority of genes responsible for IFN-I signaling (89%; 31/35) were upregulated in the acute phase (Fig. 2D). IFI27 has been previously shown to be upregulated speci cally during virus infection and has the ability to restrict several viral infections. Apart from this ISG20, ISG15, Mx1, Mx2 and several IFN-α-inducible proteins (IRFs) showed upregulation in the acute phase.
Quantitative temporal proteomics analysis identi ed modulation of key metabolic process and signaling pathwaysin vitro. Our longitudinal transcriptomics analysis of CCHF patient samples revealed alternation in the several key metabolic process and signaling pathways during acute phase of infection. To understand the global changes in the host response during CCHFV infection, we infected Huh7 cells with CCHFV and used a time-course proteomic experiment for 24 and 48hpi using TMT-labelling based massspectrometric analysis (Fig. 3A). The global proteome analysis revealed changes in several pathways that were related to metabolism and the mTOR/HIF-I signaling pathways (Fig. 3B). This was in line with the changes we observed in the patient transcriptomics analysis ( Fig. 1B and 1C). Overall, 16 pathways were observed both in patients' transcriptomics and in vitro infection assays in Huh7 cells ( Fig. 3B and Fig. S4).
Both our patient data and the in vitro infection model indicated hijacking of the central carbon metabolism (CCM) by CCHFV to meet biosynthetic and bioenergetic demands that could be linked with the Akt/mTOR/HIF-1 pathway (Fig. 4A). To determine the requirement of the CCM and Akt/mTOR/HIF-1 pathway, we blocked glycolysis and glutaminolysis, the two key pathways of CCM in SW13 cells, using 2-Deoxy-D-glucose (2-DG) and 6-diazo-5-oxo-L-norleucine (DON) respectively ( Fig. 4A) following infection.
Infectivity of CCHFV, quanti ed as relative CCHFV L-gene levels in cells lysates, showed signi cant decrease in 2-DG treated cells (12-fold, p < 0.0001) and DON treated cells (33-fold, p < 0.0001) (Fig. 4B). We used Akt inhibitor MK2206, to modulate the Akt/mTOR/HIF-1 pathway that we earlier showed effective in blocking SARS-CoV-2 replication in vitro 14 . MK2206 signi cantly inhibited the CCHFV replication (3-fold, p < 0.0001) to a greater extent. The blocking of the glutaminolysis had maximum effect in blocking the viral replication. It was interesting to note that while 2-DG and DON induced inhibition of viral replication also inhibited the stimulation of CCHFV induced ISGs; ISG15, IFIT1, MX1 and MX2, blocking of viral replication by MK-2206 did not cause any signi cant reduction in the levels of the ISGs (Fig. 5C).
Temporal dynamics of interferon responsein vitro. To identify the CCHFV induced changes in the interferon protein landscape, we used the GO terms for the cellular response to IFN-γ (GO:0071346) and the Type-I interferon signaling pathway (GO:0060337) in our time-course proteomic data. The temporal changes in the interferome (cluster of interferon genes) were represented as a heat-map in Fig. 4A. At 24hpi, several ISGs, like Mx1, ISG20 and IFI16, that were upregulated in the acute-phase patient samples were also signi cantly elevated in infected Huh7 cells (Fig. 4B). Mx1 retained a sustained elevation 48 hpi, and ISG15 was also signi cantly elevated (Fig. 4C). The most interesting ISG was Mx2 that showed increased mRNA levels 24hpi in SW13 cells, showed a protein level suppression at 24hpi and a signi cant elevation at 48hpi in Huh7 cells (Fig. 4D). We have previously shown the anti-CCHFV role of Mx1 and the interplay between ISG15 and CCHFV was well documented by us 16 as well as others 17 . Both our transcriptomics and proteomics data from infected patients and infected cell lines respectively, showed induction of ISG20, a 3′-to-5′ exonuclease. ISG20 has been reported to have anti-viral activity against several bunyaviruses 18 . To ascertain its effect on CCHFV replication, we used doxycycline inducible stable cell lines that expressed human ISG20 and CAT protein as a control. Following infection with CCHFV at a MOI of 1 for 48 h, there was some degree of inhibition noted in the doxycycline induced ISG20 cell line (p = 0.26) and no inhibition was observed in the CAT control cell line (Fig. 4E). Though not signi cant, the results are suggestive of an antiviral property of ISG20 that is dependent on the enzymatic activity and this requires further investigation. Furthermore, there is also a possibility that the virus has evolved measures to counteract these antiviral responses. During the full course of infection, there was signi cant suppression of IRF3, the transcription factor for IFNβ production, indicating viral mediated evasion or delay in activation of the immune response. The time-dependent stimulation of different ISGs by CCHFV suggests differential regulation of the IFN pathway and does not rule out an IFN-I independent mechanism of activation of ISGs.

Discussion
Metabolic reprogramming during innate antiviral immune responses changes the subsequent phases of disease progression and in uences the adaptive immune response. In this study, using the genome-wide transcriptomic analysis of a longitudinal patient cohort, and temporal proteomics from in vitro infection assays in Huh7 cells, we show that during CCHFV infection there is metabolic reprogramming of host cells towards central carbon metabolism that play a central role in viral replication. Blocking the two key central carbon metabolic pathways, glycolysis and glutaminolysis and Akt/mTOR/HIF-1 signaling pathway controlled viral replication in vitro. Moreover, IFN-I mediated antiviral mechanisms were also activated with elevated key antiviral ISGs (ISG12, ISG15, ISG20), and MXs (Mx1 and Mx2).
Both DNA and RNA viruses have been known to alter the host cell metabolism to promote their replication 19 . The transcriptomics data on patient materials and in vitro cell culture assays indicated transient dysregulation of key metabolic processes of the central carbon metabolism, glycolysis, TCAcycle after CCHFV infection, which is known to promote replication in several other RNA viruses including HIV-1, rubella virus, dengue virus, rhinovirus, hepatitis C virus (HCV), in uenza virus etc 19,20,21,22,23 . Blocking glycolysis and glutaminolysis resulted in severe suppression of CCHFV replication suggesting the viral need for these pathways for its e cient replication. The upregulation of Glycolysis in CCHFV infection could be mediated by Akt/mTOR/HIF pathway, as blocking Akt/mTOR signaling following infection showed dramatic decrease in viral replication and furthermore in our previous study we have shown that exogenous nitric oxide that is known to regulate the HIF-1 via the Akt/mTOR pathway under normoxic conditions 24 , inhibited CCHFV in vitro 25 . However, in our patient transcriptomics data we also noted signi cant downregulation of genes belonging to the PI3K/Akt pathway during acute CCHFV infection, indicating the dysregulation of a pathway that is critical for the cellular defense against invading pathogens. Interestingly, an in vitro study in another Bunyaviridae virus, Rift valley fever virus (RVFV), identi ed the inhibition of the PI3K/Akt pathway by dephosphorylation of the AKT and Forkhead box protein O1 (FoxO1) 26 . In our study we observed that the majority of the coding genes in the FoxO signaling pathway were downregulated by CCHFV, including the FOXO transcription factors FOXO1 and FOXO3, that can act as negative feedback regulators of the innate cellular antiviral response 27 . FoxO1 and FoxO3 also play an essential role in the immunometabolic dynamics and are important targets for glycolysis and gluconeogenesis 28, 29 . Another critical signaling pathway that was downregulated during the CCHFV infection was AMP-activated protein kinase (AMPK) signaling that has been known to support IFN-I production through metabolic reprogramming. 30 In recent years, it has been recognized that metabolic pathways play a role in mounting effective IFN-I response 31 and the mechanisms are still not well understood. We observed a marked inhibition of CCHFV infection induced ISGs like ISG15, IFIT1, MX1 and MX2 mRNA levels upon blocking glycolysis and glutaminolysis. This could be either due to inhibition of CCHFV replication resulting in reduced replication products that could not elicit type-I IFN signaling pathways or blocking of glycolysis and glutaminolysis did not meet the energy requirement for successful progression of the IFN signaling cascades. Interestingly the inhibition of CCHFV replication by MK2206 that blocks Akt did not cause signi cant reduction in ISGs mRNA. Akt has been shown to regulate mRNA translation of ISGs 32 and thus its effect on the mRNA levels may not be profound and in this study, we did not determine the functionality of the IFN response. However, ability of the CCHFV virus to induce ISG expression in absence of viral replication under Akt inhibition suggests towards the activation of IFN-signaling cascades and importance of CCM in regulation of IFN-I pathways during viral infection. pre-treatment of cells with IFN-α was found to be inhibitory to CCHFV 12 . Though CCHFV is inhibited by the IFN -response, not many ISGs with anti CCHFV activity have been identi ed apart from MxA although, ISG20 and PKR have been proposed 16 to have anti-CCHFV activity. In the present study we observed that several ISGs with known or proposed anti-CCHFV activity, i.e. Mx1, ISG15 and ISG20 or not de ned for CCHFV like IFIT3, IFITM3, IFI16 and OAS3 were upregulated in the acute phase CCHFV patient samples as well as in the cell-infection model. Among these, ISG20 has been shown to have antiviral activity against related-Bunyaviruses 18 , and we observed some degree of inhibitory activity of ISG20 on CCHFV replication.
Like IFN-I, proin ammatory cytokines are important for effective innate immunity. CCHFV patients are known to have higher levels of plasma cytokine and chemokine compared to healthy people. Elevated levels of IL-6, IL-10 and TNF-α are most commonly associated with fatal outcomes 36, 37,38 . A few studies have suggested that the co-elevation of IL-6 and TNF-α are strongly associated with fatality 38 . Our study did not include samples from fatal cases and thus we performed a cytokine array and compared moderate and severe cases. Our results further con rmed the roles of TNF-α, IL-10, IL-8 and GM-CSF since these showed signi cant or near-signi cant elevation in severe cases. Korva et al., also observed signi cant differences in IL-10 and GM-CSF between moderate and severe cases 36 . In another study by Hulva et al., changes in IL-6 and TNF-α were noted between moderate and severe cases, but no signi cant change in IL-8 was observed 39 . Although IL-6 is often associated with disease severity and fatality, we did not observe any signi cant difference in IL-6 levels.
In conclusion, our study provides a comprehensive description of the host-immune response against CCHFV that can explain viral pathogenesis. The interplay of the metabolic reprogramming and the IFNmediated host antiviral mechanism could provide attractive options for therapeutic intervention of CCHF. The speci c role of the different ISGs that were regulated during the course of infections need to be studied further to provide important clues for the development of antivirals or vaccines.

Ethics statement and biosafety
This study was approved by the Local Research Ethics Committee of the Ankara Numune Education and Research Hospital, Turkey (Protocol # 17-1338) and Regional Ethics Committee, Stockholm (Dnr. 2017-/1712-31/2). All patients and/or their relatives were informed about the purpose of the study and signed a consent form before collection.

Study design, patients, and sample collection:
We enrolled 18 adult patients (≥18 years) diagnosed with CCHF who were followed up by the clinical service of Infectious Diseases and Clinical Microbiology of Sivas Cumhuriyet University Hospital, Sivas, Turkey. The CCHF patients were divided into three groups using the SGS scores of 1, 2 and 3 40 . Blood samples were collected on the admission day (acute stage) and from the survivors one year after their recovery (Supplementary Table 1 Factor alpha (TNF-a). Two quality control samples and a standard included with the kit were prepared as per the manufacturer's instructions. A Luminex MAGPIX instrument using Exponent software (Invitrogen, Paisley, UK) was used to obtain results in the form of MFI (median uorescence intensity), which was converted to concentration (pg/ml) using the standard curve prepared. Plate data was accepted based on negative background wells and QC data within the expected ranges. Heatmaps were generated using the R/Bioconductor package ComplexHeatmap v2.2.0 48 . Bubble plots, MA plots, Volcanoplots, violin plots and bar plots were created using R package ggplot2 v3.3.2. Network visualization was performed using Cytoscape v3.6.1 49 . Venn diagrams were constructed using the online tool InteractiVenn 50 . All the codes are available in GitHub (https://neogilab.github.io/CCHF-Turkey/). Raw RNAseq data is available in Sequence Read Archive (SRA) with temporary id SUB8608640.

Cells and viruses:
The CCHFV strain IbAr10200 (originally isolated from Hyalomma excavatum ticks from Sokoto, Nigeria, in 1966) was used in this study. The SW13-ATCC ® -CCL-105 TM and human hepatocyte-derived cellular carcinoma cell line Huh7 was obtained from Marburg Virology Laboratory, (Philipps-Universität Marburg, Marburg, Germany) and matched the STR reference pro le of Huh7 51 . CCHFV was propagated and titered in SW13 cells. The virus was serially diluted 10-fold and 100ul of each dilution was added to SW13 cells in a 96 well plate. At 48 h post-infection (hpi), cells were xed in ice-cold acetone-methanol (1:1) and stained using a rabbit polyclonal anti-CCHFV nucleocapsid antibody followed by a uorescein isothiocyanate (FITC)-conjugated anti-rabbit antibody (Thermo Fisher Scienti c, US) and DAPI (Roche, US). The number of uorescent foci in each well was counted using a uorescence microscope and the titer was determined. Human 293 FLP-IN TRex cells expressing CAT or ISG20 under the control of a Tet-on promoter were kindly provided by Ju-Tao Guo (Drexel University College of Medicine, Pennsylvania, USA). 52 The cells were propagated in DMEM supplemented with 5% Tet-negative FCS (PAA Lab, US).
In vitro infection assay in Huh7 cells and tandem mass tag (TMTpro) labelled reverse phase liquid chromatography mass-spectrometric (RPLC-MS/MS) analysis: Huh7 cells were infected with the CCHFV in triplicate, as described by us previously 14 . Brie y, Huh7 cells were infected with CCHFV IbAr10200 at a multiplicity of infection (MOI) of 1. After 1 h of incubation (37°C, 5% CO 2 ) the inoculum was removed, the cells were washed with PBS, and 2 ml DMEM supplemented with 5% heat-inactivated FBS was added to each well. Samples were collected in triplicate at 24 and 48 hpi along with controls. Following the protein digestion in S-Trap microcolumns (Proti , Huntington, NY), the resulting peptides were labeled with TMTpro tags. Labeled peptides were fractionated by high pH (HpH) reversed-phase chromatography, and each fraction was analyzed on an Ultimate 3000 UHPLC (Thermo Scienti c, San Jose, CA) in a 120 min linear gradient. Proteins were searched against the SwissProt human database using the search engine Mascot v2.5.1 (MatrixScience Ltd, UK) in Proteome Discoverer v2.4 (Thermo Scienti c) software allowing up to two missed cleavages.

Proteomics data analysis:
The raw data were rst ltered to remove missing data. Proteins detected in all samples were retained for analysis resulting in 8,501 proteins in the ltered dataset. The ltered data was then normalized by applying eight different methods using R/Bioconductor package NormalyzerDE v1.4.0 53 . The quantile normalization was found superior to other methods and was selected for further use. Differential protein expression analysis was performed using R/Bioconductor package limma v3.42.2 54 . Proteins with adjusted p-values of less than 0.05 were regarded as signi cant. KEGG pathway enrichment analysis of signi cantly regulated proteins was performed using the enrichr module of python package GSEAPY v 0.9.16. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identi er PXD022672.

ISG20 inhibition assay:
Tet-inducible HEK-293 FLP-IN TRex cells containing a CAT or ISG20 expression cassette were seeded on in six-well dishes and treated with doxycycline at a nal concentration of 2 μg/ml and incubated for 36 h. Cells were infected with CCHFV strain IbAr10200 at an MOI of 1, and supernatants were titrated at 48 hpi by enumeration of fluorescent foci and calculation of progeny virus titers as described by us previously 11 .

Metabolic perturbation and virus infection.
To inhibit glycolysis, glutaminolysis and Akt/mTOR pathway, following 1hpi (moi 0.1) the cells were Reverse transcription of extracted RNA was performed using a high capacity reverse transcription kit (Applied Biosystems, USA) for 10 min at 25°C, followed by 37°C for 120 min and 85°C for 5 min.
Quantitative RT-PCR assays were setup using the Power SYBR Green PCR Master Mix (Applied Biosystems, UK) using 250nM of primer pairs with cycling conditions: initial denaturation 95°C for 10 min, followed by 40        hpi. The size and colour gradient of the dots correspond to the adjusted P value of differential expression analysis and the log2 fold change, respectively. (E) HEK293-FLP-IN T-Rex cells that conditionally express ISG20 or human CAT protein were treated with doxycycline or left untreated for 36 h, followed by infection with CCHFV at an MOI 1 for 48hours. Cell-free culture supernatants were harvested, and progeny viral