key: cord-0867505-gv7pnnja
authors: Karlowitz, Rebekka; Stanifer, Megan L.; Roedig, Jens; Andrieux, Geoffroy; Bojkova, Denisa; Smith, Sonja; Kowald, Lisa; Schubert, Ralf; Boerries, Melanie; Cinatl, Jindrich; Boulant, Steeve; van Wijk, Sjoerd J. L.
title: USP22 controls type III interferon signaling and SARS-CoV-2 infection through activation of STING
date: 2022-02-02
journal: bioRxiv
DOI: 10.1101/2022.02.01.478628
sha: 5b9dc67520468763f054ce499be7053f407f9c2e
doc_id: 867505
cord_uid: gv7pnnja

Pattern recognition receptors (PRRs) and interferons (IFNs) serve as essential antiviral defense against SARS-CoV-2, the causative agent of the COVID-19 pandemic. Type III IFN (IFN-λ) exhibit cell-type specific and long-lasting functions in autoinflammation, tumorigenesis and antiviral defense. Here, we identify the deubiquitinating enzyme USP22 as central regulator of basal IFN-λ secretion and SARS-CoV-2 infections in native human intestinal epithelial cells (hIECs). USP22-deficient hIECs strongly upregulate genes involved in IFN signaling and viral defense, including numerous IFN-stimulated genes (ISGs), with increased secretion of IFN-λ and enhanced STAT1 signaling, even in the absence of exogenous IFNs or viral infection. Interestingly, USP22 controls basal and 2’3’-cGAMP-induced STING activation and loss of STING reversed STAT activation and ISG and IFN-λ expression. Intriguingly, USP22-deficient hIECs are protected against SARS-CoV-2 infection, viral replication and the formation of de novo infectious particles, in a STING-dependent manner. These findings reveal USP22 as central host regulator of STING and type III IFN signaling, with important implications for SARS-CoV-2 infection and antiviral defense.

type III (IFN-λ1, -λ2, -λ3 and -λ4) IFNs, the expression of IFN-specific receptors is cell 86 type restricted and determines IFN responses. For example, the type I IFN receptor 87 (IFNAR) is ubiquitously expressed in many tissues, whereas expression of the type III 88 IFN receptor IFNLR1 is mainly limited to epithelial cells, e.g. the gastro-intestinal and 89 respiratory epithelium 6-8,11,12 . Although type I and type III IFNs induce similar ISG 90 signatures, type I IFNs generally trigger a more rapid increase and decay of ISG 91 expression 7 . In addition, IFN-λs have been described to be first-in-line against viral 92 infections and might inhibit viral spread without triggering inflammatory responses, 93 depending on IFN-λ receptor expression 7,13,14 . IFNs and IFN-related cytokines revealed only low basal levels of secreted   225 IFN-α and IFN-β, suggesting that these cytokines might only weakly contribute to the 226 observed ISG signature ( Figure 3C) . Surprisingly, the secretion and expression of IFN-227 λ1, a type III IFN, was strongly upregulated in USP22 KO HT-29 cells compared to 228 control cells (Figure 3C & D) . In addition, loss of USP22 expression also induced 229 elevated basal secretion of the pro-inflammatory cytokines CXCL10 and IL-8 and minor 230 changes in the secretion of IFN-α2 and GM-CSF, compared to controls. (Figure 3C ). Figure 5D ). In line with 315 these observations, basal and 2'3'-cGAMP-induced STING ubiquitination was also 316 increased in USP22 KO HT-29 cells, compared to NHT control cells ( Figure 5E ). 317 Together, these findings suggest that USP22-mediated effects on type III IFN might be 318 predominantly regulated by activating STING ubiquitination and lesser through auto-319 or paracrine IFN signaling. 320 321 fold change mRNA normalized to NHT, untreated 

Germany for help and support with performing the microarray analysis and Christina 677 Hugenberg for proofreading

WI 5171/1-1, FU 436/20-1 and project-ID 259130777 679 -SFB 1177), the Deutsche Krebshilfe (70113680), the Frankfurter Stiftung für 680 krebskranke Kinder and the Dr. Eberhard and Hilde Rüdiger Foundation. M.B. is 681 supported by the DFG -CRC 850 subprojects C9 and Z1, CRC1479 (Project ID: 682 441891347-S1)

the German Federal Ministry of Education and Research by 684 MIRACUM within the Medical Informatics Funding Scheme (FKZ 01ZZ1801B). S.B. 685 was supported by DFG project numbers 415089553

21/5/1), the BMBF (01KI20198A) and within the 688

COVID-19. M.L.S. was supported by the 689 BMBF (01KI20239B) and DFG project 416072091

Ubiquitination, ubiquitin-like modifiers, and 940 deubiquitination in viral infection

The Role of Atypical Ubiquitin Chains in the 943 Regulation of the Antiviral Innate Immune Response

Ubiquitin in the activation and 946 attenuation of innate antiviral immunity

The ubiquitin ligase TRIM56 regulates innate immune 949 responses to intracellular double-stranded DNA

TRIM32 protein modulates type 952 I interferon induction and cellular antiviral response by targeting MITA/STING 953 protein for K63-linked ubiquitination

Ubiquitination of STING at lysine 224 controls 956 IRF3 activation

Loss of Usp22 enhances histone H2B monoubiquitination and 1014 stimulates intracellular and systemic interferon immunity. bioRxiv

Positively Regulates RCAN1 Protein Levels Through RCAN1 De-Ubiquitination

USP18 -a multifunctional component 1020 in the interferon response

Constitutive type I interferon modulates homeostatic balance through tonic 1023 signaling

Unphosphorylated STAT1 prolongs the expression of 1025 interferon-induced immune regulatory genes

IFN-lambdas mediate antiviral protection through a distinct 1028 class II cytokine receptor complex

IL-28, IL-29 and their class II cytokine receptor IL-28R

The H2B deubiquitinase Usp22 promotes antibody class switch 1040 recombination by facilitating non-homologous end joining

Recognition of cytosolic DNA activates an IRF3-1043 dependent innate immune response

STING regulates intracellular DNA-1046 mediated, type I interferon-dependent innate immunity

Decreased H2B 1049 monoubiquitination and overexpression of ubiquitin-specific protease enzyme 1050 22 in malignant colon carcinoma

Histone crosstalk between H2B monoubiquitination and H3 1053 methylation mediated by COMPASS

Histone H2B ubiquitylation is associated with elongating RNA 1056 polymerase II

Monoubiquitinated H2B is associated with the transcribed 1059 region of highly expressed genes in human cells

Functional landscape of SARS-CoV-2 cellular 1062 restriction

Adenovirus evasion of interferon-mediated innate immunity 1065 by direct antagonism of a cellular histone posttranslational modification

Monoubiquitination of human histone H2B: the factors involved 1068 and their roles in HOX gene regulation

Protease cleavage of RNF20 facilitates 1071 coronavirus replication via stabilization of SREBP1

STING is an essential mediator of the Ku70-mediated production 1074 of IFN-λ1 in response to exogenous DNA

STING-Dependent Interferon-λ1 Induction in HT29 Cells

Human Colorectal Cancer Cell Line, After Gamma-Radiation. International 1078 journal of radiation oncology

The cGAS-STING pathway drives type I IFN 1081 immunopathology in COVID-19

Analysis of relative gene expression data using 1084 real-time quantitative PCR and the 2(-Delta Delta C(T)) Method

A framework for oligonucleotide microarray 85

limma powers differential expression analyses for RNA-1090 sequencing and microarray studies

1093 generally applicable gene set enrichment for pathway analysis

Gene set enrichment analysis: A knowledge-based 1096 approach for interpreting genome-wide expression profiles

Efficient protection and isolation of ubiquitylated proteins using 1100 tandem ubiquitin-binding entities

Supplementary Information USP22 controls type III interferon signaling and SARS-CoV-2 infection through activation of STING

Institute for Experimental Cancer Research in Pediatrics

Email: vanWijk@med.uni-frankfurt.de, s.wijk@kinderkrebsstiftung-frankfurt.de Summary: Four Supplementary Figures including Supplementary Figure legends and a Supplementary Table Supplemental Figure 1 (related to Figure 1). A. Scatter plot demonstrating the changes in gene expression of CRISPR/Cas9 control (NHT) HT-29 cells with two independent single-cell HT-29 USP22 KO clones (#16 and #62)

A. Western blot analysis of phosphorylated γ-H2AX (p.γ-H2AX) and USP22 expression levels in control (nonhuman target: NHT) and CRISPR/Cas9-generated USP22 knock-out (KO) HT-29 cells (USP22 KO) subjected to vehicle or etoposide (100 µM) for 2 h. GAPDH served as loading control. Representative blots of at least two different independent experiments are shown. B. Western blot analysis of phosphorylated and total p65

Representative blots of at least two different independent experiments are shown. C. Basal mRNA expression levels of TMEM173/STING in control and USP22 KO HT-29 cells (USP22 KO #62) using qRT-PCR. Gene expression was normalized against 28S mRNA and is presented as x-fold mRNA expression compared to NHT. Mean and SD of three

A. mRNA expression levels of TMEM173/STING in control (non-human target: NHT) and CRISPR/Cas9-generated USP22 knock-out (KO) HT-29 cells (USP22 KO) using qRT-PCR. Cells were treated with ruxolitinib (5 µM) for the indicated timepoints. Gene expression was normalized against 28S mRNA and is presented as x-fold mRNA expression compared to NHT

Mean and SD of three independent experiments in triplicate are shown. ***P < 0.001, n.s. not significant

Basal mRNA expression levels of IFNA, IFNB and IFNL1 in wild-type (WT), control (non-human target: NHT) and CRISPR/Cas9-generated USP22 knock-out (KO) Caco-2 single clones (USP22 KO #1 and #6)

TCID50/mL, determined via titration of supernatant from SARS-CoV-2-infected WT, control and USP22 KO Caco-2 cells (USP22 KO #1 and #6) 6 and 24 hpi on Vero cells. Mean and SD of three independent experiments in triplicate are shown

24 hpi, supernatant additionally of NHT-, USP22 KO #1-and #6-STING dKO Caco-2 cells

The authors thank the members of the van Wijk lab for advice, discussions and support 674