Regulation of the linear ubiquitination of STAT1 controls antiviral interferon signaling

Linear ubiquitination is a critical regulator of inflammatory signaling pathways. However, linearly ubiquitinated substrates and the biological significance of linear ubiquitination is incompletely understood. Here, we show that STAT1 has linear ubiquitination at Lys511 and Lys652 residues in intact cells, which inhibits STAT1 binding to the type-I interferon receptor IFNAR2, thereby restricting STAT1 activation and resulting in type-I interferon signaling homeostasis. Linear ubiquitination of STAT1 is removed rapidly by OTULIN upon type-I interferon stimulation, which facilitates activation of interferon-STAT1 signaling. Furthermore, viruses induce HOIP expression through the NF-κB pathway, which in turn increases linear ubiquitination of STAT1 and thereby inhibits interferon antiviral response. Consequently, HOIL-1L heterozygous mice have active STAT1 signaling and enhanced responses to type-I interferons. These findings demonstrate a linear ubiquitination-mediated switch between homeostasis and activation of type-I interferon signaling, and suggest potential strategies for clinical antiviral therapy.

In this manuscript, the authors reported that the linear ubiquitin chain assembly complex (LUBAC) ubiquitinates STAT1 in resting cells. Linear ubiquitination of STAT1 blocks the recruitment of STAT1 to IFNAR2 and subsequent phosphorylation by JAK1. Upon IFN-IFNAR2 engagement, OTULIN removes the linear ubiquitination of STAT1 to promote its activation. Furthermore, prototype RNA viruses (Sendai virus and VSV) can induce the expression of HOIP to promote the linear ubiquitination of STAT1. Overall, this discovery is very interesting and informative in the regulation of STAT-mediated immune response. However, there are a few questions that the authors may address.
Major issues: 1. Scientifically, the manuscript is very solid and findings are supported by extensive experiments. Regarding the implication of linear ubiquitination of STAT1 in viral immune evasion, their findings did not support the conclusion. In order to support the author's conclusion, one may have to show that viruses, such as SeV and VSV, deploy specific strategies to induce this activity. Rather, it is conceivable that the virus-induced linear ubiquitination of STAT1 is a negative feedback mechanism that cells developed to avoid overt inflammation and subsequent collateral damage. 2. IFNγ, in addition to type I interferons, can induce STAT1 phosphorylation at Tyr701 that drives the homodimerization of STAT1. In the nucleus, STAT1 homodimers bind to GAS elements and promote ISG transcription. It will be interesting and meaningful to detect whether linear ubiquitination can block the activation of STAT1 in type II interferon signaling.
Minor issues: 1. Many co-IP experiments do not have input, including those shown in Figure 1d, 1g, 3f, 5e, 6e, 6f, 6g, 6h, etc. 2. When detecting the phosphorylation of STAT1, the levels of total STAT1 should also be showed. 3. In Figure 4a and 4b, the protein levels of JAK1 and STAT1 should be showed in input. 4. In Figure 4d, the protein levels of STAT1 and IFNAR2 should be showed in input. 5. In general, the manuscript is well-written. However, I would recommend proof-reading by a professional or native English-speaking individual. Some examples from the manuscript: Study the effect…to a large extent… the sentence in line 194-195… Line 215-216,… in conjugation with (line 217)(In conjunction with?)… Reviewer #2 (Remarks to the Author): In the manuscript by Zuo et al, the authors demonstrate a novel biochemical mechanism that negatively regulates STAT1 antiviral activity. Using extensive biochemical and genetic tools, the authors demonstrate that STAT1 is linearly ubiquitinated on at least two lysine residues by LUBAC. The linear ub of STAT1 inhibits its ability to signal by IFN-alpha, an antiviral cytokine. As a part of the mechanism, the linear ub of STAT1 prevents its interaction with IFNAR2 to inhibit IFN signaling. Furthermore, the authors also report that the deubiquitinase, Otulin removes the linear ub chains from STAT1 to restore the IFN signaling. Finally, the authors utilize Hoil-1+/-mice to study the biological significance of their study, and these mice exhibit reduced viral replication presumably due to the enhanced IFN-STAT1 signaling.
The study is very interesting, and using highly sophisticated biochemical approaches, the authors clearly established a novel posttranslational modification on STAT1 to regulate the IFN signaling. IFN is an antiviral cytokine; however, excessive IFN signaling can lead to undesired outcomes including autoimmune reactions and diseased states. Therefore, this study uncovers a new mechanism that may help develop new therapy based on the linearly ubiquitinated STAT1.

This is a very good comment! We absolutely agree with the reviewer about the difference between viral evasion and host negative feedback.
Thus, we corrected our description and conclusion in the "viral evasion" section to reflect the actual negative feedback mechanism of STAT1 linear ubiquitination induced by viral infection. Please see Line 306-335.
In addition, to better demonstrate the negative feedback mechanism,

When detecting the phosphorylation of STAT1, the levels of total STAT1
should also be showed.
We have added the total STAT1 levels in the new Fig. 2c-e, 5i, 6k, 7b, S2c, S3d, S4a and S5i. Figure 4a and 4b, the protein levels of JAK1 and STAT1 should be showed in input.

In
We have added the protein levels of JAK1 and STAT1 as the input in Fig.   4a and 4b. Figure 4d, the protein levels of STAT1 and IFNAR2 should be showed in input.

In
Thanks! We have added the protein levels of STAT1 and IFNAR2 as the input in Fig. 4d.

Reviewer #2 (Remarks to the Author):
In the manuscript by Zuo et al, the authors demonstrate a novel biochemical mechanism that negatively regulates STAT1 antiviral activity. Using extensive biochemical and genetic tools, the authors demonstrate that STAT1 is linearly ubiquitinated on at least two lysine residues by LUBAC. The linear ub of STAT1 inhibits its ability to signal by IFN-alpha, an antiviral cytokine. As a part of the mechanism, the linear ub of STAT1 prevents its interaction with IFNAR2 to inhibit IFN signaling. Furthermore, the authors also report that the deubiquitinase, Otulin removes the linear ub chains from STAT1 to restore the IFN signaling. Finally, the authors utilize Hoil-1+/-mice to study the biological significance of their study, and these mice exhibit reduced viral replication presumably due to the enhanced IFN-STAT1 signaling.

Virus infection induces IFN-beta in all cell types whereas IFN-alpha expression is restricted primarily to the myeloid cells. It would be relevant to also show that the new mechanism is valid in IFN-beta treated cells as well to
broaden the significance of the study.