Chronic CRYPTOCHROME deficiency enhances cell-intrinsic antiviral defences

Abstract

The within-host environment changes over circadian time and influences the replication and severity of viruses. Genetic knockout of the circadian transcription factors CRYPTOCHROME 1 and CRYPTOCHROME 2 (CRY1^-/-/CRY2^-/-; CKO) leads to altered protein homeostasis and chronic activation of the integrated stress response (ISR). The adaptive ISR signalling pathways help restore cellular homeostasis by downregulating protein synthesis in response to endoplasmic reticulum overloading or viral infections. By quantitative mass spectrometry analysis, we reveal that many viral recognition proteins and type I interferon (IFN) effectors are significantly upregulated in lung fibroblast cells from CKO mice compared with wild-type (WT) mice. This basal 'antiviral state' restricts the growth of influenza A virus and is governed by the interaction between proteotoxic stress response pathways and constitutive type I IFN signalling. CKO proteome composition and type I IFN signature were partially phenocopied upon sustained depletion of CRYPTOCHROME (CRY) proteins using a small-molecule CRY degrader, with modest differential gene expression consistent with differences seen between CKO and WT cells. Our results highlight the crosstalk between circadian rhythms, cell-intrinsic antiviral defences and protein homeostasis, providing a tractable molecular model to investigate the interface of these key contributors to human health and disease.This article is part of the Theo Murphy meeting issue 'Circadian rhythms in infection and immunity'.

Keywords: CRYPTOCHROME; circadian rhythms; integrated stress response; interferon; protein homeostasis; virus.

Figure 1.

Aberrant constitutive type I IFN signalling in CRYPTOCHROME (CRY) knock-out (CRY1^−/−/CRY2^−/−; CKO) cells restricts virus growth. (a) Schematic representation of antiviral defence proteins significantly upregulated in CKO mouse lung fibroblasts compared with wild-type (WT) controls (created using BioRender). (b) Ranked gene ontology (GO) enrichment analysis was carried out using GOrilla on protein lists ordered by absolute fold change in abundance in CKO primary lung fibroblasts compared with WT controls (n = 4 biological replicates). For further data analysis please refer to Munns et al. [19]. FDR, false discovery rate. (c) Single-cycle influenza A virus (IAV) growth curves were performed on WT and CKO mouse lung fibroblasts (n = 3 biological replicates; mean ± s.e.m.). IAV-infected cells and supernatant were harvested at the stated times and infectious IAV particle levels were determined by plaque assay. PFU, plaque forming units. Two-way ANOVA (time × genotype): time *p < 0.05; genotype *p < 0.05; interaction *p < 0.05. Multiple comparison significance: *p < 0.05, ***p < 0.001. (d,e) Representative western blots and quantitative analysis of STAT1 abundance in WT and CKO mouse lung fibroblasts treated with anti-IFNAR1 monoclonal antibodies (anti-IFNAR) or isotype control antibody for 144 h. Summary graph shows percentage change in expression at 144 h relative to WT isotype control antibody at time = 0 h (n = 3 biological replicates; mean ± s.e.m.). Two-way ANOVA (time × treatment): time ****p < 0.0001; treatment ****p < 0.0001; interaction ****p < 0.0001. (f,g) Representative blots and quantitative analysis of STAT1 abundance in WT and CKO mouse lung fibroblasts treated with BX795, ruxolitinib (RUX) or dimethyl sulfoxide (DMSO) vehicle control for 48 h. Summary graph shows percentage change in expression at 48 h relative to WT DMSO vehicle control at time = 0 h (n = 3 biological replicates; mean ± s.e.m.). Two-way ANOVA (time × treatment): time **p < 0.01; treatment ****p < 0.0001; interaction *p < 0.05; n.s., not significant.

Figure 2.

Loss of CRYPTOCHROME (CRY) proteins does not directly induce the altered type I interferon signature of CRY knock-out (CRY1^−/−/CRY2^−/−; CKO) cells. Ranked GO enrichment analysis was carried out using GOrilla on protein lists ordered by absolute fold change in abundance. Terms with a false discovery rate q-value < 0.05 were considered significant. Wild-type (WT) (a) and CKO (b) mouse fibroblasts were analysed after treatment with C8 or dimethyl sulfoxide (DMSO) vehicle control for 10 days (n = 2 biological replicates; n = 4 technical replicates). For further data analysis please refer to Munns et al. [19]. (c) Relative induction of interferon-stimulated genes (ISGs) in CKO cells compared with WT controls versus C8-treated WT cells compared with DMSO controls. KO, knockout. (d) Representative western blots and quantitative analysis of STAT1 abundance in WT and CKO cells treated with C8, KS15 control or DMSO vehicle control (veh) for 10 days. Summary graph shows percentage change in expression relative to WT DMSO vehicle control (n = 3 biological replicates; mean ± s.e.m.). One-way ANOVA multiple comparison significance: *p < 0.05; **p < 0.01. n.s., not significant. (e) Single-cycle influenza A virus (IAV) growth curves in WT and CKO mouse lung fibroblasts following 10 days of treatment with C8 CRY-degrader or KS15 CRY-binder control (n = 3 biological replicates; mean ± s.e.m.). IAV-infected cells and supernatant were harvested at stated times and infectious IAV particle levels were determined by plaque assay. PFU, plaque forming unit. Two-way ANOVA (time × treatment): time ****p < 0.0001; genotype p = 0.6554; interaction p = 0.9983.

Figure 3.

Altered proteostasis and stress responses underpin differences in wild-type (WT) and CRYPTOCHROME knock-out (CRY1^−/−/CRY2^−/−; CKO) antiviral state. Representative western blots and quantitative analysis of protein samples harvested from WT and CKO mouse lung fibroblasts treated with an integrated stress response (ISR) activator or inhibitor over 24 h (n = 3 biological replicates; mean ± s.e.m). (a) Phosphorylated eIF2α during tunicamycin treatment, (b) STAT1 during tunicamycin treatment, (c) RIG-I during tunicamycin treatment, (d) RIG-I during transISRIB treatment, and (e) phosphorylated TBK1 during tunicamycin treatment. Two-way ANOVA (time × treatment) multiple comparison significance: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. (f) Summary illustration of the interplay between cell-intrinsic innate defence signalling and circadian clock pathways, and future research questions (created using BioRender).