TYK2 kinase activity is required for functional type I interferon responses in vivo

Abstract

Tyrosine kinase 2 (TYK2) is a member of the Janus kinase (JAK) family and is involved in cytokine signalling. In vitro analyses suggest that TYK2 also has kinase-independent, i.e., non-canonical, functions. We have generated gene-targeted mice harbouring a mutation in the ATP-binding pocket of the kinase domain. The Tyk2 kinase-inactive (Tyk2(K923E)) mice are viable and show no gross abnormalities. We show that kinase-active TYK2 is required for full-fledged type I interferon- (IFN) induced activation of the transcription factors STAT1-4 and for the in vivo antiviral defence against viruses primarily controlled through type I IFN actions. In addition, TYK2 kinase activity was found to be required for the protein's stability. An inhibitory function was only observed upon over-expression of TYK2(K923E)in vitro. Tyk2(K923E) mice represent the first model for studying the kinase-independent function of a JAK in vivo and for assessing the consequences of side effects of JAK inhibitors.

Figure 1. TYK2^K923E is enzymatically inactive and generation of Tyk2^K923E mice.

A. The in vitro kinase activity assay was performed in a TYK2-deficient cell line transiently transfected with plasmids encoding GFP, wild-type TYK2 or kinase-inactive TYK2^K923E. TYK2 and TYK2^K923E proteins were immunoprecipitated from cell extracts and subjected to an in vitro kinase assay using GST-IFNARcyt as an exogenous substrate (left panel). TYK2 was immunoprecipitated from whole cell extracts and Western Blot analysis performed to detect phosphorylated TYK2 (pTyk2, upper right panel) or TYK2 protein (lower right panel). B. Scheme of the murine Tyk2 locus from exons 9-24 (black boxes). The point mutations introduced in exon 20 resulting in the amino acid exchange K>E and the introduction of the BspTI restriction endonuclease site are depicted. The neomycin resistance cassette (neo^r, white box) flanked by loxP sites (black triangles) was inserted into the intron sequence between exons 21 and 22. The lower scheme shows the targeted locus with the restriction sites important for Southern blot analysis. Note that after germline transmission the neo^r cassette was excised to leave a single loxP site in the mutated allele. C. Southern blot analysis using a non-radioactively labelled 471 bp neo^r probe verified correct targeting and lack of heterologous integration in the ES cell clone 1, whereas two other clones (2 and 3) were not correctly targeted. D. DNA from WT (+/+), heterozygous (+/m) or homozygous Tyk2^K923E (m/m) mouse tails was used to amplify a 710 bp fragment with primers surrounding exon 20. The amplicons were digested with BspTI resulting in a 498 bp and a 212 bp fragment only in the Tyk2^K923E alleles. E. Conventional genotyping of mouse tails results in a 678 bp fragment corresponding to the WT and a 778 bp fragment specific for Tyk2^K923E.

Figure 2. TYK2^K923E protein level is reduced and TYK2 differs organ-specifically.

A. WT, Tyk2^−/− and Tyk2^K923E mice were used to prepare whole cell extracts from BMMΦs, T cells and various organs (as indicated). Levels of expression of TYK2 and JAK1 were determined by immunoprecipitation and Western blot analysis. NFκB-p65 was used as input control. TYK2^K923E protein levels were quantified using ImageJ software for Mac OS X (open source, http://rsb.info.nih.gov/ij/index.html) and were between 13% and 30% in BMMΦs and approximately 58% in T cells compared to WT. B. Total RNA was isolated from WT and Tyk2^K923E BMMΦs and cDNA was used to analyse Tyk2 mRNA expression normalized to the housekeeping gene Ube2D2. Results from 4 independent experiments are shown (n = 6 per genotype). C. BMMΦs were treated with the proteasomal inhibitor MG-132 (50 µM), the autophagy-lysosome inhibitor 3-MA (10 mM) or the lysosome-acidification inhibitor bafilomycin A₁ (80 nM) for the indicated period of time (upper panel), for 11 h (middle panel) or 48 h (lower panel). Whole cell extracts were used to determine TYK2 and JAK1 expression levels by immunoprecipitation and Western blot analysis. As a control, a Western blot for HO-1 was performed. D. From day 5 after isolation of WT BMMΦs, cells were treated with JAK inhibitor I (panJAK inhibitor; 15 nM upper panel and 300 nM lower panel) for the indicated period of time. TYK2 and JAK2 expression levels were analysed as described in (A and C); NFκB-p65 was used as input control.

Figure 3. IFN treatment leads to similar activation of JAKs and STATs in Tyk2^K923E and Tyk2-deficient cells.

BMMΦs were treated with IFNβ (500 U/ml) for 20 min or left untreated. Whole cell extracts were used to determine levels of JAK1 tyrosine phosphorylation and JAK1 expression (left panel) and of TYK2 tyrosine phosphorylation and TYK2 expression (right panel) by immunoprecipitation and Western blot analysis. B.-D. BMMΦs were treated with IFNα (500 U/ml), IFNβ (100 U/ml) or IFNγ (100 U/ml) for 20 min or left untreated. Whole cell extracts were used to determine STAT1α/β tyrosine phosphorylation and levels of STAT1α/β expression (B), levels of STAT2 tyrosine phosphorylation and STAT2 expression (C) and levels of STAT3 tyrosine phosphorylation and STAT3 expression (D) by Western blot analysis. E. BMMΦs were treated with IFNβ (10, 100 or 500 U/ml) for 20 min or left untreated and Western blot analysis performed as described in (B). F. NK cells were treated with the indicated doses of IFNβ for 20 min or left untreated. Levels of tyrosine phosphorylation and protein expression of STAT1α/β and STAT4 were analysed by Western blot. G. NK cells were treated with IFNα (500 U/ml) for the times indicated and STAT1 and 4 analysed as described in (F); H. NK cells were treated with IFNβ (100 U/ml) for the times indicated and STAT1 and 4 analysed as described in (F); ERK p85 (B, C, E-H) and ERK p42 (D) served as a loading control.

Figure 4. Transcriptional induction of IFN-responsive genes is similar in Tyk2^K923E and Tyk2^−/− cells.

A.-C. WT, Tyk2^−/− and Tyk2^K923E BMMΦs were treated with IFNα (500 U/ml), IFNβ (100 U/ml) or IFNγ (100 U/ml) for 6 h or left untreated. Total RNA was extracted, reverse-transcribed and analysed by RT-qPCR for expression of Oas1a, Ifit1 (A), Cxcl1, Socs1 (B) and Irf7, Tap1 (C). Ube2D2 was used for normalization and expression levels were calculated relative to untreated WT cells. Data are derived from three independent experiments and depicted as mean values (+/− SE). D. WT, Tyk2^−/− and Tyk2^K923E BMMΦs were treated with indicated doses of IFNβ for 6 h. Target gene expression was determined as described in A-C. Mean values (+/− SD) derived from two independent experiments are depicted. Note that due to sample size a statistical analysis was not performed.

Figure 5. Tyk2^−/− and Tyk2^K923E mice show similarly increased susceptibility to virus infection.

VSV (A) was administered intranasally (i.n.) and EMCV (B) intraperitoneally (i.p.). Mice were monitored twice daily for survival over a two-week period. Data are derived from two independent experiments (n = 20/genotype and n = 21/genotype for A and B, respectively).