Signal transducer and activator of transcription 2 deficiency is a novel disorder of mitochondrial fission

Abstract

Defects of mitochondrial dynamics are emerging causes of neurological disease. In two children presenting with severe neurological deterioration following viral infection we identified a novel homozygous STAT2 mutation, c.1836 C>A (p.Cys612Ter), using whole exome sequencing. In muscle and fibroblasts from these patients, and a third unrelated STAT2-deficient patient, we observed extremely elongated mitochondria. Western blot analysis revealed absence of the STAT2 protein and that the mitochondrial fission protein DRP1 (encoded by DNM1L) is inactive, as shown by its phosphorylation state. All three patients harboured decreased levels of DRP1 phosphorylated at serine residue 616 (P-DRP1(S616)), a post-translational modification known to activate DRP1, and increased levels of DRP1 phosphorylated at serine 637 (P-DRP1(S637)), associated with the inactive state of the DRP1 GTPase. Knockdown of STAT2 in SHSY5Y cells recapitulated the fission defect, with elongated mitochondria and decreased P-DRP1(S616) levels. Furthermore the mitochondrial fission defect in patient fibroblasts was rescued following lentiviral transduction with wild-type STAT2 in all three patients, with normalization of mitochondrial length and increased P-DRP1(S616) levels. Taken together, these findings implicate STAT2 as a novel regulator of DRP1 phosphorylation at serine 616, and thus of mitochondrial fission, and suggest that there are interactions between immunity and mitochondria. This is the first study to link the innate immune system to mitochondrial dynamics and morphology. We hypothesize that variability in JAK-STAT signalling may contribute to the phenotypic heterogeneity of mitochondrial disease, and may explain why some patients with underlying mitochondrial disease decompensate after seemingly trivial viral infections. Modulating JAK-STAT activity may represent a novel therapeutic avenue for mitochondrial diseases, which remain largely untreatable. This may also be relevant for more common neurodegenerative diseases, including Alzheimer's, Huntington's and Parkinson's diseases, in which abnormalities of mitochondrial morphology have been implicated in disease pathogenesis.

Keywords: JAK-STAT signalling; STAT2; dynamin-related protein 1 (DRP1); mitochondrial disease; mitochondrial fission; mitochondrial fusion.

See Dasgupta et al. (doi:10.1093/awv237) for a scientific commentary on this article. Shahni et al. identify a novel mutation in STAT2, which encodes a component of the JAK-STAT cytokine signalling pathway, in three patients with severe neurological deterioration following viral infection. STAT2 is shown to regulate mitochondrial fission, suggesting a new avenue for treatment of mitochondrial diseases and possibly common neurodegenerative disorders.

Figure 1

Analysis of muscle and cultured skin fibroblasts. (A) Representative electron micrographs of muscle (longitudinal sections, top) and fibroblasts (bottom) in patients and controls. Red arrows show mitochondria. (B) Top: Electropherograms showing Sanger sequence confirmation of novel homozygous stop gain mutation c.1836 C>A (p.Cys612Ter) in STAT2 identified by whole exome sequencing in Patients 1 and 2. Parents are both heterozygous for the mutation. Bottom: The mutation leads to a stop-gain at amino acid position 612 of the highly conserved SH2 domain of STAT2 protein. P represents phosphorylation sites on STAT2.

Figure 2

Mitochondrial length analysis using TMRM staining and confocal microscopy. (A) TMRM labelled mitochondria in Patients 1–3 fibroblasts (top), transduced with wild-type STAT2 (middle), and DRP1⁻, STAT1⁻ and healthy control (HC) cells (bottom); all cells at 70% confluency. Minus sign represents autosomal dominant mutation. (B) TMRM labelled mitochondria in SHSY5Y cells, wild-type (left) and STAT2 knockout (right). (C) ImageJ quantitation of mitochondrial perimeter length in fibroblasts (top) and SHSY5Y cells (bottom). **P < 0.005.

Figure 3

Image analysis of mitochondria in transduced patient and control fibroblasts. (A) Confirmation of lentiviral transduction in patient fibroblasts using GFP tag. Patient fibroblasts transduced with STAT2, stained with TMRM (left) and GFP expression upon reintroduction of STAT2 (right). Scale bar = 20 μm. (B) TMRM stained mitochondria in STAT2 transduced patient and control fibroblasts. Mito-Morphology Macro was installed on ImageJ, which measures mitochondrial interconnectivity and elongation from epifluorescence micrographs of cells stained for mitochondria. The perimeter length of the selected mitochondria is then measured using the software. Scale bar = 50 μm. (C) Three-dimensional representations of mitochondria in Patient 1 fibroblasts before (top) and after (middle) transduction with STAT2, and control fibroblasts (bottom). Confocal z-stack images of mitochondria stained with TMRM were reconstructed using IMARIS X64 software (version 7.6.3, Bitplane) to give 3D representations of the mitochondria present in each cell. Single connected mitochondria depicted as yellow.

Figure 4

Transcript and protein expression of fission and fusion proteins and DRP1 localization. (A) Quantitative PCR analysis of DRP1, MFN1, MFN2 and OPA1 in three controls (C1–C3) and two patients (Patients 1 and 3). The samples were normalized to C1 and ACTB encoding β-actin was used as an endogenous control (n = 3). (B) Western blot analyses of 14 μg protein extracted with 1.5% n-dodecyl-β-d-maltoside from control (C1–6) and patient (Patients 1 and 3) fibroblast cultures. Blots were probed with antibodies raised against DRP1, MFN1, MFN2, OPA1, MTCO2 and TOM20 as indicated. An antibody raised against β-actin served as loading control. (C) DRP1 localization was carried out by labelling the mitochondria with TOM20, DRP1 with FITC conjugated secondary antibody and nuclei with Hoechst and visualized under a confocal microscope. Overlaid images show DRP1 co-localization with TOM20 on the mitochondria.

Figure 5

DRP1 phosphorylation in patients and controls. (A) Protein samples (20 μg) from fibroblasts and SHSY5Y cells were separated by sodium dodecyl sulphate polyacrylamide gel electrophoresis and immunoblotted with the indicated antibodies. STAT2 protein was not detectable in the three patients (P1–3) confirming nonsense mediated decay; STAT2 levels were restored after lentiviral transduction with STAT2 (top). Total DRP1 protein levels remained constant throughout (second panel from top), but DRP1 phosphorylation at serine 616 was very low in STAT2 deficient cells and increased after lentiviral transduction with wild-type STAT2 (third panel), whilst phosphorylation at serine 637 reduced after STAT2 transduction (fourth panel). Actin was used as a loading control (bottom). (B) Quantitative analysis of phosphorylation at serines 616 and 637 of DRP1 in STAT2 transduced patient and control fibroblasts (left) and SHSY5Y wild-type and STAT2 knockout cells (right). Data were normalized to the total level of DRP1 and represent the mean ± SD of three independent experiments. *P < 0.05, **P < 0.005.

Figure 6

STAT1 phosphorylation before and after stimulation with IFNα, membrane potential and apoptosis levels. (A) Representative FACS dot blot graph showing phosphorylated STAT1 in STAT2 transduced cells compared to non-transduced STAT2 deficient cells after stimulation (results for Patient 3 shown). (B) The percentage change in phosphorylated cells was calculated from the data from A [patient cells, transduced cells and SHSY5 wild-type (WT) and STAT2 knockout (KO) cells]. Data represents the mean ± SD (n = 3). *P < 0.05, **P < 0.005 (C) Membrane potential analysis in control and patient cells using TMRM staining. The mean fluorescence intensity was analysed via confocal microscopy and Image J (n = 3;>27 cells analysed for each n, ± SEM) *P < 0.05, **P < 0.005. (D) Apoptosis was analysed in fibroblasts derived from the STAT2 deficient patient (Patient 1) and compared to fibroblasts derived from healthy individuals. Cultured fibroblasts were left unstimulated or stimulated with 100 U/ml IFNα, 10 µg/ml anti-fasL IgM or phytohaemagglutinin. Cells were analysed using CellQuest Pro and the percentage of apoptotic cells defined as annexin V+7AAD+.

Figure 7

Activation of the JAK-STAT pathway by viruses and the biogenesis, fission and fusion life cycle of mitochondria. Viral infections activate a cascade of signals to produce cytokines, which then activate JAK-STAT proteins which then form a complex and translocate into the nucleus. The STAT1/2 dimer protein complex binds to IRF9 in the nucleus and the resulting trimer interacts with DNA at transcription factor binding sites, and upregulates gene expression. The dotted red line indicates effects of STAT2 on mitochondrial fission, mediated by DRP1 phosphorylation in healthy cells. Mitochondrial morphology is maintained by a balance of fission and fusion events, with disparity leading to a distinct shift in viability of the organelle, which can result in disease states. Phosphorylation of DRP1 at differing serine residues (Ser616 and Ser637) within the protein results in varied effects on DRP1 activity. Phosphorylation at Ser637 inhibits the GTPase activity of DRP1, leading to a defect in mitochondrial fission—promoting mitochondrial elongation.