Hypoxia Regulates Endogenous Double-Stranded RNA Production via Reduced Mitochondrial DNA Transcription

Hypoxia is a common phenomenon in solid tumours strongly linked to the hallmarks of cancer. Hypoxia promotes local immunosuppression and downregulates type I interferon (IFN) expression and signalling, which contribute to the success of many cancer therapies. Double-stranded RNA (dsRNA), transiently generated during mitochondrial transcription, endogenously activates the type I IFN pathway. We report the effects of hypoxia on the generation of mitochondrial dsRNA (mtdsRNA) in breast cancer. We found a significant decrease in dsRNA production in different cell lines under hypoxia. This effect was HIF1α/2α-independent. mtdsRNA was responsible for induction of type I IFN and significantly decreased after hypoxia. Mitochondrially encoded gene expression was downregulated and mtdsRNA bound by the dsRNA-specific J2 antibody was decreased during hypoxia. These findings reveal a new mechanism of hypoxia-induced immunosuppression that could be targeted by hypoxia-activated therapies.

2 ABSTRACT: 28 Hypoxia is a common phenomenon in solid tumours strongly linked to the hallmarks of cancer. 29 Hypoxia promotes local immunosuppression and downregulates type I interferon (IFN) 30 expression and signalling, which contribute to the success of many cancer therapies. Double-31 stranded RNA (dsRNA), transiently generated during mitochondrial transcription, 32 endogenously activates the type I IFN pathway. We report the effects of hypoxia on the 33 generation of mitochondrial dsRNA (mtdsRNA) in breast cancer. We found a significant 34 decrease in dsRNA production in different cell lines under hypoxia. This was HIF1α/2α-   As 0.1% hypoxia had a greater effect than 1% hypoxia ( fig. 1a), subsequent experiments were 109 performed under 0.1% hypoxic conditions. A time course in hypoxia for 4h, 8h, 16h, 24h and 110 48h showed that 4h in hypoxia was enough to lower IFNβ promoter stimulation and it was 111 maintained up till 48h ( fig. 1b left panel). The time course for recovery after reoxygenation 112 following 48h hypoxia was evaluated at 15min, 30min, 1h, 2h, 4h, 8h, 16h and 24h.    To analyse which RNA species from the total RNA were responsible for the IFNβ promoter     fig. 1a). This is potentially related to its cytosolic thymidine kinase (TK1) 192 deficiency. Mitochondrial thymidine kinase (TK2) is not cell cycle regulated 26 . As the cytosolic 193 and mitochondrial thymidine triphosphates are in rapid equilibrium and mainly produced by 194 TK1 27 , it is possible that there is a more steady state of mtDNA replication, with a stable source 195 of nucleotides from one compartment, which is not cell cycle dependent, in TK1 deficient cells.

196
It was also reported that hypoxia decreased protein expression of mitochondrial ribosomal  Altogether, these data suggest that hypoxia leads to lower mtDNA transcription and thus lower 203 production of dsRNA available to trigger the type I IFN response.  Interestingly, mitochondrial encoded genes were downregulated by hypoxia in the 215 mitochondrial fraction but not affected in the cytosolic fraction ( fig. 6a).

216
The IFNβ promoter assay was then performed using these mtRNA and cytosolic RNAs.  To assess the composition of the dsRNA pool in hypoxia, dsRNA pull-down was performed 223 using the J2 antibody in MCF7 cells exposed to normoxia or 0.1% hypoxia for 48h, and the 224 resultant RNA was sequenced. Reads were normalised as transcript per million (TPM).

225
Interestingly, the percentage of mitochondrial reads was significantly lower in hypoxia than in  Protein 3) in MCF7 exposed to 0.1% hypoxia for 48h (siBNIP3 and control siRNA, siCON).

235
In the knock-down cells, BNIP3 showed no induction in hypoxia, either at mRNA or protein  which could decrease gene transcription. G-TPP has also been described to induce 318 mitophagy 42 , but as shown above, it is not likely to be the mechanism to decrease dsRNA. The 319 similarity of reduction by these 2 inhibitors to that induced by hypoxia, and lack of further 320 suppression suggest these pathways could overlap e.g. by inhibiting RNA synthesis.

321
Basal levels of dsRNA in different tissues without hypoxic stress was assessed using human 322 total RNA, assuming that the assay measured only dsRNA. RNA from testis, brain, heart and 323 kidney were strong activators of IFNβ promoter, and these tissues also showed higher concentrations in tissues such as brain, heart and kidney could stimulate higher turnover 44 .

331
Interestingly, some tumour types such as bladder, breast, esophageal, head and neck, kidney 332 and liver showed significantly lower mtDNA content than paired adjacent normal tissue, and 333 this was associated with lower patient survival 45 . Expression could be even lower in hypoxic 334 areas having impact in anticancer therapies that rely on functional type I IFN signalling.

335
To sum up, we have shown that hypoxia caused significantly lower mtdsRNA production, 336 probably due to a decrease in mitochondrial transcription rather than increased degradation, 337 thus leading to lower activation of IFNβ promoter, and consequently to lower type I IFN     Immunofluorescence for dsRNA 400 Cells were plated on coverslips (VWR Collection) and exposed to 0.1% O2 hypoxia for 48h.