MRE11-independent effects of Mirin on mitochondrial DNA integrity and cellular immune responses

Abstract

Mirin, a chemical inhibitor of MRE11, has been recently reported to suppress immune response triggered by mitochondrial DNA (mtDNA) breakage and release during replication stalling. We show that while Mirin reduces mitochondrial replication fork breakage in mitochondrial 3´-exonuclease MGME1 deficient cells, this effect occurs independently of MRE11. We also discovered that Mirin directly inhibits cellular immune responses, as shown by its suppression of STAT1 phosphorylation in Poly (I:C)-treated cells. Furthermore, Mirin also altered mtDNA supercoiling and accumulation of hemicatenated replication termination intermediates-hallmarks of topoisomerase dysfunction-while mitigating topological changes induced by the overexpression of mitochondrial TOP3A, including TOP3A-dependent strand breakage at the noncoding region of mtDNA. Although Mirin does not seem to inhibit TOP3A activity in vitro, our findings demonstrate its MRE11-independent effects in cells and give insight into the mechanisms of the maintenance of mtDNA integrity.

FIGURE 1:

MIRIN inhibits STAT1 activation and mtDNA breakage in MGME1 KO cells. (A) MGME1 KO cells show immune response activation, indicated by the elevated levels of phosphorylated STAT1. This activation is almost completely abolished by a 48 h treatment with 100 µM Mirin. (B) Schematic illustration of human mtDNA, showing the OriH-OriL fragment present in MGME1 KO cells as well as the location of HindIII restriction sites and the NCR probe used for Southern blot hybridization. The predicted fragments recognized by the probe are shown below. (C) A Southern blot of HindIII-digested DNA from wildtype (WT) and MGME1 KO cells, was probed with the NCR probe. Note the increase in the dimeric 20 kb DNA species as well as the significant reduction in the 4 kb OriH-OriL fragment abundance after 48 h of 100 µM Mirin. Ratios of 4 kb OriH-OriL /10 kb fragment from four independent experiments are shown with standard error of the mean. Statistical significance (p-value 0.0015) was calculated with a two-tailed Student's t-test. Datapoints for the measurements are shown in the graphs together with their median values and SDs.

FIGURE 2:

MRE11 knockdown does not phenocopy the effects of Mirin. (A) Knockdown of MRE11 in wildtype or MGME1 KO cells does not influence STAT1 activation or the abundance of OriH-OriL fragment (B). The original Southern blot is shown in Supplemental Figure S2. (C) In contrast to known mitochondrial proteins, such as TOMM40, COX2, and TWNK helicase, MRE11 is not protected from protease treatment in mitochondrial preparations. Digitonin permeabilizes the outer membranes, rendering the outer membrane protein TOMM40 sensitive to the proteinase treatment, while the matrix proteins COX2 and TWNK remain protected. Triton X-100 solubilizes mitochondria and exposes matrix proteins to the proteinase. Nuclear DNA polymerase δ is used as a nuclear protein contaminant control, being sensitive to proteinase already prior to detergent treatments. (D) The majority of the MRE11 is localized in the nucleus (stained with DAPI) in HEK cells. While some cytoplasmic signals are visible in long exposures, this does not colocalize with the mitochondrial network (anti-TFAM staining). Datapoints for the measurements are shown in the graphs together with their median values and SDs.

FIGURE 3:

Mirin blocks the RIG-I-agonist-dependent immune reaction in HEK cells. (A) STAT1 phosphorylation was induced in both wildtype (WT) and MGME1 KO cells by the addition of 1 µg/ml RIG-I agonist Poly (I:C) for 48 h. STAT1 activation was abolished by the simultaneous addition of 100 µM Mirin but not by 100 µM PFM39. (B) 48 h treatment of cells with PFM39 reduced the baseline STAT1 phosphorylation.

FIGURE 4:

Increased levels of mtDNA replication termination intermediates after Mirin treatment. (A) Schematic illustration of human mtDNA showing restriction cut sites and the probe locations. Details of non-relevant cut sites are omitted for clarity. (B) Replication intermediates present in the OriH-containing HincII fragment and their migration patterns on 2D-AGE. Single-stranded bubble arc (s-b) represents partially single-stranded (ss) DNA replication bubbles. Slow-moving y-forms (smy) are partially ssDNA replication forks that have progressed beyond the downstream restriction sites (*), giving rise to high-molecular weight replication intermediates. Double-stranded (ds)DNA replication intermediates are represented by bubbles (d-b) and y-arcs (y). Note that the y-arc is not complete because of the position of OriH within the fragment; its descending tip corresponds to persistent replication forks at this locus. Four-way junctional termination intermediates (ter) form at the end of replication close to the OriH. (C) 48 h 100 µM Mirin treatment of cells increases the termination intermediates (ter), as well as replication forks, paused at OriH (y) in MGME1 KO cells. In contrast, (D) a 48 h treatment with 100 µM PFM39 had no significant effect on the mtDNA replication intermediates. Representative gel images are shown from three independent experiments.

FIGURE 5:

Mirin treatment alters mtDNA topology. (A) A 48 h 100 µM Mirin treatment of wildtype (WT) HEK293 cells causes an increase in fully supercoiled (SC) mtDNA and generates a ladder of additional supercoiled forms (SC*) below the fully relaxed open circles (OC). The effect on supercoiling is not as dramatic in MGME1 KO cells. (B) 100 µM Mirin treatment not only reduces MRE11 but also TOP3A protein levels in the same cells. Some mtDNA replication-related proteins such as POLG and POLRMT are upregulated, whereas mtDNA-encoded COX2 remains unaltered.

FIGURE 6:

Mirin treatment partially reverses the effects of mitochondrially targeted TOP3A (mtTOP3A) overexpression. (A) A mtDNA topology gel from 293Trex cell lines with an empty vector insertion or tetracycline-inducible mtTOP3A construct showing the effects of transgene induction with and without 100 µM Mirin for 48 h. The bottom panel shows the 7S signal of the same gel. Note the accumulation of linear mtDNA after mtTOP3A induction, the increase in supercoiled mtDNA after Mirin treatment (compare to Figure 5) as well as the lengthening of the 7S DNA. (B) Quantitation of the abundance of linear and supercoiled mtDNA in mtTOP3A expressing cells after 48 h of 100 µM Mirin. (C) Mirin treatment decreases the mtTOP3A-induced breakage in the NCR. Compare with Figure 1C. (D) Schematic illustration of the HindIII NCR fragment analyzed in C. HMW = High-molecular weight mtDNA. Datapoints for the measurements are shown in the graphs together with their median values and SDs.