Frataxin Deficiency Promotes Excess Microglial DNA Damage and Inflammation that Is Rescued by PJ34

Abstract

An inherited deficiency in the frataxin protein causes neurodegeneration of the dorsal root ganglia and Friedreich's ataxia (FA). Frataxin deficiency leads to oxidative stress and inflammatory changes in cell and animal models; however, the cause of the inflammatory changes, and especially what causes brain microglial activation is unclear. Here we investigated: 1) the mechanism by which frataxin deficiency activates microglia, 2) whether a brain-localized inflammatory stimulus provokes a greater microglial response in FA animal models, and 3) whether an anti-inflammatory treatment improves their condition. Intracerebroventricular administration of LPS induced higher amounts of microglial activation in the FA mouse model vs controls. We also observed an increase in oxidative damage in the form of 8-oxoguanine (8-oxo-G) and the DNA repair proteins MUTYH and PARP-1 in cerebellar microglia of FA mutant mice. We hypothesized that frataxin deficiency increases DNA damage and DNA repair genes specifically in microglia, activating them. siRNA-mediated frataxin knockdown in microglial BV2 cells clearly elevated DNA damage and the expression of DNA repair genes MUTYH and PARP-1. Frataxin knockdown also induced a higher level of PARP-1 in MEF cells, and this was suppressed in MUTYH-/- knockout cells. Administration of the PARP-1 inhibitor PJ34 attenuated the microglial activation induced by intracerebroventricular injection of LPS. The combined administration of LPS and angiotensin II provoke an even stronger activation of microglia and neurobehavioral impairment. PJ34 treatment attenuated the neurobehavioral impairments in FA mice. These results suggest that the DNA repair proteins MUTYH and PARP-1 may form a pathway regulating microglial activation initiated by DNA damage, and inhibition of microglial PARP-1 induction could be an important therapeutic target in Friedreich's ataxia.

Fig 1. Introcerebraventricular injection of LPS induced more microglia activation and astrocyte activation in FA mice.

(A) FA mice receiving introcerebraventricular injection of LPS exhibited more Iba-1 immunoreactivity compared with WT mice receiving introcerebraventricular injection of LPS, and WT or FA mice received introcerebraventricular injection of vehicle (PBS). Scale bar: 20μm. (B) Quantitative measurement of Iba-1 immunoreactivities showed FA mice treated with LPS had significantly higher Iba-1 staining intensity compared with other FA mice only receiving PBS, and WT mice with or without LPS treatment. FA mice treated with LPS had significantly more Iba-1 positive staining microglia compared with other groups of mice. (C) FA mice receiving introcerebraventricular injection of LPS exhibited more GFAP immunoreactivity compared with WT mice received introcerebraventricular injection of LPS and WT, or FA mice receiving introcerebraventricular injection of vehicle (PBS). Scale bar: 20μm. (D) Quantitative measurement of GFAP immunoreactivity showed FA mice treated with LPS had significantly higher Iba-1 staining intensity compared with other FA mice only receiving PBS and WT mice with or without LPS treatment. Data are expressed as mean ± s.e.m. (t test, *p<0.05, ** p<0.01, n = 4). (E) Western blot analysis of Iba-1 and CD11b showed expression level of Iba-1 and CD11b are increased in FA cerebellum after LPS introcerebraventricular injection. (F) Quantitative measurement of Western blot bands showed cerebellums of FA mice treated with LPS had significantly higher expression level of Iba-1 and CD11b. Data are expressed as mean ± s.e.m. (t test, ** p< 0.01, n = 5).

Fig 2. 8-oxoG, PARP-1 and MUTYH are increased in cerebellum of FA mice treated with introcerebraventricular injection of LPS compared with wild-type mice received the same treatment.

(A) DNA damage marker, 8-oxoG was double stained with Iba-1. Cerebellum of FA mice has more 8-oxoG immunoreactivities and 8-oxoG/Iba-1 double-stained cells. (B) MUTYH was double stained with CD11b. Cerebellum of FA mice has more MUTYH immunoreactivity and MUTYH/CD11b double-stained cells. (C) PARP-1 was double stained with Iba-1. Cerebellum of FA mice has more PARP-1 immunoreactivity and PARP-1/Iba-1 double stained cells. Scale bar:10μm. Data are expressed as mean ± s.e.m. (t test and one way ANOVA, * p< 0.05, ** p<0.01, *** p<0.001, n = 4).

Fig 3. 8-oxoG, MUTYH and PARP-1 are increased in frataxin knockdown BV2 cells.

(A) 8-oxoG was labeled in BV2 cell culture. BV2 cells treated with frataxin siRNA have more 8-oxoG immunoreactivity compared to BV2 cells treated with control siRNA. Data are expressed as mean± s.e.m. (t test, ** p < 0.01, n = 3) (B) Western-blot of MUTYH and PARP-1 in BV2 cells showed that BV2 cells treated with frataxin siRNAs (siR5, siR6 and siR7) expressed more MUTYH and PARP-1 compared to BV2 cells treated with control siRNA. Data are expressed as mean± s.e.m. (one way ANOVA, ** p < 0.01, *** p<0.001, n = 3).

Fig 4. PARP-1 is less inducible in MUTYH-/-MEF cell by MNNG and frataxin siRNA.

(A) Western blot of PARP-1 in MEF cells showed that MNNG treatment induced higher level of PARP-1 in WT MEF cells and in MUTYH-/- MEF cells expressing human MUTYH (labeled as hMUTYH) compared to MUTYH -/- MEF cells. Data represent the densitometry results of the western blots. Data are expressed as mean± s.e.m. (one way ANOVA, * p<0.05, ** p < 0.01, n = 3)(B) Western blot of PARP-1 in MEF cells showed that frataxin siRNA transfection induced higher level of PARP-1 in WT MEF cells and in MUTYH-/- MEF cells expressing human MUTYH (labeled as hMUTYH) compared to MUTYH -/- MEF cells. Data represent the densitometry results of the western blots. Data are expressed as mean± s.e.m. (t test, * p<0.05, n = 3).

Fig 5. PARP-1 inhibitor, PJ34 attenuated microglia activation in cerebellum of FA mice treated with LPS.

(A) Both Iba-1 and PARP-1 immunoreactivities are attenuated in cerebellum of LPS treated FA mice by administration of PARP-1 inhibitor, PJ34. Merged PARP-1 and Ib1-1 staining showed less double stained cells with PJ34 treatment compared to vehicle. (B) Quantification of Iba-1 staining intensity and cell count of PARP-1/Ib1-1 doubled cells showed significant differences between PJ34 treated group and Vehicle treated group. Data are expressed as mean± s.e.m. (t test, ** p<0.01, n = 5).

Fig 6. Angiotensin II treatment exacerbates microglial activation and astrocyte activation and causes neuronal cellular damage.

(A) Western blot of angiotensin II type 1 receptor (AT1R) showed that LPS treatment induced more AT1R expression in cerebellum of FA mice compared to wild type mice. Data are expressed as mean± s.e.m. (one way ANOVA, * p<0.05, n = 5). (B) Angiotensin II infusion increased microglia activation and astrocyte activation in cerebellum of LPS treated mice showed by Iba-1 and GFAP staining. Quantitative analyses of Iba-1 and GFAP staining intensity showed significant differences between angiotensin II infusion group and PBS infusion group. Data are expressed as mean± s.e.m. (t test, * p<0.05, n = 5). (C) Angiotensin II infusion exacerbated neuronal damage showed with TUNEL staining. More TUNEL immunoreactivity was seen in cerebellum of FA mice treated with LPS combined with angiotensin II infusion compared to FA mice treated with LPS combined with PBS infusion. Partial of TUNEL staining were overlapped with NeuN staining. Angiotensin II infusion group had more TUNEL/NeuN doubled stained cells.

Fig 7. Level beam tests showed that PJ34 treatment attenuates behavior impairments caused by combined LPS and angiotensin II treatment in FA mice.

Level beam tests showed that FA mice treated with LPS combined angiotensin II infusion walked significantly slower than other groups on (A) 21mm, (B) 12mm and (C) 9mm level beams. PJ34 treatment attenuated the behavior impairments caused by LPS and angiotensin II combined treatment. After receiving PJ34 treatment, FA mice treated with LPS and angiotensin II infusion walked significantly faster than FA mice treated with LPS and angiotensin II infusion but not receiving PJ34 on the level beams. Data are expressed as mean± s.e.m. (t test or one way ANOVA, * p<0.05, ** p<0.01, *** p< 0.001, n = 5).

Fig 8. Treadscan tests showed that PJ34 treatment attenuates behavior impairments caused by combined LPS and angiotensin II treatment in FA mice.

Treadscan tests showed that FA mice treated with LPS combined angiotensin II infusion had significantly less normal stepsequence numbers than other groups (A). FA mice treated with LPS combined angiotensin II infusion had significantly smaller regularity index compared to other groups (B). LPS combined angiotensin II infusion on FA mice increased average stance time of front feet (C) and average swing time of rear feet (D) compared to untreated mice and LPS or ANG II only treated FA mice. PJ34 treatment reversed the effects of LPS combined with angiotensin II infusion. LPS injection only or angiotensin II infusion only didn’t cause any behavior deficit on either wild type mice or FA mice. Data are expressed as mean± s.e.m. (t test or one way ANOVA, * p<0.05, ** p<0.01, *** p< 0.001, n = 5).