Analysis of targeted mutation in DJ-1 on cellular function in primary astrocytes

Abstract

DJ-1 mutation induces early-onset Parkinson's disease, and conversely over-expression of DJ-1 is associated with cancer in numerous tissues. A gene-trap screening library conducted in embryonic stem cells was utilized for generation of a DJ-1 mutant mouse. Real-time PCR and immunoblotting were utilized to confirm functional mutation of the DJ-1 gene. Normal DJ-1 protein expression in adult mouse tissue was characterized and demonstrates high expression in brain tissue with wide systemic distribution. Primary astrocytes isolated from DJ-1(-/-) mice reveal a decreased nuclear localization of DJ-1 protein in response to rotenone or LPS, with a concomitant increase in mitochondrial localization of DJ-1 found only in the rotenone exposure. Resting mitochondrial membrane potential was significantly lower in DJ-1(-/-) astrocytes, as compared to controls. Our DJ-1 knockout mouse provides an exciting tool for exploring the molecular and physiological roles of DJ-1 to further explicate its functions in neurodegeneration.

Fig. 1

DJ-1 protein is widely distributed in the mouse. (A) A representative blot depicts relative expression of DJ-1 in 25 μg of protein from each respective tissue. (B) DJ-1 protein levels were assayed and are expressed relative to β-actin in each tissue (n ≥ 4), and demonstrate high expression in brain tissue. Bars with differing letters are significantly different from one another (p < 0.05).

Fig. 2

Gene-trap insertion into DJ-1 induces a functional mutation. (A) Schematic representation of the gene-trap in the endogenous DJ-1 gene. Following identification of the gene-trap insertion site, primers encompassing this sequence were designed, and, in combination with primers in the exogenous DNA, differentiation of DJ-1^+/+, DJ-1^+/−, or DJ-1^−/− animals was accomplished. (B) DJ-1 mRNA expression was measured via real-time PCR in whole brain tissue. DJ-1^−/− brain tissue has significantly decreased expression of DJ-1 compared to DJ-1^+/+ or DJ-1^+/− (n ≥ 4, p < 0.001). (C) DJ-1 protein expression is evident in DJ-1^+/+ and DJ-1^+/− animals, but absent in DJ-1^−/−.

Fig. 3

Rotenone increases mitochondrial localization of DJ-1, and both rotenone and LPS decrease nuclear localization. Primary cortical astrocytes were exposed to 50, 500 or 1000 nM rotenone + PBS, 10, 25, or 50 μg/mL LPS + DMSO, or control (PBS + DMSO) for 24 h, then loaded with MitoTracker Red and probed for DJ-1 immunofluorescence. To allow nuclear visualization, cells were mounted in DAPI-containing media. Pearson's correlation coefficients were calculated using SlideBook software image (A) A representative image used for analysis: (1) DAPI-stained nucleus, (2) DJ-1 immunofluorescence, (3) MitoTracker, and (4) merged image, for color image please see Supplementary Fig. 1. (B) DJ-1 correlation with the mitochondria (λ519:λ599) increased upon exposure to rotenone but not LPS (n ≥ 71, p < 0.001). (C) Nuclear correlation DJ-1 (λ519:λ460) decreased following exposure to either rotenone or LPS (n ≥ 71, p < 0.001).

Fig. 4

Mutation of DJ-1 decreases resting mitochondrial membrane potential. Astrocytes were loaded with the ratiometric mitochondrial dye JC-1 (5 μM) for 30 min, then rinsed once in MEM with 10 mM HEPES without phenol red, and images were collected every 2 min for 8 min. Mitochondrial membrane potential, expressed as the ratio of λ599:λ519 was measured using SlideBook image analysis software. (A) Representative image of the mitochondrial dye JC-1, (1) fluorescence emitted at λ519, (2) fluorescence emitted at λ599, and (3) merged images of 1 and 2, for color image please see Supplementary Fig. 2. (B) DJ-1^−/− astrocytes have decreased basal membrane potential compared to DJ-1^+/+ (n = 48, p = 0.02).