Behavioral and Neurotransmitter Abnormalities in Mice Deficient for Parkin, DJ-1 and Superoxide Dismutase

Parkinson’s disease (PD) is a progressive neurodegenerative disease characterized by loss of neurons in the substantia nigra that project to the striatum and release dopamine. The cause of PD remains uncertain, however, evidence implicates mitochondrial dysfunction and oxidative stress. Although most cases of PD are sporadic, 5-10% of cases are caused by inherited mutations. Loss-of-function mutations in Parkin and DJ-1 were the first to be linked to recessively inherited Parkinsonism. Surprisingly, mice bearing similar loss-of-function mutations in Parkin and DJ-1 do not show age-dependent loss of nigral dopaminergic neurons or depletion of dopamine in the striatum. Although the normal cellular functions of Parkin and DJ-1 are not fully understood, we hypothesized that loss-of-function mutations in Parkin and DJ-1 render cells more sensitive to mitochondrial dysfunction and oxidative stress. To test this hypothesis, we crossed mice deficient for Parkin and DJ-1 with mice deficient for the mitochondrial antioxidant protein Mn-superoxide dismutase (SOD2) or the cytosolic antioxidant protein Cu-Zn-superoxide dismutase (SOD1). Aged Parkin -/- DJ-1 -/- and Mn-superoxide dismutase triple deficient mice have enhanced performance on the rotorod behavior test. Cu/Zn-superoxide dismutase triple deficient mice have elevated levels of dopamine in the striatum in the absence of nigral cell loss. Our studies demonstrate that on a Parkin/DJ-1 null background, mice that are also deficient for major antioxidant proteins do not have progressive loss of dopaminergic neurons but have behavioral and striatal dopamine abnormalities.


Introduction
Parkinson's disease (PD) is the second most common neurodegenerative disorder after Alzheimer's disease and afflicts millions of people worldwide. The primary clinical symptoms are bradykinesia, resting tremor, rigidity, and postural instability. These symptoms are caused by the loss of dopaminergic innervation of the striatum and increase in severity over time due to selective, progressive nigral dopaminergic neuron loss. Most cases of PD are sporadic and the underlying cause of neuronal death remains unknown. The greatest risk factor for PD is age. About 5 to 10% of all cases are caused by inherited mutations [1][2][3][4][5][6]. Loss-of-function mutations in the Parkin and DJ-1 genes were the first mutations to be causally linked to recessive parkinsonism [7,8]. Both genes are widely expressed throughout the brain and other tissues [9][10][11][12][13]. The mechanism by which loss of Parkin or DJ-1 function causes parkinsonism remains unclear.
Together, these data suggest that oxidative damage may be an important factor in the development of PD caused by Parkin and DJ-1 mutations. Oxidative stress has been implicated as a potential cause of idiopathic PD because postmortem examinations of PD patients show increased oxidative damage in neurons [54][55][56][57]. Furthermore, the capacity of cells to clear reactive oxygen species and repair oxidative damage to proteins, lipids and nucleic acids diminishes with age [58].
Two superoxide dismutase proteins, cytoplasmic Cu/Znsuperoxide dismutase (SOD1) and mitochondrial Mnsuperoxide dismutase (SOD2) are the among the most abundant antioxidant proteins in the brain and are important for protecting neurons from oxidative stress. The superoxide dismutase proteins remove toxic superoxide by catalytically converting it to oxygen and hydrogen peroxide. Some studies have suggested that abnormalities in SOD1 or SOD2 may contribute to the development of PD [59][60][61], although no mutations in SOD1 or SOD2 have been causally linked to PD. In flies, expression of human SOD1 is protective against neuronal loss due to inactivation of PINK1, another gene linked to recessive parkinsonism [62]. In mice, overexpression of SOD2 is protective against nigral dopamine neuron loss induced by the neurotoxin 6-hydroxydopamine while partial SOD2-deficiency increases 1-methyl-4-phenyl-1,2,3,6tetrahydropyridine (MPTP) sensitivity [63,64]. Levels of SOD1 mRNA and SOD1 activity are significantly reduced in PD patients [65][66][67]. Overexpression of SOD1 or SOD2 in vitro and in vivo is protective against MPTP, methamphetamine and 6hydroxydopamine toxicity [64,66,[68][69][70][71]. MPTP exposure increases the protein levels of both SOD1 and SOD2 [72], suggesting that these proteins are important for mitigating oxidative stress in response to toxins. Together, these studies indicate that SOD1 and SOD2 protein or activity levels may be key determinants of susceptibility to nigral cell loss in mice.
Aged DJ-1 -/-mice have increased SOD2 in brain mitochondria, suggesting that up-regulation of antioxidant proteins may prevent PD-linked loss-of-function mutations from causing nigral cell loss in mice [14]. We hypothesized that eliminating potential compensatory upregulation of superoxide dismutase activity would result in PD-related neurodegeneration in Parkin and DJ-1 knockout mice. To test this hypothesis and to potentially generate better PD animal models, we crossed mice deficient for Parkin and DJ-1 with mice deficient for SOD1 to generate triple mutant mice deficient for all three genes (Parkin -/-DJ-1 -/-SOD1 -/-). We also generated Parkin -/-DJ-1 -/-SOD2 +/-mice by crossing Parkin -/-DJ-1 -/-mice with heterozygous SOD2 knockout mice because homozygous SOD2 deficiency is perinatal lethal [73] while heterozygous SOD2 knockout mice have approximately 50% decreased SOD2 activity and increased oxidative damage [74,75]. The triple mutant mice did not exhibit age-dependent nigral neuron loss, dopamine depletion, or motor behaviors characteristic of PD. Contrary to our hypothesis, Parkin -/-DJ-1 -/-SOD1 -/-triple mutant mice had increased striatal dopamine levels and Parkin -/-DJ-1 -/-mice showed improved rotarod performance. These results are consistent with our previous study showing improved rotarod performance in Parkin -/-DJ-1 -/-mice and increased striatal dopamine in Parkin -/-DJ-1 -/-Gpx1 -/-mice [76] and consistent with the increased striatal dopamine observed in aged Parkin -/-DJ-1 -/-PINK1 -/-mice [23]. This study demonstrates that superoxide dismutase is not critical for preventing nigral cell loss caused by Parkin and DJ-1 deficiencies. The increase in striatal dopamine in Parkin -/-DJ-1 -/-SOD1 -/-mice indicates functional abnormalities within the nigrostriatal pathway. These behavioral and neurotransmitter abnormalities suggest that the pathogenic mechanisms of Parkin and DJ-1 mutations may involve dysregulation of nigrostriatal dopaminergic neurotransmission.

Ethics Statement
Experimental procedures involving the use of animals or animal tissue were performed in accordance with the NIH Guidelines for Animal Care and Use and approved by the Institutional Animal Care and Use Committee at The University of Texas Southwestern Medical Center.

Animals
Parkin knockout mice and DJ-1 knockout mice were generated as previously described [19,20] and backcrossed to strain C57BL/6J for 10 generations, then intercrossed for two generations to obtain homozygous double knockout mice (Parkin -/-DJ-1 -/-) and wild type controls. SOD1 and SOD2 knockout mice on a C57BL/6 background were kindly provided by Dr. Holly VanRemmen with permission from Dr. Ting-Ting Huang [73,77]. SOD1 and SOD2 knockout mice were crossed with Parkin, DJ-1 double knockout mice for two generations to produce Parkin -/-DJ-1 -/-SOD1 +/-, Parkin -/-DJ-1 -/-SOD1 -/-and Parkin -/-DJ-1 -/-SOD2 +/-mice. When possible, paired littermates were used as controls; when not possible, wild type mice from the same line to which the mutant mice were backcrossed were used as controls (C57BL/6). Animals were housed in a climatecontrolled facility with ventilated cages and standard commercial lab diet. Behavioral tests were performed between 10 AM and 4 PM during the 6 AM to 6 PM light cycle.
beam breaks was recorded in 5-minute bins as a measure of locomotor activity.
Open Field Activity. Mice were placed in the periphery of a novel open field environment (44 cm x 44 cm, walls 30 cm high) and allowed to explore for five minutes. The animals were monitored from above by a video camera connected to a computer running video tracking software (Ethovision 3.0, Noldus, Leesburg, Virginia) to determine the time, distance moved and number of entries into two areas: the periphery (5 cm from the walls) and the center (14 cm x 14 cm). The open field arenas were wiped and allowed to dry between mice.
Elevated Plus Maze. Mice were placed in the center of a black Plexiglas elevated plus maze (each arm 30 cm long and 5 cm wide with two opposite arms closed by 25 cm high walls) elevated 31 cm from the floor and allowed to explore for five minutes. The animals were monitored from above by a video camera connected to a computer running video tracking software (Ethovision 3.0, Noldus, Leesburg, Virginia) to determine time spent in the open and closed arms, time spent in the middle, and the number of entries into the open and closed arm. The apparatus was wiped and allowed to dry between mice.
Dark-Light Activity.
Mice were placed into a black Plexiglas chamber (25 cm x 26 cm) and allowed to explore for two minutes. After the habituation period, a small door was opened allowing them to access the light side of the apparatus (25 cm x 26 cm lit to approximately 1700 lux) for ten minutes. The animals were monitored by seven photobeams in the dark compartment and eight photobeams on the light side connected to a computer which recorded the time spent in each compartment, latency to enter the light side and the number of entrances to each compartment (Med-PC IV, Med Associates, St. Albans, VT). The dark-light apparatus was wiped and allowed to dry between mice.
Acoustic Startle Response. Acoustic startle response behavior was measured using a San Diego Instruments SR-Lab Startle Response System (San Diego, CA). Mice were placed into Plexiglas holders and allowed to acclimate to the chamber and background white noise (70 dB) for five minutes. After acclimation, startle stimuli (80, 90, 100, 110 and 120 dB, 40 ms, white noise) were presented in pseudorandom order with an average interstimulus interval of 20 seconds (range 13 -27 seconds). Each startle response was computer recorded by force transducers. The Plexiglas holders were wiped and allowed to dry between mice.
Forced Swim Test. Mice were placed in a 16.5 cm diameter beaker of water (21-25° C) filled to a depth of seven inches. The mice remained in the water for 6 minutes and were then removed and allowed to dry in a clean dry cage before returning to their home cage. The water was changed between each subject. The mice were monitored from the side by camera and video images were recorded for later analysis. The last four minutes of the test were scored for latency to the first immobility and total time spent immobile. The experimenter scoring the behavior was blind to the genotypes. Immobility was defined as no body or limb movement other than a minimal forelimb movement required for keeping the head above water.
Grip strength. Mice were suspended by the tail and allowed to clasp a grip bar with both forelimbs. Gentle constant horizontal resistance was applied until the mouse released the bar. Force was measured in grams by a force transducer attached to the bar. The greatest three of five trials per mouse were averaged.

Stereology of Dopaminergic Neurons
Brains were removed and placed in 10% neutral buffered formalin at 4° overnight, processed for paraffin embedding and sectioned in the coronal plane at 20 micron thickness. Every fifth slide was stained and used for unbiased stereology. Slides were deparaffinized and rehydrated and blocked with 5% normal goat serum for 1 hour prior to incubation in primary antibody (anti-tyrosine hydroxylase AB152, Chemicon) diluted 1:1000 at 4°C overnight. Sections were washed, incubated with biotinylated goat anti-rabbit secondary antibody, horseradish peroxidase conjugated avidin (ABC Elite, Vector) and were developed in DAB solution with NiCl enhancement prior to dehydration and coverslipping. A microscope with a motorized stage and Stereoinvestigator software was used to count the tyrosine hydroxylase positive neurons in the substantia nigra of each section, with a counting frame of 50 microns by 50 microns and a grid size of 100 microns by 100 microns. The total number of bilateral substantia nigra dopamine neurons was estimated by counting neurons from one hemisphere.

Measurement of tissue neurotransmitter levels
Mice were euthanized and the striatum was quickly dissected on an ice-cold glass dish, weighed, frozen on dry ice and stored at -80 prior to analysis. Samples were combined with 50-fold (weight:volume) ice-cold 0.1 N perchloric acid containing 0.2 mM sodium metabisulfite. The tissue was disrupted by brief sonication and centrifuged at 4°C for 20 minutes at 15,000 x g to pellet proteins and cell debris. 200 µL of the supernatant was transferred to a clean tube, and 20 µL was injected onto an HPLC with a C18 column and eluted with isocratic MDTM mobile phase (ESA) at a rate of 0.6 mL/min. Monoamines were detected with a model 5014B electrochemical cell (ESA) set to a potential of +220 mV. Peak areas were normalized to tissue weight and compared to external standards for quantification.

Tritiated Dopamine Ligand Binding to D1 and D2-like Receptors in the Striatum
The rostral half of unfixed mouse brains were sectioned in the coronal plane on a cryostat at a thickness of 20 microns. Slides were preincubated in assay buffer (50 mM Tris, 120 mM sodium chloride, 5 mM potassium chloride, 1 mM magnesium chloride, 40 nM ketanserin) for 20 min at room temperature. Slides were then incubated in buffer containing either 2 nM tritiated SCH 23390 (PerkinElmer, Boston, MA) or 5 nM tritiated spiperone (PerkinElmer, Boston, MA). Cold competition of tritiated siperone with 10μm cold siperone, or of tritiated SCH 23390 with 1μm cold SCH 23390 was used to assess nonspecific signal. Following a 1 hour incubation, slides were washed twice in ice-cold buffer and then rinsed in cold water prior to incubation overnight at 4C in a desiccator with paraformaldehyde powder in the bottom to fix the tissue without washing away the ligand. Slides were dried for 2 hours in a desiccator with Dri-Rite and exposed to Kodak BioMax MS film with the Kodak BioMax TranScreen-LE Intensifying Screen for seven days (D1) or 5-7 weeks (D2). Films were analyzed for both density and area of binding using Adobe Photoshop software. Measurements from twenty sections per animal were averaged prior to statistical analysis 4 mice per genotype were included in the analysis.

Hematoxylin and eosin (H&E) staining
Muscle tissue was taken from the extensor digitorum longus (EDL) and soleus (SOL) muscles and frozen sections were incubated 45 minutes in hematoxylin stain followed by a water rinse, then a 0.3% hydrochloric acid in ethanol rinse. Slides were incubated in eosin for 2 minutes, dehydrated and mounted for imaging with a light microscope, as above.

Muscle fiber typing
Extensor digitorum longus (EDL) and soleus (SOL) muscles were dissected from wild type and Parkin -/-DJ-1 -/-mice and flash frozen. Muscle were cut into thin sections and maintained at -20°C. Sections were stained with Metachromatic ATPase as previously published (Ogilvie 1990). Briefly, sections were preincubated with ATPase 8 minutes (pH 4.5), rinsed twice for 3 minutes in Tris buffer (pH 7.8), then incubated with ATP (pH9.4) for 25 minutes at room temperature. Slides were rinsed 3 times with calcium chloride and counterstained with Toluene blue 0.1% for 1 minute, cleared in ethanol, then xylene, and mounted for imaging with a light microscope, as above.

Statistical Analysis
All analysis was performed using SigmaPlot. All experiments were analyzed by ANOVA to account for multiple comparisons. Two way repeated measures ANOVA was performed where appropriate to account for multiple samplings over time. Pairwise and posthoc comparisons were performed by Tukey test, unless otherwise noted. Student's t-test was used in lieu of ANOVA where appropriate. Results are presented as mean ±standard error of the mean.

No neuronal loss in the substantia nigra of mutant mice
Triple knockout mice bearing combined loss-of-function mutations in the PD-linked genes Parkin and DJ-1, as well as the antioxidants SOD1 or SOD2, were born at the expected Mendelian ratio and had no apparent differences in viability compared to wild type mice. Because the pathological hallmark of PD is the progressive loss of dopamine neurons in the substantia nigra and previously published data indicate that Parkin -/-and DJ-1 -/-mice do not have changes in numbers of nigral dopaminergic neurons [14,16,[18][19][20][21][22][23][24]28,33], we postulated that mice with a deficiency for major antioxidant proteins in addition to Parkin and DJ-1 would show an agedependent loss of dopamine neurons in the substantia nigra and thereby model human PD neuropathology. We used rigorous stereology to obtain unbiased estimates of the number of dopaminergic neurons, marked by immunohistochemical staining for tyrosine hydroxylase (TH), in coronal paraffin sections of wild type and mutant mice. We analyzed cohorts of mice at ages 7, 16 and 18 months. There was no statistically significant difference between the wild type mice and any of the mutant mice at any age by ANOVA (Figure 1), indicating no change in the number of nigral dopaminergic neurons in Parkin -/-DJ-1 -/-SOD1 -/-mice or Parkin -/-DJ-1 -/-SOD2 +/-mice compared to wild type mice.

Dopamine levels are elevated in the striatum of triple mutant mice
Although we observed no change in dopaminergic nigral neurons, loss of dopaminergic terminals in the striatum may precede loss of cell bodies in the substantia nigra. Therefore, we used HPLC with electrochemical detection to measure levels of dopamine, serotonin and their metabolites in the striatum of wild type and mutant mice. Our previous studies showed no changes in the level of striatal dopamine in Parkin -/mice or DJ-1 -/-mice compared to wild type mice [19,20]. Remarkably, we found a consistent and statistically significant increase in dopamine in the striatum of Parkin -/-DJ-1 -/-SOD1 -/mice, but not in Parkin -/-DJ-1 -/-SOD2 +/-mice, compared to wild type mice ( Figure 2) and no differences in the levels of dopamine or serotonin between wild type and Parkin -/-DJ-1 -/mice (Table 1).
Separately, we examined TH immunoreactivity in the striatum of wild type, Parkin -/-DJ-1 -/-, Parkin -/-DJ-1 -/-SOD1 -/-and Parkin -/-DJ-1 -/-SOD2 +/-mice to see whether the increased striatal dopamine was accompanied by increased TH levels. Images of striata stained with DAB to measure THimmunoreactivity were analyzed by densitometry using Image J software. In agreement with previous data [28], we found no differences in the intensity of TH immunostaining between wild type, Parkin -/-DJ-1 -/-, Parkin -/-DJ-1 -/-SOD1 -/-or Parkin -/-DJ-1 -/-SOD2 +/-mice (p>0.5) (data not shown).  The significant increase in dopamine levels in triple mutant mice prompted us to measure levels of dopamine receptors in the striatum. We expected to see a compensatory change in dopamine receptor density in response to elevated dopamine levels in Parkin -/-DJ-1 -/-SOD1 -/-mice compared to wild type mice. Using radiolabeled dopamine receptor ligands, we quantified the abundance of D1-like and D2-like dopamine receptors in the striatum in fresh frozen sections. Contrary to the expected results, we saw no change in the binding of radiolabeled ligands to D1-or D2-type dopamine receptors (p > 0.4) (Figure 3).

Locomotor behavior is unchanged in mutant mice
Because Parkinson's disease causes deficits in motor function, we investigated whether mice with combined PDlinked mutations have altered performance in established behavioral tests of motor function. The locomotor test measures total activity, and is used to identify mice that are hypoactive or hyperactive compared to wild type mice. Previous reports have indicated normal locomotor activity in Parkin -/-and DJ-1 -/-mice [19,33,78]. We hypothesized that mice deficient for Parkin, DJ-1 and antioxidants would have an agedependent locomotor deficit, manifested as reduced ambulatory behavior compared to age matched controls. The activity of all experimental groups was affected by time in the apparatus, indicating that all mice acclimated to the novel environment over the course of two hours, as indicated by a gradual decrease in the number of beam breaks. Seven-month old Parkin -/-DJ-1 -/-, Parkin -/-DJ-1 -/-SOD2 +/-, and Parkin -/-DJ-1 -/-SOD1 -/-mice exhibited no locomotor deficits compared to wild type (p > 0.126) ( Figure 4A).

Rotarod behavior is altered in mice with combined mutations linked to PD
The rotorod task is commonly used to measure the ability of a rodent to sustain complex coordinated movement over time and has been used as a measure of basal ganglia function and as a functional measure of neurodegeneration in mouse models of amyotrophic lateral sclerosis (ALS), Huntington's disease (HD) and other neurodegenerative diseases [79]. No abnormal rotorod phenotypes have been reported for Parkin -/-, DJ-1 -/-, SOD1 -/-, or SOD2 +/-mice, including results from our previously published studies [15,19,20,30,31,76,80,81]. We assessed the rotorod performance of wild type, Parkin -/-DJ-1 -/-, SOD1 -/-, SOD2 +/-, Parkin -/-DJ-1 -/-SOD1 -/-, and Parkin -/-DJ-1 -/-SOD2 +/-at various ages ( Figure 5). As expected, there was a main effect of trial for each experiment, indicating that there is a learning process associated with the task and all groups improve over the course of the 8 trials. There was also a main effect of genotype in each experiment and the Parkin -/-DJ-1 -/mice had an increased latency to fall instead of the expected decreased latency, compared to wild type mice. Although puzzling, this result confirms data that we previously reported on separate cohorts of Parkin -/-DJ-1 -/-mice [76].
Rotarod performance is typically considered to be a measure of motor ability but other factors, such as weight, have also been reported to affect latency to fall [82]. We analyzed whether body weight of the mice in this study has an effect on rotarod performance by ANCOVA. There was no effect of body weight on latency to fall from the rotarod (p > 0.05). The increase in rotorod performance of Parkin -/-DJ-1 -/-mice prompted us to determine whether these mice have increased ability to grip the rotating rod, thus facilitating their ability to remain on the rod for longer periods of time. We analyzed the grip strength of a cohort of 6-month old wild type, Parkin -/-DJ-1 -/-, Parkin -/-DJ-1 -/-SOD1 -/-, and Parkin -/-DJ-1 -/-SOD2 +/-mice and found no difference between genotypes (p=0.352) ( Figure  6A). Therefore the improved rotorod performance of the double and triple mutant mice was not due to increased grip strength.
Other characteristics that could underlie the striking differences observed in rotarod behavior are differences in reflexes or sensory functions. This prompted us to measure startle response behavior in the mice using the acoustic startle test, which measures the innate reflex of the mouse in response to a sudden and unexpected acute stimulus, such as a loud noise. Analysis of a cohort of 7 month old wild type, Parkin -/-DJ-1 -/-, Parkin -/-DJ-1 -/-SOD1 -/-, and Parkin -/-DJ-1 -/-SOD2 +/mice, showed a significant decrease in startle response in Parkin -/-DJ-1 -/-and Parkin -/-DJ-1 -/-SOD1 -/-(p < 0.001) compared to wild type ( Figure 6B). There was also a trend toward reduced startle response in Parkin -/-DJ-1 -/-SOD2 +/-mice compared to wild type mice, but this was not significant, perhaps due to the smaller sample size of this genotype. Decreased startle responses in Parkin -/-DJ-1 -/-and Parkin -/-DJ-1 -/-SOD1 -/-indicate impairment of reflexes and/or sensory functions compared to wild type mice. However, this would not explain the observed increased latency to fall in these genotypes.
Because decreased startle amplitudes could indicate decreased anxiety levels and because fear of falling off the rotarod apparatus could contribute to rotarod behavior performance, we tested wild type and mutant mice for changes  in fear or anxiety using several established behavioral paradigms including the light/dark test, the open field test and the elevated plus maze. We did not observe any difference between genotypes in any of these behavior tests (data not shown). Therefore, the increased rotorod performance of the double and triple mutant mice does not appear to be due to altered anxiety. Finally, we conducted the forced swim test as part of our comprehensive battery of behavioral tests in an effort to better characterize these mice and possibly explain the altered  rotarod performance. The forced swim test (FST) is a test of forced motor behavior developed to predict the efficacy of antidepressant drugs and is particularly sensitive to compounds acting on the serotonin and norepinephrine systems. Compared to wild type, we found significantly reduced immobility time in 6-, 14-and 16-month old Parkin -/-DJ-1 -/-mice (p = 0.005), 6-month old Parkin -/-DJ-1 -/-SOD2 +/mice (p = 0.027) and 16-month old DJ-1 -/-mice (p < 0.001) ( Figure 7A-C). Post hoc analysis of a second 14 month old cohort of wild type, Parkin -/-, and Parkin -/-DJ-1 -/-demonstrated a significant decrease in immobility time Parkin -/-DJ-1 -/-mice compared to wild type (p < 0.001). Together these data indicate that the loss of DJ-1 alone is sufficient to confer the increased mobility observed in the forced swim test.

Normal muscle, heart, and cerebellar histology in Parkin -/-DJ-1 -/-mice
Parkin is highly expressed in the heart and muscle tissue, as well as most areas of the brain [7]. Additionally, the ability to stay on the rotorod is highly dependent on cerebellar function. Therefore, histology of muscle, heart and cerebellum was performed in order to test for abnormalities that might account for enhanced rotarod performance in the Parkin -/-DJ-1 -/-mice. Hematoxylin and eosin staining of paraffin sections of hearts from 18 month old wild type and Parkin -/-DJ-1 -/-mice showed normal organ structure as well as normal cellular appearance of heart muscle in mice of both genotypes (Figure 8). Hematoxylin and eosin staining of coronal brain sections showed normal anatomical and cellular organization of the cerebellum (Figure 8). We hypothesized we would see a difference in the ratio of fast twitch to slow twitch muscle fiber types, providing the Parkin -/-DJ-1 -/-mice fatigue-resistant muscles and better stamina on the rotorod. The ratio and morphology of type I and type II muscle fiber types appeared similar in metachromatic ATPase stained soleus muscle from Parkin -/-DJ-1 -/-mice compared to wild type mice ( Figure 8).

Discussion
Despite years of effort by many researchers generating and characterizing mice with recessive mutations in Parkin and DJ-1 linked to familial PD, no Parkin or DJ-1 knockout mouse line exhibits all, or even most, of the key clinical and neuropathological features of PD [14,16,17,[19][20][21][22][23][24][25][26][27]30,31]. This contrasts with mouse models of other neurodegenerative diseases such as Huntington's disease and amyotrophic lateral sclerosis that reproduce the major neuropathological and behavioral features of these diseases, including delayed onset of motor behavior deficits as well as selective and progressive neurodegeneration and protein inclusions [83,84]. It is possible that compensatory changes protect mice from the neurodegeneration caused by Parkin and DJ-1 mutations or that additional stress is required to initiate neurodegeneration in Parkin and DJ-1 knockout mice. Indeed, we have shown that Parkin knockout mice are more susceptible to nigral cell loss induced by chronic lipopolysaccharide exposure [18] and others have shown that DJ-1 knockout mice are more susceptible to MPTP-induced nigral cell loss [22]. Here we sought to generate better mouse models of progressive nigral cell loss induced only by genetic mutations. We combined PDlinked mutations in Parkin and DJ-1 with deficiency for antioxidant enzymes that have been shown to be important determinants of susceptibility to nigrostriatal degeneration in mice [59][60][61][62][63][64][66][67][68][69][70][71][72]85,86].
The increased striatal dopamine did not cause hyperactivity or other apparent behavioral abnormalities specific to Parkin -/-DJ-1 -/-SOD1 -/-mice. Dopamine levels are tightly regulated by presynaptic autoreceptors that signal the regulation of tyrosine hydroxylase to alter dopamine synthesis. Therefore, we speculate that the increased striatal dopamine levels indicate that Parkin -/-DJ-1 -/-SOD1 -/-mice either have dopamine autoreceptor signaling defects or that the increased dopamine compensates for other defects in dopaminergic signaling. It is possible that, in the absence of the compensatory increase in striatal dopamine, we would have detected behavioral phenotypes specifically in Parkin -/-DJ-1 -/-SOD1 -/-mice. Alternatively, the increase in dopamine may be indicative of a compensatory change that is occurring in the striatum that might ultimately lead to neuronal death, either due to overwhelming the compensatory mechanisms of the cell or due to neurotoxicity of dopamine itself [93,94]. The increased striatal dopamine also could be due to interactions between the three genes to affect release, trafficking, dopamine production, or degradation. Consistent with this possibility, Parkin and DJ-1 have been found in cell membranes and vesicles [95,96] and mutations in Parkin and DJ-1 can cause altered dopamine release, reuptake and signaling [20,97,98].
The increased striatal dopamine levels cannot explain the rotarod and forced swim test behavioral phenotypes because we did not observe an increase in striatal dopamine in the DJ-1 single knockout mice (which show the forced swim test phenotype) or Parkin-DJ-1 double knockout mice (which show the improved rotarod performance phenotype). Therefore, the behavioral and the neurochemical phenotypes observed in these mice are likely independent of each other. Mice deficient for both Parkin and DJ-1 consistently stayed on the rotarod longer than wild type mice. This finding holds true in both the young and aged cohorts of mice, although the phenotype is more pronounced in aged mice ( Figure 6). The rotarod behavioral phenotype is consistent with independent cohorts of Parkin -/-DJ-1 -/-mice that we studied in combination with glutathione peroxidase deficiency [76], which suggested that the increased rotarod latency of Parkin -/-DJ-1 -/-mice may be due to decreased distraction from the task compared to wild type mice.
The decreased startle response in Parkin -/-DJ-1 -/-mice might be due to Parkin deficiency alone because Parkin -/-mice have previously been shown to have decreased startle response [31]. The greater decrease in startle response in Parkin -/-DJ-1 -/-SOD1 -/-mice is likely due to a loss of hearing in mice deficient for SOD1. The cochleae of SOD1 -/-mice have been shown to have severe spiral ganglion cell degeneration by 7 months of age [99].
It remains possible that abnormalities in antioxidants such as SOD1 or SOD2 are involved in the development or progression of PD. However, our study demonstrates that deficiency for these antioxidants does not directly cause nigral cell loss or parkinsonian-like locomotor deficits in the context of mice with PD-linked mutations in Parkin and DJ-1. The increase in striatal dopamine resulting from the combination of SOD1, Parkin and DJ-1 deficiency might represent early-stage nigrostriatal dysfunction and these mice may be useful for efforts to develop neuroprotective therapies targeting early stage abnormalities in PD pathogenesis.