Aldehyde accumulation and/or impaired aldehyde detoxification has been hypothesized to play a role in the pathogenesis of PD
[6],
[19]–
[22]. It was recently reported that patients who died with PD had an elevated DOPAL-to-dopamine ratio in caudate and putamen
[11]. Aldehydes, including DOPAL and 4-HNE, are highly reactive and can modify proteins
[6]–
[9]. Aldehydes are also cytotoxic and have been demonstrated to kill dopaminergic cells
in vitro and after intracranial injection
[20],
[21],
[23]–
[26]. A loss of function polymorphism in
ALDH2 has been reported to be a risk factor for Alzheimer's disease
[16],
[18],
[27]. Others have reported that
ALDH1A1 mRNA is reduced in substantia nigra dopaminergic neurons in brains of PD patients
[15] and
ALDH1A1 expression level in peripheral tissues has been reported as a candidate biomarker for PD diagnosis
[28]. These findings are strongly suggestive of a role for reduced aldehyde detoxification in the development of neurodegenerative diseases. However, until now, direct evidence for a causal relationship between impaired aldehyde detoxification and neurodegeneration in PD has been lacking. We hypothesized that chronically decreased function of multiple aldehyde dehydrogenases consequent to impaired complex I, and/or reduced
ALDH expression, plays a role in the pathophysiology of PD. To test this hypothesis, we generated a mouse model null for both
Aldh1a1 and
Aldh2, the two isoforms of aldehyde dehydrogenase that are known to be expressed in midbrain dopaminergic neurons
[13]–
[15]. The results show that mutations in these two aldehyde dehydrogenase isoforms resulted in neurochemical evidence for ALDH inhibition and elevated striatal tissue DOPAL, followed by aging-related development of a PD-like phenotype.
Mice deficient in
Aldh1a1 and
Aldh2 exhibited age-related deficits in tests of motor performance that require a high degree of motor coordination. We found progressive age-related impairment in rotarod performance in the
Aldh1a1−/−×Aldh2−/− mice as compared to wild-type controls. These effects were not due to alterations in muscle strength, body weight, motor learning, visual impairment or changes in circadian patterns of activity, factors that may confound the interpretation of rotarod performance
[29]. This impairment in rotarod performance is consistent with what others have observed in animal models of PD. Rotarod performance has been shown to be impaired in genetic
[30],
[31] and toxin-induced
[32] animal models of PD.
Altered gait pattern is symptomatic of PD
[33]–
[35]. A behavioral hallmark of PD is a shortened stride length, which is manifested as a shuffling gait
[34],
[35]. Alterations in gait have also been observed in animal models of PD and in models of other basal ganglia diseases, including Huntington's disease
[36]–
[38]. Conversely, others have reported that gait analysis did not detect alterations in the MPTP model of PD or in a model of amyotrophic lateral sclerosis
[39]. In the present study, gait analysis revealed asignificant 8.3% reduction of stride length in the oldest
Aldh1a1−/−×Aldh2−/− mice. A similar magnitude of reduction in stride-length was reported for mice treated with MPTP
[40]. That paper also reported a similar shorter stride duration and stride frequency
[40]. We found no differences in stride duration and frequency. On the other hand, others have reported no differences in stride length, but shorter stride duration and frequency in mice overexpressing human α-synuclein
[41]. Unlike MPTP-treated mice, those mice showed no loss of dopamine neurons. Thus, gait parameters may be differentially affected by the underlying pathology.
Responsiveness to L-DOPA treatment is diagnostic of PD
[42],
[43]. To test whether
Aldh1a1−/−×Aldh2−/− mice were responsive to L-DOPA treatment, we treated mice with L-DOPA plus benserazide, a peripheral aromatic amino acid decarboxylase inhibitor. The dose that we used (25 mg/kg body weight) was chosen based on calculations of the human equivalent dose (~2.03 mg/kg) as reported by FDA Draft Guidelines
[44],
[45]. This dosage is also one that is commonly reported for mice in the literature
[46],
[47]. L-DOPA treatment did not alter rotarod or gait performance of wild-type mice as compared to the baseline measurements. However, L-DOPA treatment alleviated the age-related deficits in stride length and rotarod performance in
Aldh1a1/Aldh2-deficient mice. Furthermore, deficits in stride length reappeared a week after L-DOPA treatment, a time when L-DOPA had been completely eliminated. Thus, with respect to responsiveness to L-DOPA, the
Aldh1a1/Aldh2-deficient mice model this aspect of Parkinson's disease. One might argue that L-DOPA administration would enhance aldehyde accumulation and toxicity and thus impair performance in
Aldh1a1/Aldh2-deficient mice. However, we feel that it is unlikely that a single L-DOPA injection would cause immediate and measureable cytotoxicity and/or augmentation of dopaminergic neurodegeneration.
Cognitive impairment is often found in PD
[48]. For example, patients with mild, early stage, Parkinson's disease are impaired in planning and spatial working memory tests
[49]. Therefore, we measured cognitive function by a Y-maze test, an assessment of short-term spatial working memory that is based on the innate tendency of rodents to explore novel environments. The advantage of the Y-maze test is that it is not as dependent on motor function as compared to other tests of memory, such as the Morris water maze
[50]. The Y-maze paradigm is also less stressful and less influenced by emotionality
[51]. In the present study, there were no differences due to genotype or age in total or novel arm entries, measures of locomotion and exploratory behavior respectively. However, there was a significant effect of age and genotype on the percentage of arm alternations, a measure ofspatial working memory. Thus, at the oldest age, the
Aldh1a1−/−×Aldh2−/− mice were significantly impaired as compared to wild-type mice. Deficiency in short-term spatial working memory, as measured in a T-maze, has been previously reported in mice treated with moderate doses of MPTP to mimic early stage PD
[52]. In that study, the dose used produced a 50 to 60% decrease in dopamine, but had no effect on locomotor activity in an open field test, similar to the levels of impairment in these parameters that we observed in our model in the present study. Similar effects of low dose MPTP on spatial working memory have been reported in monkeys
[53]. Thus, the effects of age and
Aldh1a1/
Aldh2-deficiency on spatial memory recapitulate the cognitive deficits seen in early stage Parkinson's disease.
We also measured monoamines and metabolites in the neostriatum. The serotonergic system has been reported to be affected in PD patients, although to a smaller extent than the dopaminergic system
[54]. Reductions in serotonin and its metabolite 5-HIAA in caudate nucleus and putamen of PD patients have been reported
[55],
[56]. Several studies have linked complications found in PD, such as fatigue, depression and dementia, to dysfunction of the serotonergic system
[1],
[57],
[58]. Consistent with these findings, the
Aldh1a1−/−×Aldh2−/− double knockout mice exhibited decreased striatal 5-HT (by 39%) and its metabolite 5-HIAA (by 38%), only at the oldest age.
Deficiency in
Aldh1a1 and
Aldh2 was also associated with significant decreases in dopamine and its metabolites DOPAC and HVA in middle aged and old mice. The deficits in dopamine and HVA content were only significant in the middle-aged and oldest group, respectively. In contrast, DOPAC was reduced by 48 to 67% of control at all ages and DOPAL was increased at all ages. The deficiency in DOPAC and the increase in DOPAL is consistent with the idea that Aldh1a1 and/or Aldh2 play a role in metabolism of DOPAL to DOPAC. We previously reported that deficiency in
Aldh2 was not associated with alterations in dopamine or metabolites
[59]. Recently, it was reported that in
Aldh1a1 deficient mice 2–3 months of age, there was no reduction in striatal dopamine levels and that DOPAC was slightly but significantly reduced by 20% as compared to wild-type control mice
[60]. The up to 67% reduction in DOPAC in the present study, as compared to the much smaller reduction in the
Aldh1a1-deficient mice reported previously
[60], supports the idea that more than one aldehyde dehydrogenase participates in metabolizing DOPAL to DOPAC. A recent clinical study reported that DOPAL-to-dopamine ratio was significantly elevated in the caudate and putamen of PD patients
[11]. The reduced ability to detoxify DOPAL by metabolizing it to DOPAC may contribute to the degeneration of dopamine neurons in PD. Thus, in the present study, elevated striatal DOPAL preceded evidence of loss of dopaminergic terminals.
In addition to the elevated DOPAL content, we observed increases in 4-HNE adducted proteins in the midbrain. Together, these results provide direct evidence that Aldh1a1 and Aldh2 are necessary to detoxify multiple biogenic aldehydes. Since each of these is known to be cytotoxic, this suggests that accumulation of multiple biogenic aldehydes may contribute to the neurodegeneration observed in PD. Thus, DOPAL and 4-HNE have been reported to induce death of cultured cells
[22],
[61]. Moreover, DOPAL has been reported to cause degeneration of dopaminergic neurons when injected intracranially
[20]. In addition, DOPAL and 4-HNE have been reported to accelerate oligomerization of α-synuclein and such oligomers may also be cytotoxic
[8],
[9].
A loss of TH-immunoreactive cells was observed in
Aldh1a1−/−×Aldh2−/− mice. A statistically significant reduction of up to 50% was observed in middle-aged and old
Aldh1a1−/−×Aldh2−/− mice as compared to their age-matched wild-type controls. The loss of TH-immunoreactive neurons is consistent with the loss of striatal dopamine. However, the loss of striatal dopamine was smaller in magnitude than the reduction in TH-immunoreactive neurons. It has long been reported that in studies of post-mortem human brain tissue, the degree of loss of substantia nigra dopamine neurons is greater than the loss of dopamine and metabolites in striatal regions, suggesting that surviving neurons have an increased capacity to make and store dopamine
[62]. There was no significant difference in the number of Nissl-stained cells, indicating that the reduction in TH
+ neurons was relatively specific. There was also no difference in the size of the TH
+ cells. There is no preclinical or clinical evidence showing a relationship between
Aldh2 deficiency and neurochemical and behavioral manifestations of PD. However, the loss of TH
+ cells in the
Aldh1a1−/−×Aldh2−/− mice is in contrast to a recent report that TH
+ neurons are increased in
Aldh1a1-deficient mice
[60]. Taken together, these findings suggest that deficiency in only one aldehyde dehydrogenase isoform is insufficient to cause neuropathology and that multiple redundant mechanisms for aldehyde detoxification participate in protecting dopamine neurons from the toxic effects of aldehydes.
In summary, deletion of two isoforms of aldehyde dehydrogenase, Aldh1a1 and Aldh2, which are known to be present in dopamine neurons, resulted in a Parkinsonian phenotype characterized by age-dependent deficits in motor performance, age-related reductions in monoamines and their metabolites and loss of neurons in the substantia nigra. These results provide the first direct support for the hypothesis that impaired aldehyde detoxification plays a role in PD. The Aldh1a1−/−×Aldh2−/− mouse may be a useful new model of early stage Parkinson's disease.
=
![[ratio]](/corehtml/pmc/pmcents/x2236.gif)
80 parts of 0.2 M acetic acid and phosphoric acid solution. After homogenization, samples were centrifuged and the supernatant was frozen and stored at −80°C freezer for storage until assayed for DOPAL by HPLC with electrochemical detection after batch alumina extraction. Briefly, mobile phase containing octanesulfonic acid as an ion pairing agent was pumped isocratically through a reverse phase liquid chromatographic column. Catechols were quantified by the current produced upon exposure of the eluate to flow-through electrodes set to oxidizing and then reducing potentials in series, with recording from the last electrode reflecting reversibly oxidized species.