Inverse correlation between brain dopamine loss in Parkinson disease (PD) and tissue noradrenaline levels. A, Noradrenaline levels in subdivisions of the nucleus accumbens (NACS) of control subjects and the percentage dopamine loss in the corresponding accumbens subdivisions of patients with PD. Data are mean ± SEM. *P<.001, †P=.007, and ‡P=.07, PD vs control in dopamine levels (2-tailed t tests with Bonferroni adjustment). c Indicates caudal; i, intermediate; and r, rostral. Note the high concentration of noradrenaline and the low extent of dopamine loss in the medial portion of the caudal accumbens (solid bar) in PD. B, Correlation (Pearson product moment correlation) between dopamine loss in PD and levels of noradrenaline in control subjects in the 20 brain areas as given in Table 2. The solid circles indicate subdivisions of the striatum (caudate, putamen, and NACS). Noradrenaline levels in the NACS of the control subjects have been published11 and are reproduced with permission from the International Society for Neurochemistry.
Tong J, Hornykiewicz O, Kish SJ. Inverse Relationship Between Brain Noradrenaline Level and Dopamine Loss in Parkinson DiseaseA Possible Neuroprotective Role for Noradrenaline. Arch Neurol. 2006;63(12):1724-1728. doi:10.1001/archneur.63.12.1724
Experimental findings using animal models of Parkinson disease (PD) suggest that noradrenaline might protect dopamine neurons from damage.
To assess whether human brain regions having high levels of noradrenaline are less susceptible to dopamine loss in PD.
Autopsied brains of patients with PD and of healthy control subjects.
Main Outcomes Measures
We compared the extent of dopamine loss in different regions relative to levels of noradrenaline found in healthy brain, with special attention devoted to the dopamine-rich nucleus accumbens, which has noradrenaline-rich and noradrenaline-poor subdivisions.
Among 20 brain areas, dopamine loss in PD was negatively correlated with healthy noradrenaline levels (r = 0.83), with regions rich in noradrenaline (eg, the noradrenaline-rich portion of the nucleus accumbens) spared from dopamine loss. However, within the striatum, noradrenaline levels in the caudate and putamen were similar, despite dopamine's being more markedly reduced in the putamen.
Our postmortem data are consistent with animal findings suggesting that noradrenaline might affect dopamine neuron loss in PD and that a noradrenergic approach (although not aimed at the as yet unknown primary cause of PD) could be neuroprotective. This possibility should also be considered when noradrenergic therapy is provided for symptomatic purposes in PD.
Idiopathic Parkinson disease (PD) is characterized by heterogeneous degeneration of brain dopamine neurons.1 The cause of PD is unknown, and the reason why dopamine neurons in some brain areas are much more vulnerable to damage than those in others is unknown.
Neuropathological observations in PD of the loss of noradrenergic neurons originating in the locus coeruleus area2 suggest that noradrenaline therapy might be helpful in symptomatic treatment of some aspects of PD.3 Moreover, animal findings suggest that the loss of brain noradrenergic neurons in PD might exacerbate dopamine neuron damage and that noradrenaline could actually be neuroprotective.4 The possible involvement of noradrenaline is especially interesting given findings of brain microglial activation in PD5- 8 and the emerging role for noradrenaline as an endogenous anti-inflammatory agent.9,10
If the experimental animal data are relevant to the human, one might expect that brain regions having high levels of noradrenaline could be more protected against dopamine loss in PD. To address this possibility, we compared the extent of dopamine loss in different regions in autopsied brains of patients with PD with levels of noradrenaline found in healthy brain. Of 20 areas examined, an area of focus was the dopamine-rich nucleus accumbens (NACS), a striatal subdivision that contains a noradrenaline-poor rostral portion and a caudomedial subdivision that is strikingly enriched in noradrenaline.11
Autopsied brains were obtained (February 1982 to December 1990) from patients with PD (n = 10) and from healthy control subjects (n = 11). The control subjects and the patients with PD did not differ significantly in their mean (SD) ages (70.6 [9.9] and 75.8 [7.4] years, respectively; P = .20 or in their mean postmortem intervals before autopsy (9.3 [4.2] and 12.8 [5.9] hours, respectively; P = .14) (2-tailed t tests). Half of the brain was used for neuropathological examination, whereas the other half was frozen for neurochemical analyses. Clinical, drug history, and brain neuropathological findings for the patients with PD are summarized in Table 1. All patients had received the clinical diagnosis of PD except for patient 8, who had been observed only briefly by a neurologist a few days before death and was considered at that time to have a neurological illness characterized by tremor. Postmortem brain neurochemical analyses of this patient showed the pattern of striatal dopamine loss that is characteristic of PD, and brain neuropathological analyses of this patient and all other patients with the clinical diagnosis of PD disclosed the characteristic histopathological signs of substantia nigra cell loss and the presence of Lewy bodies. All patients except for patient 8 had received dopamine substitution medication. No detailed information on the neuropsychological or mental function of the patients was available. Control subjects had died without evidence of neurological or psychiatric disease and showed no brain abnormality on neuropathological examination. The causes of death for the control subjects were myocardial infarction (n = 3), cardiac failure (n = 2), cancer (n = 2), pulmonary embolism (n = 2), diffuse interstitial pulmonary disease (n = 1), and unknown (n = 1).
Dissection of the brains and subdivisions of the striatum, including the NACS, followed procedures that have been published elsewhere.11 Levels of dopamine and noradrenaline were measured by high-performance liquid chromatography–electrochemical detection.12 The primary outcome measure was correlation (Pearson product-moment correlation) among the brain areas examined between percentage dopamine loss in PD vs noradrenaline levels in healthy brain. Two-tailed t tests with Bonferroni correction for multiple comparisons were performed to examine differences between the control and PD groups.
Some neurochemical data have been previously published. These include caudate and putamen dopamine levels in 10 control subjects and in 8 patients with PD13; hypothalamus dopamine and noradrenaline levels in 8 control subjects and in 9 patients with PD14; and NACS dopamine and noradrenaline levels in 11 control subjects.11 No significant correlation was observed between age or postmortem interval of the subjects and tissue levels of the monoamines in the control or PD group.
Table 2 gives the control and patient levels of dopamine and noradrenaline in 1 cerebral and 19 subcortical cortical brain areas. Analysis in the cerebral cortex was limited to the cortical region, Brodmann area 25 (parolfactory cortex), which in the human contains a quantifiable amount of dopamine.
As expected, dopamine was decreased in all examined brain regions, but the magnitude of the reduction in the different brain areas was variable. Of the regions examined, the caudal portion of the putamen was the most affected (−97%) in PD, and the preoptic area of the hypothalamus was the least affected (−9%) (Table 2). Concentrations of noradrenaline were generally more modestly decreased (≤57% maximum reduction). No significant correlation (r = −0.17, P>.05) was observed between the extent of dopamine and noradrenaline loss in the 20 examined brain regions.
In the PD group, brain areas that in healthy subjects have high (>1 ng/mg) noradrenaline concentrations (the caudomedial NACS and rostral portions of the hypothalamus) showed less severe dopamine loss (−9% to −54%) compared with brain areas with low levels (<0.4 ng/mg) of the neurotransmitter (−61% to −97%) (Table 2). This is best exemplified by the NACS and the striatal nuclei. In the NACS, the noradrenaline-rich caudomedial portion was spared (relative to the other NACS subdivisions) from severe dopamine depletion (Figure, A), and in the noradrenaline-poor striatum, the dopamine loss was severe in the caudate and in the putamen. A correlational analysis using values from the 20 examined brain areas confirmed that the extent of the mean dopamine loss in the different brain regions was significantly inversely correlated with the levels of noradrenaline in healthy control subjects (r = 0.83, P<.001; Figure, B) and in subjects with PD (r = 0.82, P<.001). However, within the striatum (caudate and putamen), noradrenaline concentrations were low and were similar in the subdivisions that are most (caudal putamen, −97% dopamine loss) and least (caudal caudate, −61% dopamine loss) affected in PD (Table 2).
Notwithstanding the many uncertainties inherent in human postmortem brain studies (eg, possible medication effects), we found that regions rich in noradrenaline had smaller dopamine losses than those in noradrenaline-poor areas in PD. This is best illustrated in the dopamine-rich NACS, which has a noradrenaline-poor rostral portion and a caudal subdivision that is strikingly rich in noradrenaline11 and was particularly spared from dopamine loss in PD. This notion receives further support from the observation that dopamine loss is marked in the striatum (caudate and putamen), which contains low levels of noradrenaline. However, it is unlikely that a lack of noradrenaline could be the primary factor responsible for the dopamine neuron death. This is shown by our additional finding that levels of noradrenaline within the different subdivisions of the caudate and putamen were similar, although dopamine was much more markedly decreased in the subdivided putamen. This points to the subregionally selective effect of the still unknown primary factor (or factors) responsible for the nigrostriatal neurodegenerative process in PD.
In studies using animal models of PD caused by dopaminergic neurotoxins (eg, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine and 6-hydroxydopamine), striatal dopamine loss is more severe following lesions of the locus coeruleus noradrenergic pathway.15- 18 Conversely, pharmacological enhancement of noradrenergic activity by an α2-adrenergic antagonist,19 by a noradrenaline transporter inhibitor,20 or by genetic deletion of the noradrenaline transporter20 was neuroprotective. This suggests the possibility that noradrenaline might have a neuroprotective role in PD.4
The mechanism of neuroprotective action of noradrenaline in experimental studies is still debated4 but could involve the neurotrophic function of the noradrenergic brain innervation21 or, especially, the amine anti-inflammatory potential.9,10 The loss of dopamine neurons in PD is accompanied by the appearance of activated microglial cells in brain5- 8 that would be expected to have the capacity to produce cytokines (eg, tumor necrosis factor α and interleukin 1β) and other proinflammatory molecules, thereby aggravating neuronal cell death.22 In this context, experimental studies9,10 have revealed that noradrenaline can inhibit several aspects of the microglial reaction (including the expression of tumor necrosis factor α and interleukin 1β) and thus has the potential of an endogenous anti-inflammatory agent in the brain. Future studies should establish whether the regional extent of microglial activation in PD is inversely related to the brain regional noradrenergic innervation.
Our postmortem brain findings (although only correlational) provide support to the notion (based on animal data) that clinical therapeutic approaches designed to enhance brain noradrenergic activity (eg, by α2-adrenergic antagonists and noradrenaline transporter inhibitors) should be considered as possible neuroprotective strategies in PD. This potential neuroprotective action of noradrenaline should also be kept in mind when treating patients having PD with noradrenergic compounds (eg, noradrenaline transporter inhibitors for depression23,24) for symptomatic purposes.
Correspondence: Junchao Tong, PhD, Human Neurochemical Pathology Laboratory, Centre for Addiction and Mental Health, 250 College St, Toronto, Ontario, Canada M5T 1R8 (firstname.lastname@example.org).
Accepted for Publication: June 16, 2006.
Author Contributions:Study concept and design: Kish and Hornykiewicz. Acquisition of data: Tong, Kish, and Hornykiewicz. Analysis and interpretation of data: Tong, Kish, and Hornykiewicz. Drafting of the manuscript: Tong, Kish, and Hornykiewicz. Critical revision of the manuscript for important intellectual content: Tong, Kish, and Hornykiewicz. Statistical analysis: Tong. Obtained funding: Kish. Administrative, technical, and material support: Kish and Hornykiewicz. Study supervision: Kish.
Financial Disclosure: None reported.
Funding/Support: The study was supported by grant DA07182 from the National Institute of Drug Abuse (Dr Kish).