Parkinson’s disease (PD) affects approximately 1% of the population above the age of 50 years, worldwide [1
]. Costs of treatment per capita in the U.K. alone can be up to £13,804 annually [2
]. Aging populations are generally at greatest risk [3
]. Up to 18 genetic loci have been demonstrated to contribute to familial cases of the disease [4
]. It is estimated that the incidence of PD worldwide will double within the next decade [1
]. All these factors combined are increasingly directing research and development not only towards novel therapies, but also technologies for better diagnosis and disease management. Diagnosis of PD is primarily based on clinical symptoms. However, the error rate is high [5
]. Of these, the main features used for clinical diagnosis are heavily relied on symptoms such as bradykinesia, rigidity, tremors, postural instability and freezing [6
Metabolomics is a well-defined approach used for biomarker discovery and investigating disease mechanisms. Burté et al. undertook metabolomic profiling of the serum samples of early stage PD patients and found an increase in metabolites of the fatty acid beta oxidation pathways [7
]. In one longitudinal study, it was found that a combination of plasma and CSF xanthine and fatty acid metabolites showed significant changes between baseline and the study end point (two years from baseline) [8
]. Markers that were highly correlated with a change in UPDRS scores, indicative of PD progression, included benzoate in the CSF and phenylcarnitine and aspartylphenylalanine in the plasma. Several studies on urine samples of PD patients showed an increase in amino acid metabolism, including phenylalanine [9
], histidine, glycine and tryptophan/kynurenine [10
]. A study by Ohman et al., using NMR based metabolomics on CSF, demonstrated the role of the amino acid alanine, energy metabolism (creatinine) and glucose metabolism (mannose) in distinguishing PD patients from controls [11
]. Metabolic profiling of CSF has also been useful in differentiating newly diagnosed PD patients from controls. Trupp et al. revealed an increase in levels of the amino acids, alanine and methionine, and a reduction of saturated and unsaturated fatty acids in the CSF of newly diagnosed PD patients [12
]. A differential role of glutathione metabolism has also shown a link to PD. While one study found increased glutathione in the plasma [5
], another study revealed lower oxidized glutathione levels in the CSF [13
], further suggesting the involvement of free radicals in PD.
A wide range of in vivo transgenic- and toxin-based models of PD are available. Studies on the mesencephalon of MPTP-induced PD in mice have revealed the role of altered energy [14
], ceramide and sphingolipid metabolism [15
]. A range of lipid species, including lysophosphatidylcholines and phosphatidylcholines are shown to be involved in 6-OHDA toxicity in rat midbrains [16
]. An MPTP-treated goldfish model also indicated the role of phosphocholine metabolism, along with amino acids such as leucine, valine and glutamine [17
]. Additionally, cardiolipins and mitochondria-associated phospholipids were detected in mesencephalon and plasma samples of rotenone-treated rats [18
The aim of this study is to discover metabolic pathways in different tissues in a unilateral PD model. Plasma and liver profiling are often used as an indicator of metabolic changes, whereas in this case the midbrain is the site of primary pathological changes in the brain.
The aim of this study was to elucidate metabolite changes in an in vivo model of PD. To confirm the effectiveness of our model, we performed behaviour tests and immunohistochemistry to examine the loss of dopaminergic neurons in the SNpc. Unilateral lesions of 6-OHDA successfully resulted in the manifestation of motor symptoms, as observed by the cylinder test, and the apomorphine-induced rotation test indicating the intensity of the lesions. Additionally, a significant loss of dopaminergic neurons was observed on the ipsilateral side, as measured by counting TH immuno-reactive positive cells. All these results proved the validity of a “hemi-parkinsonian” model.
Palmitic acid and stearic acid were significantly increased in the plasma, while an imbalance of their monoglyceride forms in the midbrain was also observed. Fatty acids can be transported through the blood–brain barrier via two major routes, either passive diffusion [19
], or facilitated by transporters [20
]. The transport of these two upregulated fatty acids in the context of PD needs to be further studied. One study has shown a decrease of palmitic and linoleic acid in the human plasma [12
]. Moreover, the significant correlation between plasma palmitate and stearate and the cylinder test is of particular interest, suggesting an association with symptoms. This association of the fatty acids with motor symptoms is indicative of the severity of the 6-OHDA lesion and thus damage induced by it. Owing to a lack of biomarkers for PD, diagnosis currently relies heavily on symptoms. It is worthwhile to investigate the clinical potential of these fatty acids in PD, further. Moreover, longitudinal studies will aid in assessing whether palmitate and stearate levels change as neuronal damage progresses in this model. Finding a biomarker that correlates with worsening of symptoms is ideal for tracking disease progression.
Saturated free fatty acids are released into the blood by two major pathways. Lipolysis [21
] is the breakdown of fats in adipose tissue to release triglycerides and free fatty acids into the blood, whereas de novo lipogenesis (DNL) occurs when saturated fatty acids are synthesized from glucose and its metabolites in the liver are subsequently released into the plasma to target tissues in need. There is also evidence to show that palmitic acid may be the major product of DNL [22
]. However, our results show that there was no increase of palmitic or stearic acid in the liver tissue. This observation can be attributed to a swift clearance of the liver fatty acids by the plasma. In addition, oxidation of fatty acids, which takes place in the mitochondria, is an important pathway providing energy [23
]. Mitochondrial dysfunction has also been implicated in PD [24
]. The increased levels of stearic acid and palmitic acid in plasma could, therefore, be a consequence of impaired mitochondria in hepatic or extrahepatic tissues.
Saturated free fatty acids have known effects in the context of neuronal conditions. In one study, a diet rich in palmitic acid (30% palmitic acid of total fat) fed to mice resulted in reduced hippocampal neurogenesis [25
]. Additionally, a diet supplemented with palmitate (2.2% w
) induced endoplasmic reticulum (ER) stress in murine hippocampi and cortices. This study also assessed effects of increasing concentrations of palmitate, of up to 500 µM, on human neuroblastoma SH-SY5Y cells, which are commonly used for PD studies. It was found that 100 µM and higher concentrations of palmitic acid led to an upregulation of the ER stress-associated pro-apoptotic signalling machinery CHOP [26
]. Additionally, 0.2 mM palmitic and stearic acid induced hyperphosphorylation of tau in rat primary cortical neurons, which was facilitated by astrocyte-induced oxidative stress [27
]. Hyperphosphorylation of tau is a hallmark of Alzheimer’s disease. An increased uptake of labelled palmitate into the brain was also observed in patients with metabolic syndrome compared to that of control subjects. Elevated levels of free fatty acids in the plasma have also been linked to metabolic syndrome [28
]. This is particularly important given that PD affects an aging population, many of which suffer from metabolic syndrome as well.
On the other hand, monoglycerides of palmitic acid and stearic acid had a significant difference between the lesioned and non-lesioned sides of the mesencephalon, with no corresponding changes in the cerebellum. The cerebellum was used as a control region because this region remains unaffected by 6-OHDA lesions into the MFB. Monoglycerides are an intermediate product formed during the breakdown of triglycerides by lipolysis [33
]. Whether the imbalance of these metabolic features can be stem from the imbalance of dopaminergic neurons must be further assessed. Additionally, the corresponding levels of these metabolites in the striatal terminals will give a more holistic idea about their association with dopaminergic loss; 1-monopalmitin and 1-monostearin have been shown to be altered in the CSF of patients with inflammatory demyelinating disease such as multiple sclerosis, a disorder affecting nerve fibres [34
In this study, myo-inositol also showed an imbalance in the mesencephalon. Studies on the basal ganglia of patients with a PINK1
mutation have reported an increase in myo-inositol by using MR spectroscopy [35
]. Myo-inositol is purported to be a marker of glial cell death and neuroinflammation [36
]. It is noteworthy that monoglycerides and myo-inositol are also downstream products of the IP3
-DAG signalling pathway [37
]. This pathway plays a role in facilitating release of Ca2+
, which is important in cellular growth and synaptic plasticity [37
Our results are in line with some of the findings from other studies. Lu et al. studied the changes in metabolites in the brains of goldfish treated with MPTP. 1
H NMR-based metabolomics revealed an increase of myo-inositol and linoleic acid in the PD brain [17
]. In another independent study, a paraquat-treated Drosophila
model was used to elucidate changes of metabolites. It was shown that myo-inositol, 1-monopalmitin, 1-monostearin and the fatty acids palmitate and oleate were increased in the heads of the paraquat-treated flies [38
]. In these studies, however, specific changes in the midbrain only were not determined. Furthermore, it must be noted that the findings from this study must be further validated for specificity to dopaminergic loss. This can be confirmed using a lesion that spares dopaminergic neurons but selectively damages surrounding neurons, such as an excitotoxic lesion of the striatum.
4. Materials and Methods
Twenty-six, four to six weeks old male Sprague-Dawley rats were purchased from the Laboratory Animal Unit at The University of Hong Kong. All experimental procedures were in accordance with the Committee on the Use of Live Animals in Teaching and Research of The University of Hong Kong (3491-14). The animals weighed 200 g at the beginning of treatment and were housed in pairs, in a temperature-controlled room with a 12-h dark/light cycle and free access to food and water.
All chemicals and reagents were obtained from Sigma Aldrich (United Kingdom), unless stated otherwise.