Accumulated evidence suggests that PD may arise from a combination of genetic susceptibility and exposure to environmental toxins [32, 4, 33]. Indeed, several environmental risk factors such as metals, solvents, carbon monoxide and herbicides have been linked to the incidence and progression of PD [34–37]. Amongst these factors, the herbicide paraquat (PQ) shows clear neurotoxicity in the central nervous system. PQ can enter the CNS through neutral amino acid transporters associated with the blood brain barrier system  and selectively damage the SNpc neurons in mouse models . It has also been reported that a prolonged exposure to non-pneumotoxic levels of PQ causes the damage to basal ganglia . According to the published data, a range of xenobiotics, including paraquat and menadione, can undergo monoelectronic redox cycling in isolated brain capillaries, giving a rise to increased production of ROS [40, 41]. Menadione, for example, is shown to significantly increase the permeability of endothelial cell monolayers in vitro . Thus, there is a correlation between free radical formation and passive permeability of the cerebral microvascular endothelium; therefore, it is possible that the neurotoxicity of PQ may, in part, be linked to its ability to alter the permeability of the blood brain barrier.
Our data clearly confirmed the toxicity of systemically administered PQ, especially towards DA neurons of the SNpc. Three weekly intraperitoneal injections of this herbicide effectively killed between 50–70% of DA neurons in the midbrain of adult male Long-Evans hooded rats and induced PD-like symptoms. The effectiveness of PQ in killing DA neurons seems to vary from 40–70% in reported experimental studies, but its selectivity towards TH-positive neurons in the SNpc region is consitent in all studies. Ossowska et al., (2005) showed that long term paraquat administration caused somewhat less progressive neuronal loss of about 37% loss of dopamine neurons in adult male Wistar rats . McCormack et al., (2002) found only about a 25% reduction in these DA neurons in C57BL/6 mice . However, Brooks et al. (1999) found much greater losses of TH-positive neurons in the substantia nigra pars compacta (~60%) and their terminals in the striatum (~90%) after PQ exposure in C57BL/6J mice given the same dosage used in our study . In the present study, we observed a loss of ~65% dopaminergic neurons using a 3 injection PQ regime. Perhaps, among other possible factors, the different strains of mice and rats used in these two sets of studies may account for the differences in susceptibility to PQ-induced neuronal loss.
The question remains whether our rats may have simply been susceptible to a more general, non-specific neurodegenerative effect from PQ exposure. Unfortunately, we did not test for this possibility by examining other brain areas. In view of this problem, we now use a general marker for neurons namely NeuN to stain various areas of brain in on-going experiments. Although we have yet to complete this study, our preliminary results from those rats that have completed it, clearly show a clear decrease in NueN-positive cell numbers in the midbrain area containing the SNc region but no decline in NeuN immunostaining in other areas of the brain; e.g. hippocampal region (data not shown) following PQ injections. Therefore we are confident that PQ exposure targeted DA neurons of the SNpc.
Moreover, the affected rats showed clear signs of deficiency in fine motor control as indicated by a reduced tendency to turn around and walk backwards on the rotorod. Furthermore, we have established that the brain damage and the performance deficits could be minimized if not prevented by giving the animals a water soluble formulation of CoQ10 (WS-CoQ10) throughout the experiment. This study is, in fact, the first pre-clinical evaluation of WS-CoQ10 as a neuroprotectant of DA neurons.
The neurotoxicity of PQ is linked to its ability to destabilize mitochondria and increase the production of reactive oxygen species (ROS) . Experimental evidence indicates that PQ targets mitochondria of brain cells and the affected mitochondria are a major source of ROS. Using in-vitro assays, we show that PQ can destabilize the mitochondria isolated from mouse brains and from differentiated human neuroblastoma cells , (Somayajulu et al., unpublished data). Furthermore ROS such as superoxide reacts with nitric oxide to form reactive nitrogen intermediates, which impair the mitochondrial respiratory chain and lead to decreased ATP synthesis . Furthermore, we found evidence of increased lipid peroxidation, which could lead to changes in membrane properties and affect cellular homeostasis . Therefore, the evidence suggests that dysfunctional mitochondria are most likely significant contributors to PQ-induced neurotoxicity.
Our data has also shown that placebo offers some protection against oxidative stress; however, placebo alone is not sufficient to protect the neurons from cell death induced by paraquat. The placebo formulation consists of Vitamin E and polyethylene glycol. Coenzyme Q10 on the other hand, works at the mitochondrial level and does prevents neuronal cell death induced by paraquat. Our data adds to the mounting evidence that antioxidants, especially CoQ10 and vitamin E, are important for the management of neurodegenerative diseases . Neuroprotective effects of CoQ10 in the CNS have been extensively evaluated [18, 44] and several in vivo studies demonstrate its protective role against experimental ischemia, sparing the levels of GSH and ATP [15, 23]. CoQ10 is a highly hydrophobic, naturally occurring compound that mainly functions in mitochondrial membranes as a diffusible electron carrier for the mitochondrial respiratory chain complexes. It is also a powerful antioxidant readily scavenging free radicals . Its pharmaceutical applications, however, seem to suffer from the lack of solubility and low bioavailability, both necessitating the applications of high doses to achieve a therapeutic effect. An open-label phase I clinical trial of CoQ10 in PD patients reveal a good absorption and tolerance of CoQ10, however, high dosages were required (up to 1200 mg per day) to achieve some beneficial effects [17, 45]. A recent study by Cleren et al (2008) has shown that oil soluble Tishcon CoQ10 formula provided significant protection against MPTP toxicity in a mouse model . However the effective doses used in this work were 200 mg/kg/day-1600 mg/kg/day (equivalent to 14–114 gm/day for a 70 kg patient). This dose is also extremely high and therefore unlikely to be used in human subjects. On the other hand, the effective daily dose of the water soluble formulation used in our study was 5 mg/kg/day in rats, roughly one fourth of the dose used in the cited above clinical trial and 40 times lower than the dose used by Cleren et al (2008) in mice . The results from our study offer the possibility of using CoQ10 in smaller, clinically relevant doses.
The water soluble formulation of CoQ10 has been reported to be more efficacious as it combines two potent antioxidants, i.e., derivatised vitamin E (PTS) and CoQ10 [23, 20], http://www.Zymes.com. PTS is a pro-drug form of vitamin E (alpha-tocopherol), which was chemically derivatised by sebacic acid and PEG and used as a component (carrier) in WSCoQ10 formulation . These two compounds form a stable and water soluble complex, easy to deliver and test (i.e., in drinking water). The effective daily dose of this formulation that offered significant neuroprotection in our study will translate to 350 mg/day for a 70 kg human subject (roughly one fourth of the dose used in the aforementioned clinical trial). Indeed rats fed the WS-CoQ10 containing diet have shown elevated plasma levels of CoQ10 . Due to the water-soluble nature of our formulation, we have done extensive work on its effect as neuroprotective agent in neuronal cell cultures [21, 22, 24]. In our previous in vitro studies we show that this formulation of CoQ10 protects differentiated SHSY-5Y cells against PQ toxicity by stabilizing mitochondrial membranes, maintaining mitochondrial membrane potential and sustaining ATP production . More recently, we established that it inhibits Bax activity and prevents Bax-induced destabilization of mitochondria in mammalian cells .
Of equal importance in this study is the identification of the fine behavioural abnormalities (indicative of altered motor skills), observed in PQ-treated rats as the reduced behavioural variability for balanced walking on the rotorod (PQ-placebo group, Figure 4A). This was evidenced by rats reducing their spontaneous turning around to walk backward. These PQ-injected rats were receiving drinking water supplemented with derivatised vitamin E (PTS) as placebo and lost over 50% of the midbrain DA neurons (Figure 2E). It should be noted, however, that the rats of the PQ-Placebo group did not suffer extreme behavioural motor dysfunction of slow ambulatory behaviour with hunched postures, typically found in PD patients. Such PD-like general ambulatory behaviour has been observed in mice exposed to similar PQ and MPTP dosages  and a forward rotorod gaiting in rats receiving an unilateral destruction of the nigrostriatal bundle by targeted injections of 6-hydroxydopamine . These behavioural differences, however, are not surprising given the differences in level of SNpc destruction in our animals compared with that of animals in the earlier studies.
Rats in the rotorod study described by Whishaw  obviously had most of their SNpc unilaterally destroyed which also reduced their flexibility in fore- and hind limb paw placement on the contralateral than ipsilateral side of the damage. Given the more general, but lower degree of neuronal loss in the PQ-Placebo group, it is not surprising that we observed the same flexibility of their paw placements as in the unaffected animals (both saline injected as well as PQ-WS-CoQ10 fed groups). It is noteworthy that symptoms of PD, such as resting tremors, postural instability, rigidity and bradykinesia, are not readily detectable until individuals have more than 80% loss of their SNpc neurons .