Animal models are helpful in understanding the mechanism and environmental factors which trigger the disease process. Studies of MacFabe et al.,  have demonstrated that PA intraventricularly infused to rats provides a suitable animal model to study ASD. Moreover, there are a number of inherited and acquired conditions which lead to elevations of PA and these are related to developmental delay, seizures and gastrointestinal symptoms, resembling some aspects of autism. Thus, PA may be a putative link between dietary or enterobacterially derived metabolites along with genetic predisposition and subsequent features of autism [9, 32–35]. As oxidative stress has been suggested as the primary mechanism of PA neurotoxicity, it was of great interest to test the protective and therapeutic effects of CoQ and melatonin as two supplements displaying antioxidant and free radical scavenger properties.
Hsp is generally much more sensitive to stress than other health indices, making its use as biomarker very common in toxicology . Table 1 demonstrates that PA–treated rats show lower level of Hsp70 which is not in good agreement with our previous study in which the same PA dose induced remarkable elevation of this stress-induced protein. This could be attributed to the fact that potential mechanisms of stress tolerance differ among individuals, populations, and species . This intra-specific variation might be due to genetic variation among individuals, nutritional status, reproductive status, antioxidant status of the individuals, Hsp mRNA stability and pre-existing pool of HSF (heat shock transcription factor) . In the present study, inbred rats were used and led to the absence of intra-subject variability seen as low standard deviation (SD), (Table 1). The protective and therapeutic effects of melatonin shown in Table 1 could be supported by the work of Rodella et al.  in which they reported that melatonin was effective in increasing the reduced level of Hsp70 induced by nicotine in heavy smokers.
Undoubtedly, the magnitude and potential severity of many neurotoxic agents could be measured through the consequent changes in neurotransmitters as biomarkers of brain damage . Table 1 demonstrates the remarkable lower levels of GABA, serotonin and dopamine in PA-treated rat brain homogenates compared to controls. This could be supported by the fact that, PA being capable to access to the brain could induce neurochemical effects on CNS function  including neurotransmitter synthesis and release. Of interest, PA is capable of altering dopamine, serotonin, GABA and glutamate systems in a manner similar to that observed in ASDs [40, 13], partly via changing intracellular calcium release rate . It could be easily observed that PA pre or post-treatment with Co Q (Table 1) or melatonin (Table 2) induced satisfactory amelioration of GABA levels, with Co Q being more potent compared to melatonin. Significant increased levels of serotonin and dopamine, even higher than control subjects was recorded by both supplements. This is consistent with the previous finding of Binukumar et al.  that pretreatment with Co Q caused a significant attenuation of the loss of striatal dopamine and dopaminergic neurons caused by the pesticide, dichlorvos through the activation of the mitochondrial respiratory chain reactions and mitochondrial antioxidant enzyme function. The protective potency of Co Q against brain neurotransmitters depletion as biochemical autistic features in PA-treated rat pups, could be related to the recent work of Kałużna-Czaplińska  in which he reported elevated succinic acid excretion as a marker of Co Q deficiency in children with autism.
Cardinali et al.  proved that preincubation of synaptosome-rich homogenates of rat hypothalamus with melatonin induced significant increases of norepinephrine, serotonin, dopamine and glutamate concentrations. They suggest that exogenously-administered melatonin may affect neurotransmitter accumulation and release in the hypothalamus by modification of the transmitter uptake mechanism rather than by competition with the transmitter for its uptake pump. This can help to understand the protective and therapeutic effects of melatonin reported in the present study and shown as remarkable elevation of serotonin and dopamine concentration.
It is worth noting that OT has been involved in the etiology of autism, with a sex-related pattern and hence it could be used as biomarker of PA-induced autistic features in rodent model . Tables 1 and 2 demonstrate oxytocin levels in brain of control, PA, Co Q or melatonin-treated rat pups. It could be easily noticed that PA induce oxytocin depletion while both Co Q and melatonin were potent in ameliorating the neurotoxic effect of PA to a great extent. Depleted oxytocin could be easily correlated to the recorded depletion in GABA level shown in the same table. Endogenous oxytocin additionally functions as an anxiolytic, acting to increase release of the inhibitory neurotransmitter GABA in the central amygdale . Cumulative evidence from rodent models suggests that both acute and chronic administration of oxytocin reduces physiological and behavioral stress responsively [47, 48].
Inflammation in the nervous system is widely recognized as contributing to a number of neurological conditions. However, the central nervous system (CNS) has also been classically recognized as occupying a privileged site with respect to immune-related phenomena. This dichotomy is widely understood to be a functional manifestation of known CNS. The most prominent element involved in these mechanisms is the blood–brain barrier (BBB), a physical and metabolic barrier separating the CNS from the systemic circulation, creating a unique and stable environment for neuronal activity. Table 1 demonstrates that PA induced lower level of IFI 16 compared to control untreated rat pups. This is unexpected because Gariano et al.  pointed out that there is a signaling pathway linking oxidative stress as an aspect of PA neurotoxicity to upregulation of interferon (IFN)-inducible gene IFI16. The recorded low level could be related to the early finding of Dawson et al. 1998)  that IFI 16 is expressed in CD34+ and monocytoid daughter cells, but is rapidly and markedly down-regulated at the corresponding stages of polymorphonuclear and erythroid development. Moreover, glucose restriction in cells usually accompanied by high AMP/ATP ratio (energetic stress), which activates the AMPK/p53 pathway. Depending upon the energetic stress levels, cells undergo either autophagy or cell death. Given that the activated p53 induces the expression of IFI16 protein, Duan et al.  investigated the potential role of the IFI16 protein in glucose restriction-induced responses. Lower, IFI 16 protein reported in the present study could be related to the neurotoxic effect of PA making brain cells of treated rat pups less adapted to the energetic stress as a well known phenomenon of PA neurotoxicity. Additionally, the recorded low IFI 16 could be easily related to the pro-apoptotic effect of PA. It is well known that Interferons (IFNs) are multifunctional cytokines with antiviral, anti-proliferative and immunomodulatory effects. Activation of type-I IFN-signaling in immune cells inhibits the production of proinflammatory cytokines and activates inflammasomes. Given that the IFI16 and AIM2 proteins are IFN-inducible and can heterodimerize with each other, Veeranki et al.,  investigated the regulation of IFI16, AIM2, and inflammasome proteins by type-I and type-II IFNs and explored that expression of IFI16 protein in THP-1 cells suppresses the activation of caspase-1. This explanation could be supported through considering the work of El-Ansary et al.  in which they recorded activation of caspase 3 as pro-apoptotic biomarker in brain tissue homogenates of PA-treated rat pups. The ameliorating effects of CoQ10 may be attributed to its role in activating the electron transport chain and hence make brain cells more adapted to energy depletion induced by PA. In addition, both CoQ10 and melatonin could be easily related to their antioxidant effects.
The results provide evidence that early PA treatment induces long-lasting behavioral deficits, which are possibly caused by oxygen reactive species generation, and suggest that oxidative stress may be involved in the neuropathology of propionic acidemia.
Table 3 and Figure 1 show that the comet assay is able to demonstrate the aetiology of DNA damage, induced by PA. This could be supported by certain previous studies which have shown that PA can increase ROS generation and oxidative damage to cells . This hypothesis is reinforced by the work of McLaughlin et al.  demonstrating that exposure of striatal and cortical cultures from embryonic rat brain to PA for 24 h provoked DNA laddering and dose-dependent cell death, which was attenuated by antioxidants. PA stimulates lipid peroxidation in rat brain and in the plasma of patients with PAemia [54, 55]. The recorded protective effect of Co Q shown in Table 1 could be supported by Papucci et al.  who reported that Co Q is able to counteract mitochondrial membrane potential depolarization, ATP depletion, cytochrome c release, caspase-9 activation and DNA fragmentation in keratinocytes upon apoptotic stimuli. This could also explain the neuroprotective effect of Co Q through the increase of IFI 16.
In general, the similarity in the magnitude of the protective and therapeutic effects of CoQ and melatonin against PA neurotoxicity can be explained on the basis that both stimulate the expression of antioxidant and detoxification genes, acting in turn as a glutathione system enhancer. A further mechanism of protecting or enhancing cell survival by these two antioxidants lie in the control of damage and signaling function of mitochondria that involves decreased production of ROS. This could inturn; confirm the previously reported mechanism of PPA-induced neurotoxicity through oxidative stress-associated pathways [9, 24, 57].
High values of specificity, sensitivity and area under the curve (AUC) measured by ROC analysis suggested that the studied biomarkers could be used as predictive tools in testing the neurotoxic effects of PA and the potency of CoQ and melatonin in protecting or treating intoxicated rat pups. Among the measured parameters GABA and DNA damage markers were the most predictive recording high specificity and sensitivity and AUC of almost 1.