Selective dopaminergic neurotoxicity modulated by inherent cell-type specific neurobiology

Abstract

Parkinson’s disease (PD) is a debilitating neurodegenerative disease affecting millions of individuals worldwide. Hallmark features of PD pathology are the formation of Lewy bodies in neuromelanin-containing dopaminergic neurons of the substantia nigra pars compacta (SNpc), and the subsequent irreversible death of these neurons. Although genetic risk factors have been identified, around 90% of PD cases are sporadic and likely caused by environmental exposures and gene-environment interaction. Mechanistic studies have identified a variety of chemical PD risk factors. PD neuropathology occurs throughout the brain and peripheral nervous system, but it is the loss of dopamine neurons in the SNpc that produce many of the cardinal motor symptoms. Toxicology studies have found specifically the dopaminergic neuron population of the SNpc exhibit heightened sensitivity to highly variable chemical insults (both in terms of chemical structure and mechanism of neurotoxic action). Thus, it has become clear that the inherent neurobiology of nigral dopamine neurons likely underlies much of this neurotoxic response to broad insults. This review focuses on inherent neurobiology of nigral dopaminergic neurons and how such neurobiology impacts the primary mechanism of neurotoxicity. While interactions with a variety of other cell types are important in disease pathogenesis, understanding how inherent dopaminergic biology contributes to selective sensitivity and primary mechanisms of neurotoxicity is critical to advancing the field. Specifically, key biological features of dopaminergic neurons that increase neurotoxicant susceptibility.

INTRODUCTION

Parkinson’s disease (PD) is the second-most common neurodegenerative disease, affecting millions of individuals worldwide (Brown et al., 2021; Ellis & Fell, 2017; Ostrem & Galifianakis, 2010). A pathological hallmark of PD is the manifestation of neuropathology in selective regions of the brain, which begin with early deposition of alpha-synuclein-containing Lewy bodies in the olfactory bulb and brain stem and eventually migrate to the nigrostriatal region of the brain (Braak et al., 2003; Cave et al., 2016; Ostrem & Galifianakis, 2010). Once PD motor symptoms are severe enough for clinical diagnosis, over half of the dopaminergic neuron cell bodies in the substantia nigra pars compacta (SNpc) have died (Boldyrev et al., 2008; Ostrem & Galifianakis, 2010). Although Lewy body pathology appears in other brain regions with dopaminergic neurons, such as the olfactory bulb, it is dopaminergic neurons of the SNpc that undergo extensive neurodegeneration in PD (Huisman et al., 2004, 2008; Mundiñano et al., 2011). Furthermore, it is the loss of dopaminergic neurons in the SNpc region of the brain that is the major causal factor in the development of the cardinal motor symptoms used for clinical PD diagnostic criteria (He et al., 2018; Ostrem & Galifianakis, 2010). Interestingly, dopaminergic neurons in the ventral tegmental area (VTA) (which are in close proximity to the SNpc dopaminergic neurons) demonstrate significantly less neurodegeneration in PD compared to the SNpc (Alberico et al., 2015; Brichta & Greengard, 2014; E. C. Hirsch, 1994). Moreover, there is varying susceptibility to neurodegeneration in PD in dopaminergic neuron populations within the SNpc (Double et al., 2010). In PD patients, the ventral tier of the SNpc demonstrates death of 90% of dopaminergic neurons, whereas there is only about 25% dopaminergic neuron loss in the dorsal tier (Double et al., 2010). There is no significant evidence of notable differences in the structure of organelles such as the endoplasmic reticulum (ER) or mitochondria between dopamine (DA) cells and non-DA cells, or among midbrain DA cells that exhibit varying susceptibility to degeneration (Damier et al., 1999; German et al., 1989). Therefore, the varying vulnerability should be associated with physiological, neurochemical, and/or metabolic factors that are specific to DA cells and vary among different DA subpopulations, leading to some being more susceptible or resistant to degeneration. There are biological factors that vary between the SNpc dopaminergic neuron populations and other dopaminergic neuron populations in the brain which include- the presence of neuromelanin, dopamine transporter (DAT) density (Harraz, 2023), excessive iron content (W. R. W. Martin et al., 2008), absence of calbindin (a calcium-binding protein) (Yamada et al., 1990), increased lipid peroxidation (Dexter et al., 1989), among many others. Furthermore, there is increasing evidence to demonstrate that the dopaminergic neurons of SNpc have distinct, physiological, biochemical as well as anatomical phenotypes that render them vulnerable to both exogenous as well as endogenous stress inducers. Because the vast majority PD cases are likely attributable to environmental exposure, it is critical to investigate such factors, given the sensitivity of nigral dopamine (DA) neurons to a range of chemical and use classes, such as metals, pesticides, and industrial compounds that vary tremendously in structure and primary mechanism of action (Cannon & Greenamyre, 2011). Primary mechanisms of action in PD neurotoxicity that have been shown to involve mitochondrial dysfunction, alpha synuclein aggregation, and oxidative stress have been extensively studied and published (Boldyrev et al., 2008; Brown et al., 2021; Padmaraju et al.,2011).

The focus of this review is to discuss the inherent neurobiological features within nigral dopamine neuron populations that render these cells especially sensitive to xenobiotic exposures (exemplars in Figure 1). For example, VTA dopaminergic neurons are less susceptible to neurotoxicity induced by rotenone and heterocyclic aromatic amines (HAAs) compared to the dopaminergic neurons of the SNpc (Agim & Cannon, 2018; Greene et al., 2010). It is critical to improve the understanding of how dopaminergic neuron populations differ, which will aid in the elucidation of the mechanisms that cause the selective targeting of SNpc neurons. This will aid in improving translational relevance of mechanistic and animal model studies, as well as potentially foster earlier diagnosis and development of therapeutics that prevent PD progression.

Figure 1. Cellular neurobiology of nigrostriatal dopaminergic neuron that increases selective sensitivity to insults.

Nigrostriatal dopaminergic neurons contain key cellular characteristics that are hypothesized to increase risk of neurodegeneration from endogenous sources (ROS and DA quinones) and exogenous sources (toxicants including chlorpyrifos, harmane, and 2-Amino-1-methyl-6-phenylimidazo(4,5-b)pyridine [PhIP]. Dopaminergic neurons of the substantia nigra pars compacta (SNpc) contain neuromelanin, which is formed by the aggregation of DA quinones and other DA catabolic products with eumelanin, pheomelanin, iron, and other reactive substances. Astrocytes (teal) and microglia (red) are also shown for reference because these cells are well known to be directly involved in the pathogenesis of environmentally induced dopaminergic neurodegeneration. How glia directly contribute to inherent cell-type sensitivity, prior to toxicant exposure remains a critical research need. Neuromelanin has the capability to bind and accumulate neurotoxicants, concentrating them within SNpc dopaminergic neurons. Microglia may elicit an inflammatory response following extracellular neuromelanin secretion. SNpc dopaminergic neurons also have the highest energy requirements of neurons and a reliance on ATP generated by mitochondrial oxidative phosphorylation which produces free radical ROS biproducts. Dopaminergic neurons also uniquely have α-syn localized to both the processes and the cell body. Microglia activation may occur when α-synuclein aggregation producses significant cellular dysfunction and when α-synuclein is excreted from neurons in extracellular vesicles. Low-voltage gated calcium channels (L-VGCC) Cav1.3 and Cav1.2 are both present in SNpc neurons, but the major channel that is expressed in the SNpc dopaminergic neurons is Cav1.3. The SNpc also expresses significantly larger amounts of DA receptor 1 (DR1) and DR2 compared to the ventral tegmental area (VTA). Lastly, the catabolism of DA occurs in both neurons and astrocytes, producing numerous reactive metabolites that increase the basal oxidative stress of dopaminergic neurons. These catabolic enzymes include monoamine oxidase A (MAO-A), aldehyde dehydrogenase (ALDH), catechol-O-methyl transferase (COMT), and aldose reductase/aldehyde reductases (AR) in dopaminergic neurons and MAO-B in astrocytes. MAO-B is particularly important in metabolism of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) to 1-methyl-4-phenylpyridinium (MPP+), which produces SNpc-specific neurotoxicity and neurodegeneration similar to Parkinson’s disease.

NEUROMELANIN

Biosynthesis and structure

Neuromelanin (NM) is a dark brown colored pigment belonging to the melanin family, which renders the dark coloration to the SNpc region of the midbrain. In humans and other primates, it is found in abundance and almost exclusively in the dopaminergic neurons of the SNpc and in the noradrenergic neurons in the locus coeruleus (LC) (Zecca et al., 2004; Zecca, Bellei, et al., 2008). NM pigment manifests as a dark, insoluble molecular aggregate comprised of various components, including melanin monomers (eumelanin and pheomelanin), dopamine oxidative metabolites, proteins, lipids, and metal ions (Zucca et al., 2018) . Positioned within cytoplasmic organelles, these pigments exhibit diverse sizes spanning from 0.5 to 3.0 μm, enclosed by a dual membrane alongside lipid bodies and proteins (Sulzer et al., 2008). Notably, the concentration of NM pigment increases with age, initiating its formation and accumulation early in life (Zecca et al., 2004). The organelle containing NM undergoes a remarkably slow turnover throughout the lifespan of a neuron, serving as a final intracellular destination for numerous molecules that remain undegraded by other cellular systems (Zucca et al., 2018).

The biosynthesis of NM is still not completely understood and is a complex pathway. NM is composed of two types of melanins: the outer coat of eumelanin (black/brown) and pheomelanin (yellow/reddish brown), acting as the core of the NM pigment formation, making an organization of the “casing model” (Greco et al., 2011; Ito, 2006; Y. Liu & Simon, 2005) . While the natural melanin, that is found in the skin, hair and eyes is synthesized by tyrosinase (TYR) which oxidizes tyrosine to dopaquinone , further various redox and condensation reactions lead to the formation of eumelanin and pheomelanin (J. D. Simon & Peles, 2010). The proposed brain biosynthesis pathways involve DA oxidation to DAquinone (DAQ) in the SNpc, or NE oxidation to NEquinonein the LC, and these pathways are suggested to share similarities with melanin biosynthesis through the intrinsic pathway of L-DOPA found in human skin and hair (Greco et al., 2011; J. D. Simon & Peles, 2010) . .

With respect to NM, there is no supporting evidence for enzymatic regulation of the oxidative process responsible for pigment formation.. The presence of TYR for the oxidation of DA is still controversial, but some reports still suggest the expression of TYR in the human brains, although in low levels (Greggio et al., 2005; Tief et al., 1998; Xu et al., 1997). While alternative enzymes may play a role in NM biosynthesis, for instance, enzymatic oxidation of DA (Haavik, 1997), a definitive enzymatic pathway has yet to be conclusively established.. The earliest stages of NM biosynthesis pose a challenge to trace, but the prevailing perspective suggests the participation of three key elements: cytosolic catecholamines, iron (III) ions, and “seeds” comprising aggregated oligomers of peptides/proteins. The presence of iron (III) significantly enhances the auto-oxidation of catecholamines, leading to the formation of a reactive quinone (Sulzer & Zecca, 1999).

NM is mostly found in abundant quantities in humans, which in contrast is not observed in other species at the macroscopic level (Vila, 2019). Nevertheless, it is found in monkeys (Herrero et al., 1993), frogs (Kemali & Gioffré, 1985), dolphins (Sacchini et al., 2018), however only in less quantities. Commonly used research animals, on the other hand, like rodents, completely lack neuromelanin, despite the presence of catecholamines, which are precursors for their formation. One of the reasons why there is a difference in neuromelanin content across species may be because of the different rates of DA synthesis and oxidation (Vila, 2019).

Neuroprotective and neurotoxic roles of neuromelanin

Conventionally, NM is recognized for its intracellular protective role by binding to toxic metabolites produced in SNpc cells, such as oxidized dopamine, metabolites of dopamine, and inorganic metals (D’Amato et al., 1986; Lindquist et al., 1987; Zecca et al., 1994). Additionally, there is evidence that NM functions as an antioxidant. The synthesis of NM results in removal of excessive free cytosolic catecholamines, without which accumulation of cytosolic DA would occur inducing neurotoxic effects (Sulzer et al., 2000). Furthermore, NM can protect neurons from xenobiotic damage by reducing the toxicity of MPTP by accumulating and binding to the active metabolite MPP⁺ in frogs (Lindquist et al., 1988). It is also shown to adsorb intraneuronal paraquat which is another neurotoxic herbicide in frogs, (Lindquist et al., 1988). Nevertheless, there is a suggestion that NM could potentially pose toxicity to dopaminergic neurons. This toxicity is thought to occur through the direct inhibition of proteasome function (Shamoto-Nagai et al., 2004) and the catalysis of free radical production or ROS. NM might also become a source of free radicals through its reaction with hydrogen peroxide (Wakamatsu et al., 2003).

NM sequesters iron and dopamine in the cytosol of neurons, which prevents these molecules from being freely toxic in the cytoplasm. NM has high- and low-affinity of binding sites for iron chelation.. Iron preferentially binds to the high-affinity sites, though in iron overload conditions, the high-affinity sites become saturated and the iron ends up binding to the low-affinity sites (Zecca, Bellei, et al., 2008). When bound to the latter, iron is sequestered in its reactive form and can promote toxic redox reactions. The iron-NM complex prevents the formation of ROS species and toxic quinones which could form because of dopamine oxidation hereby rendering neuroprotection (Zecca, Bellei, et al., 2008). But in PD, when the vulnerable SNpc dopaminergic neurons containing NM die, this leads to NM being secreted extracellularly and being degraded, which releases iron as well as other neurotoxic compounds triggering a cascade of neurotoxicity. This is a key example of NM initially having a protective role, but can turn deleterious when iron overload occurs in brain regions, which is typically what happens in PD (W. R. W. Martin et al., 2008).

Neuromelanin in the pathogenesis of PD

Altered structure and density of NM have been identified in the early stages of PD, implying a potential role for NM in the pathogenesis of PD. NM is considered a critical factor underlying the neuronal vulnerability in PD, where the neurons with higher NM content are more susceptible to cell death as compared to the dopaminergic neurons with lighter NM content. There are some significant studies that investigated the NM content in the vulnerable dopaminergic neurons in PD. One of the earliest studies by Mann and Yates (Mann & Yates, 1983) found lower NM-positive neurons in the dopaminergic neurons of SNpc in PD patients compared to age-matched controls.

Hirsch et al (E. Hirsch et al., 1988), were among the pioneers in establishing a connection between SNpc neuron pigmentation and neurodegeneration in PD. Their findings indicated fewer TH-positive and NM-pigmented neurons in the nigral region of PD patients.. Later, Kastner and group (Kastner et al., 1992) found that the mean NM optical density per the NM-surviving cell population in mesencephalic region was less in PD patients than the controls. In PD patients, however, the percentage of lightly melanized neurons among the total NM-positive neuron population increased.

An in-depth study examined intracellular changes between PD-susceptible and less susceptible cell groups. This study compared cell size, pigment density, lipid components, and α-syn changes in A9 pigmented dopaminergic neurons (SNpc) and A10 neurons in the ventromedial mesencephalic tegmentum. In the most vulnerable PD region, A9 neurons showed NM lipid changes before LB formation, indicating an early pathogenic event. Even in healthy A9 neurons, NM had higher optical density and abnormal α-syn aggregation compared to A10 neurons, suggesting a fundamental alteration in NM preceding A9 neuron neurodegeneration in PD. (G. M. Halliday et al., 2005). Elevated α-syn and NM levels in A9 neurons may increase susceptibility to α-syn accumulation around NM lipids under oxidative stress, causing larger intracellular α-syn aggregates and NM distribution changes. Collectively, these studies showed that there is a quantitative association between the NM-content and the vulnerability of the dopaminergic neurons. It is more than just an either “present or absent” factor for determining the susceptibility of neurons in PD.

Overall, it is established that there is a close correlation between the NM content in SNpc with the susceptibility of neurons in PD pathogenesis. The heightened vulnerability of neurons associated with NM intraneuronal activity may be linked to the substantial presence of Major Histocompatibility Complex class I (MHC-I) in NM-containing organelles within neurons of the human SNpc and LC. Notably, neurons expressing MHC-I, specifically those prone to degeneration in PD, exhibit high MHC-I levels, while other neurons unaffected by PD display low or no MHC-I expression (Cebrián et al., 2014). In individuals with PD, CD8⁺ cells were occasionally observed in close proximity to neurons containing NM and expressing MHC-I in both SN and LC (Cebrián et al., 2014) . This observation implies a potential role for these cells in triggering neuronal death.

The pathophysiology of PD is complex, and the factors which render the dopaminergic neurons susceptible to cell death in this neurological disorder is still elusive and being explored. The major histopathological hallmarks of PD that is defined in the vulnerable dopaminergic neurons of SNpc are: α-syn and NM (Nagatsu et al., 2022). Both complex molecular mechanisms act as endogenous factors that produce a myriad of neurotoxic effects like mitochondrial dysfunction, ROS formation, oxidation of cytosolic DA and its accumulation, loss of proteostasis and neuroinflammation, all leading to neurodegeneration. While both pathways individually can lead to dopaminergic cell death, the interaction of the NM with α-syn in the dopaminergic neurons of SNpc play an interconnected role in PD which is described in detail below.

Neuromelanin interactions with α-synuclein

It is well-known that even though α-syn deposits are found in many brain regions in PD (Braak et al., 2003), the neuronal loss is still limited where the most susceptible cells are the NM-containing dopaminergic neurons. There are multiple studies which demonstrate the interaction of NM with α-syn (Fasano et al., 2003; G. M. Halliday et al., 2005), but if and/or how this contributes to the vulnerability of the SNpc dopaminergic neurons is still elusive. In normal individuals, the proportion of neuronal population containing α-syn as well as NM content increases with age, however the NM negative cells containing α-syn remains constant (Xuan et al., 2011). There is a transition from low NM neurons to high NM containing neurons and its overall accumulation in the SNpc during aging (Zecca et al., 2002).

In vitro studies show that NM collaborates with oxidative stress (OS) and α-syn to enhance the susceptibility of dopaminergic cells (J. Li et al., 2012). Infact, α-syn is known to interact with the enzymes tyrosinase (TYR) (Tessari et al., 2008) and TH (D. Liu et al., 2008), which are both necessary in melanin synthesis and dopamine oxidation pathways. While α-syn inhibits melanin synthesis in melanoma cells, on the other hand, it enhances the melanin content in dopaminergic neuronal cells, serving as a positive association between PD and melanoma due to its differential role in different cells (Pan et al., 2012). Furthermore, immunoreactive free α-syn has been detected in the melanoprotein isolated from substantia nigra of PD patients when the melanin backbone of the proteic material is solubilized, which is not found in control individuals and adults with Lewy bodies (LB) (Fasano et al., 2003). This study supports the hypothesis that α-syn accumulates in granule form in PD during NM biosynthesis.

The aggregates of α-syn redistribute to NM-containing dopaminergic neurons of SNpc in early PD but not healthy individuals (Fornstedt et al., 1989). While NM in SNpc starts at a very early age (around 3 years) (Fedorow et al., 2006), the pathogenic accumulation of α-syn doesn’t begin until middle age. Additionally, α-syn accumulation happens in SNpc but not in the VTA even though both regions of the brain contain dopaminergic neurons (Chu & Kordower, 2007). The major difference between the dopaminergic neurons of SNpc and the VTA is the presence of NM, which prompts contemplation that maybe NM plays a role in age -associated increase in α-syn.

Particularly, α-syn aggregation in PD is affected by the ROS levels and iron content in the cell, and this multifactorial process of events forms a toxic loop which results in manifestation of PD pathology. LB and α-syn have high iron content and this interaction produces high ROS levels. The binding affinity of NM is also influenced by increased ROS and iron during PD. NM granules are surrounded by a dual-layered membrane containing lipid bodies and proteins, and this composition shows its connective role in α-syn pathways modulated by iron and ROS levels (Minakaki et al., 2020; Zucca et al., 2018). Changes in the spatial and temporal distribution of dopamine, NM, and iron reveal a sequential pattern: initial striatal dopaminergic denervation is succeeded by abnormal iron metabolism, culminating in eventual alterations in NM within the SNpc (Biondetti et al., 2020).

Neuromelanin and microglial inflammation

Reactive microglia are a common observation in the SNpc of PD patients indicating that neuroinflammation is a distinct hallmark of neurodegeneration (Tufekci et al., 2012). Microglia serve as vital immune cells in the central nervous system (CNS) and they play an important role in phagocytosis, antigen presentation, and acting together with other immune cells to maintain neuronal integrity and health (Block & Hong, 2005). However, this attempt of protection of nervous system from external stressors makes the microglia cytotoxic to nearby neuronal cells since they release proinflammatory factors like cytokines (IL-1, IL-6, etc.) and ROS, which end up damaging the cells of the CNS (Banati et al., 1993). While the microglia can be activated by several internal as well as external factors, it has been demonstrated that human NM can serve as an endogenous stressor. As discussed above, NM increases the vulnerability of nigral neurons by acting as an iron chelator (Zecca et al., 2001). This accumulation causes excess iron in NM, forms hydroxyl radicals via Fenton reaction, thus renders melanized dopaminergic neurons vulnerable to degeneration (Zecca et al., 2001). While NM affects the mitochondrial complex I activity, induces ROS formation and reduces the TH-positive neurons, the glia plays a role in protecting the neurons against NM toxicity (Depboylu et al., 2007).

In the SNpc of healthy brains, deposition of extracellular NM is found in older people and rarely found in younger individuals. Concordantly, all elderly subjects display a strong, intensive, and hypertrophic microglial reaction, which is rarely observed in the younger people (Beach et al., 2007). Interestingly, in the SNpc of older people, the microglia although are distributed throughout, but are found extremely concentrated near the extracellular NM deposits. These active microglia play a role in engulfing the NM deposits, which is evident by the presence of NM in the cytosol of microglia. Moreover, activation of microglia is not observed in the SNpc of healthy elderly individuals, however there is an increase in the phagocytic activity of microglia (Korzhevskii et al., 2021), and the loss of NM (+)-neurons along with absence of microgliosis is indicative of absence of neuroinflammation in SNpc during normal aging.

In the SNpc, the NM not only localizes to the cytosolic organelles of vulnerable dopaminergic neurons, but also remains in the extracellular space, gaining access to the glial environment. This induces a strong inflammatory response from the SNpc microglia, which is evident by their change in phenotype and release of proinflammatory factors. While human NM is known to be damaging to the remaining dopaminergic neurons, the cellular and molecular mechanisms of how NM damages neurons, directly or indirectly, is still being studied. In PD, since the most vulnerable neurons are the NM-containing dopaminergic neurons, the death of NM-containing neurons leads to higher accumulation of NM extracellularly, which in turn is an endogenous inducer of neuroinflammation. Intracerebral and intranigral injections of human NM in adult male Wistar rats demonstrated a strong inflammatory response incortex, and SNpc, and both dopaminergic cell loss and strong neuroinflammation was observed in the SN (Zecca, Wilms, et al., 2008). The neurotoxic effect of NM is greatly enhanced in the presence of glial cells and has been demonstrated both in vitro and in vivo (W. Zhang et al., 2011; W. Zhang, 2013). The NM-glia interaction leads to exacerbated damage to the surrounding cells by ROS, nitrogen species, chemokines, and cytokines (W. Zhang et al., 2011; W. Zhang, 2013).

There are several pathways that have been defined for the activation of microglia by human NM. In a preliminary study, it was shown that the introduction of human NM into microglia cultures resulted in the initiation of NF-κB-dependent favorable chemotactic responses. Subsequently, this chemotaxis led to the activation of microglia, prompting the release of proinflammatory and neurotoxic mediators such as TNF-α, IL-6, and nitric oxide (Wilms et al., 2003). Another study demonstrated that BV2 microglial cultures treated with NM had significantly increased caspase 8 and caspase 3/7 activity, which lead to microglia activation and increased oxidative stress (Viceconte et al., 2015). The microglia in the SNpc express Clq component on their surface, which plays a vital role in clearing the extracellular NM and debris from the dying vulnerable dopaminergic neurons in PD models (Depboylu et al., 2011). Additionally, a newly proposed degenerative mechanism, driven by inflammatory T cells, suggests the involvement of the major histocompatibility complex class I (MHC-I) (Cebrián et al., 2014). MHC-I has elevated expression in the NM (+)-dopaminergic neurons in the SNpc and LC and accumulates in NM granules. MHC-I has the capability to bind antigenic peptides, presenting them on the cell membrane of the vulnerable dopaminergic neurons. Consequently, these NM (+)containing neurons may be particularly vulnerable to T cell-mediated cytotoxic attacks, leading to cell death and neurodegeneration.

Dendritic cells were observed to phagocytize NM pigment and subsequently secrete interleukin-6 and tumor necrosis factor α, concurrently triggering T cell proliferation. However, it’s noteworthy that the density of dendritic cells in the SNpc is considerably lower than that of microglia, and moreover the response of dendritic cells to NM activation is slower than that of microglia (Zucca et al., 2017). Consequently, the contribution of dendritic cells to the neuroinflammatory and neurodegenerative mechanisms of PD may be limited in comparison to microglia. Additionally, intranigral injections of NM in the rat causes degeneration of vulnerable DA neurons, along with major microglial activation and moderate astrocytosis (W. Zhang, 2013).

Chemical binding properties of NM, relative to toxic exposures and PD relevant sensitivity

As discussed earlier, NM’s neuroprotective actions are due to its ability to bind and sequester toxicants/toxins (i.e. MPTP). However, this binding of neurotoxicants/toxins to NM may also lead to accumulation of these toxic substances resulting in higher prolonged exposures to NM(+) dopaminergic neurons. In mice systemically exposed to MPTP, striatal dopamine depletions were exacerbated in cases where the mice were pre-injected with synthetic DA-NM as compared to the mice treated with MPTP alone (Melamed et al., 1987). MPP+, an active metabolite of MPTP, has been demonstrated to bind with high affinity to NM from the SNpc of humans, as well as to retinal melanin from bovine and frog tissues. This intracellular binding of MPP+ to NM results in storage and subsequent gradual release, ultimately contributing to neurodegeneration in SNpc neurons (D’Amato et al., 1987) . Shortly after, autoradiography of radiolabeled-MPTP in frogs demonstrated high retention and accumulation in pigmented nerve cells up to 15 days post-administration (Sokolowski et al., 1990). Moreover, MPTP induced death in the pigmented neurons, and it is hypothesized that the NM stores and binds the neurotoxic metabolites, causing increase in the cytosolic toxic concentrations by continuously releasing the metabolites from the NM depot. Autoradiography in primates revealed that two heavily melanized regions in the brain, the SNpc and the LC have the ability to sequester MPP+ through terminal uptake and retrograde transport mechanisms (Herkenham et al., 1991). Interestingly, the LC showed a higher capacity for MPP+ accumulation as compared to the SNpc, which could underlie the sensitivity of the LC region in the earliest stages of PD (Braak et al., 2003) . On chronic exposure of MPP+, the SNpc neurons are destroyed, leaving LC neurons intact, although the NM granules appear clumped and extravasated (Herkenham et al., 1991).

A herbicide, paraquat, which is structurally similar to MPTP, has been reported to accumulate in NM in frogs on being injected into the peritoneal cavity, serving a neuroprotective effect, as described earlier, but an overload causing lesions in the pigmented nerve cells, leading to neurodegeneration (Lindquist et al., 1988).

Another cyanobacterial neurotoxin, β-N-methylamino-L-alanine (BMAA) is also found to be retained in melanin containing regions, like the eye, and NM-neurons in frog brain (Karlsson et al., 2009). Interestingly, in vitro studies demonstrated that BMAA interacts more strongly to melanin during its synthesis, than it does when the melanin is pre-formed. The long term exposure of BMAA leads to bioaccumulation of the neurotoxin in NM-containing neurons and changes the overall melanin characteristics. This interaction of neurotoxin with melanin provides a possible combination of linkage between PD and the associated ocular disease, pigmentary retinopathy.

β-carbolines are another class of heterocyclic aromatic amines (HAAs) that are environmental neurotoxic compounds that are found in cooked food, coffee, tobacco smoke and some alcoholic beverages (Skog et al., 2000). The harmane and nor-harmane are structurally similar to MPTP, and intranigral administration of β-carbolines in rats causes lesions in SNpc and striatal dopamine depletion (Neafsey et al., 1995). A study checked the distribution of radiolabeled harmane and nor-harmane in the brains of pigmented and albino mice and also frogs (containing NM). After only a single injection, both the compounds had a high binding affinity to NM and were retained for almost up to 30 days (Stergren et al., 2004). Various compounds show affinity to NM, and this has thought to be a way of neuroprotection, but a slow, progressive release of the compounds that are accumulated on the NM granules may eventually lead to prolonged exposure of vulnerable dopaminergic neurons to these neurotoxins. Recent studies have shown that NM plays a role in HAA-induced neurotoxicity. For this, synthetic DA-NM, and melanin from Sepia offcinalis was used and its binding efficiency was checked by cell-free assay. Both HAA compounds, harmane and 2-amino-1-methyl-6-phenylimidazo[4,5-b] pyridine (PhIP) bind strongly to the analogs of human NM (Lawana, Um, Rochet, et al., 2020). Moreover tyrosinase-expressing SH-SY5Y cells (producing enhanced NM) demonstrated more cell toxicity and oxidative stress due to increased HAA uptake. More detailed study into the mechanism of death of NM-cells by HAA-mediated neurotoxicity was done, where mitochondria was considered if it was affected in the cells. Using the same model of tyrosinase expressing SH-SY5Y cells and treating them with HAAs confirmed that the mitochondrial functions were altered, including decrease in membrane potential, oxidative damage and impaired mitophagy, as compared to the cells which do not overexpress tyrosinase (Lawana, Um, Foguth, et al., 2020).

In vitro and in vivo models of neuromelanin

Since, in PD, vulnerability of the dopaminergic neurons is clearly influenced by the presence of NM in the cell or the presence of oxidized dopamine metabolites, many models were developed that can recapitulate these conditions both in vitro and in vivo. Here it is worth noting that many of these efforts have been conducted to understand the neurobiology of PD. The vast majority of neurotoxicity studies on environmental factors of PD still rely almost exclusively on models nearly devoid of NM, thereby, underestimating internal dose (intracellular dopaminergic neuron levels) and in the absence of critical human neurobiology.

An early study showed that since the neurotransmitter dopamine triggers apoptosis in neuronal cell culture systems, the cultures were treated to synthetic dopamine-melanin to analyze the effects. The dopamine-melanin induces apoptotic cell death in PC12 neuronal cells, suggesting that neuronal damage is caused not only by DA or its metabolites, but melanin also has a toxic effect (Offen et al., 1997). Again, the critical enzyme for synthesis of melanins is tyrosinase (TYR). While TYR may not be responsible for NM production in humans, it has been found to be an excellent tool in ectopic formation in laboratory models. Indeed, TYR plays a vital role in supplying catecholamines to the brain and is required in the enzymatic step of formation of L-DOPA and its further oxidation (Rios et al., 1999). These intermediate metabolites before the formation of NM can also render neurotoxicity in susceptible dopaminergic neurons. A study showed that the SH-SY5Y cells that express tyrosinase under the induction of tetracycline result in the production of ROS species, increase in intracellular DA and enhanced formation of granular structures that resemble the NM that is observed in vulnerable dopaminergic nigral neurons (Hasegawa et al., 2003). Later, the tyrosinase-expressing neuronal cell line SH-SY5Y was considered for use as an in vitro model of PD (Hasegawa, 2010). The treatment of human NM isolated from SNpc was also shown to inhibit the 26S proteasome activity in in vitro SH-SY5Y neuronal cells (Shamoto-Nagai et al., 2004).

Rodents which are used extensively to model PD, unlike humans, almost completely lack NM. As a result, NM, a factor closely associated with PD, which renders the neurons susceptible to neurodegeneration, has surprisingly been overlooked in experimental in vivo paradigms of the disease (Barden & Levine, 1983). To overcome this major limitation, a rodent model expressing human NM in the SNpc dopaminergic neurons was generated, that produced NM at levels comparable to the levels found in an adult human brain. Adult rats which were infused with adeno-associated viral vector (AAV)_TYR vector into the substantia nigra to ectopically form NM as an in vivo model (Carballo-Carbajal et al., 2019). The ectopic expression of TYR in the SNpc of rats resulted in production of human relevant NM, the accumulation of which when reached a particular threshold, led to a PD phenotype like LB formation, disruption of neuronal proteostasis, demise of nigrostriatal neurons and motor deficits. On enhancing the lysosomal proteostasis by TFEB expression, the density of NM decreased, and neurodegeneration was prevented.

In cultured dopaminergic neurons from the rat SNpc, it has been demonstrated that NM biosynthesis depends on the cytosolic concentration of DA that is not stored in synaptic vesicles. Consequently, neurons overexpressing VMAT2 which results in low cytosolic DA, exhibit lower NM synthesis (Sulzer et al., 2000). This relationship is further supported by findings in murine neurons of the locus coeruleus (LC), suggesting an inverse correlation between VMAT2 expression and NM content in the LC (Taylor et al., 2014). Notably, an intriguing inverse relationship has been observed between NM content and susceptibility of neurons (Liang et al., 2004). Specifically, regions such as the ventral tegmental area (VTA) and SNpc exhibit increasing NM content and neuronal vulnerability in that order, while VMAT2 expression decreases from the VTA to SNpc. Clearly factors increasing the free, cytosolic DA pool are likely to increase both NM production and DA neuron sensitivity to endogenous and exogeneous insults. This also makes VMAT2 based treatment a potential approach to reduce the vulnerability of dopaminergic neurons. Based on this principle, using the humanized NM-producing rodents, the VMAT2 in SNpc was overexpressed to see if it would attenuate NM-mediated PD (Gonzalez-Sepulveda, 2023). The strategy decreased the levels of free cytosolic DA which inhibited NM production and resulted in the conservation of nigrostriatal neuronal health and motor functions in the rats.

The present deficiency in understanding the role of NM in both healthy individuals and PD patients stems from a notable distinction. Unlike humans, commonly employed laboratory animal species in experimental research, such as rodents, lack NM. Consequently, a factor closely associated with PD, like NM, has surprisingly been overlooked in experimental in vivo paradigms of the disease. For this reason, there is an urgent need to perform studies of various environmental neurotoxins in well-standardized in vivo systems forming NM which can bridge the gap of the existing translational deficiency.

CATECHOLAMINERGIC NEUROBIOLOGY AND STRESS

Catecholamine/Dopamine Neurobiology – Metabolism and Trafficking

A properly regulated “life cycle” for catecholamines, including the synthesis, vesicular packaging, synaptic release, receptor binding, reuptake, and metabolism, is critical for the health and function of the nervous system (Goldstein, 2013; Goldstein et al., 2013). Interruption of the “life cycle” will produce adverse outcomes ranging from reversible “transmission toxicity” to irreversible, progressive loss of neurons. There is particular weight on cellular metabolism and vesicular uptake to prevent the aberrant production of reactive oxygen species (ROS) and buildup of reactive/toxic intermediates (Bucher et al., 2020). For the purpose of this review and the sake of brevity, the catecholamine, DA, will be featured; however, it should be considered that the concepts discussed, including neurotransmitter disposition and production of reactive intermediates applies to catecholamines (e.g., DA and norepinephrine, NE) as well as monoamines (e.g., serotonin).

DA undergoes oxidative deamination by monoamine oxidase (MAO) A and B to produce the biogenic aldehyde 3,4–dihydroxyphenylacetaldehyde (DOPAL) plus hydrogen peroxide and ammonia. Both MAO isoforms can metabolize catecholamine neurotransmitters and are located throughout the body, though MAO-A is present in the main production sites of NE, epinephrine, and DA (W. J. Burke et al., 2004; Eisenhofer et al., 2004). In humans, MAO-B is present mostly in glia and significantly contributes to the biotransformation of DA(Di Monte et al., 1996; Glover et al., 1977). The MAO metabolite DOPAL undergoes further biotransformation by cytosolic and mitochondrial aldehyde dehydrogenase (ALDH) enzymes to 3,4–dihydroxyphenylacetic acid (DOPAC) using the cofactor NAD, or as a minor pathway by aldose reductase/aldehyde reductases (AR) to 3,4–dihydroxyphenylethanol (DOPET), via NADPH. Homovanillic acid (HVA) is formed via catechol-O-methyl transferase (COMT) conversion of DOPAC to HVA with the cofactor S-adenosyl methionine. HVA is one of the primary DA metabolites found in the circulation and excreted via urine (Eisenhofer et al., 2004).

Under non-pathologic conditions, cellular DA is readily taken up by vesicular monoamine transporter 2 (VMAT2) into vesicles for synaptic signaling and as a means of cyto-protection to prevent cytoplasmic auto-oxidation. Several lines of evidence have demonstrated the critical role of vesicular sequestration of DA via VMAT2 in maintaining the health of dopaminergic neurons (Bucher et al., 2020; Taylor et al., 2011). Transporters, such as DAT, remove DA from the synapse and promote reuptake to the presynaptic neuron, and therefore, are critical as well for maintaining DA homeostasis and possibly controlling short-term plasticity for striatal DA release (Bu et al., 2021; Condon et al., 2019). While their functions are distinct, both DAT and VMAT2 proteins are critical for regulating DA content in the neurons. A ratio of DAT to VMAT has been used to estimate levels of DA in cytoplasmic vs the extracellular space. For instance, higher DAT to VMAT2 ratio is indicative of higher cytoplasmic levels of DA which may lead to increased oxidative stress in dopaminergic cells via DA auto-oxidation or formation of reactive intermediates (Hall et al., 2014; Masoud et al., 2015). On the other hand, a higher VMAT2 to DAT protein content is neuroprotective as it leads to removal excess cytoplasmic DA, thereby limiting oxidative stress in neurons (Guillot & Miller, 2009; Lohr et al., 2014, 2016). DAT and VMAT2 also serve as avenues for pharmacological modulation, which result in differential compartmentalization of DA in the extracellular or the cytoplasmic compartment of the synaptic terminal.

Mechanisms for Catecholaminergic Stress

Catecholaminergic stress occurs following interruption of neurotransmitter homeostasis, including metabolism and trafficking, and produces oxidative stress (e.g., ROS and lipid peroxidation products) and reactive catecholamine intermediates. Disruption of catecholamine homeostasis is reported in several pathological conditions including PD, as well as the mechanism of neurotoxicity from drugs of abuse, e.g., methamphetamine (Cagle et al., 2019; Goldstein, 2020; Kim et al., 2017; Klein et al., 2017; Maia & Frank, 2017; Sato, 2012; Tripp & Wickens, 2008). Altering homeostasis can adversely affect neuronal signaling as well produce reactive/toxic intermediates proposed to injure both neurons and glia, serving as a chemical trigger for later pathogenesis (e.g., PD). The mechanism(s) for interference with normal catecholamine neurobiology may be complicated, involving genetics or xenobiotics or gene x environment interactions; however, recent evidence suggests that enzymes/protein important for the “life cycle” of neurotransmitters are mechanistic targets for neurotoxicants: VMAT2, DAT and ALDH/AR.

Impairment of VMAT2 function prevents packaging of local stores of monoamine neurotransmitters, yielding unregulated cytosolic monoamines such as DA, which can induce neurodegeneration, relevant to PD (Bucher et al., 2020; L. Chen et al., 2008; M. Chen et al., 2008; Pifl et al., 2014). Impairment of VMAT2 function facilitates progressive loss of DA as well as other monoaminergic neurons (Caudle et al., 2007; Taylor et al., 2011, 2014), and conversely, the overexpression of VMAT2 may mitigate PD-related degeneration (Lohr et al., 2014). In fact, decreased vesicular storage appears to be associated with catecholaminergic neurodegeneration supporting the idea of “catecholamine auto-toxicity” (Goldstein, Holmes, et al., 2015).

Disruption of DAT function, including structure and expression, may produce catecholaminergic stress (Bu et al., 2021). Of note, DAT is a target for drugs of abuse such as methamphetamine, which alters both structure and activity of the transporter, yielding DA-mediated dopaminergic toxicity (Fumagalli et al., 1998; Hadlock et al., 2009). Interestingly, there is evidence that a decrease in DAT for methamphetamine users augments the risk of PD pathogenesis (Granado et al., 2013; Jayanthi et al., 2022). Conversely, the function of DAT requires fine-tuning as an increase in its cellular expression and activity can be detrimental to neurons, especially those lacking VMAT2. Previous work has demonstrated that overexpression of DAT or the generation of neurons engineered to express DAT post-synaptically in neurons lacking VMAT2 function produces progressive neuron loss (L. Chen et al., 2008; Masoud et al., 2015). In addition, genetic DAT variants combined with pesticide exposure may increase PD risk (Kelada et al., 2006; Ritz et al., 2009).

MAO plays a key role in the metabolism of monoamines, including DA, and aberrant activity is proposed to mediate neurodegeneration (Goldstein, 2020). As noted above, the initial products following oxidative deamination are aldehydes (i.e., DOPAL for DA), hydrogen peroxide and ammonia. Elevation of (astrocytic) MAO B expression in mice produces selective and progressive loss of DA neurons in mice (Chamoli et al., 2018; Mallajosyula et al., 2008).

DA metabolism, including ALDH and AR enzymes, is hypothesized to be a mechanistic target of neurotoxicants producing catecholaminergic stress. Exposure to pesticides is a risk factor for PD, with some specific agents known to alter DA homeostasis, including rotenone and the organochlorine, dieldrin (Hatcher et al., 2007; Richardson et al., 2006; Ritz et al., 2016; D. K. Simon et al., 2020; Tanner et al., 2011). Rotenone decreases ALDH activity by inhibiting mitochondrial complex I, thereby reducing the available NAD+ for ALDH to detoxify biogenic aldehydes, such as DOPAL (Goldstein, Sullivan, et al., 2015; Landau et al., 2022). A previous report demonstrated that dieldrin elevated extracellular DOPAL in vitro, likely via oxidative stress-mediated impairment of ALDH and AR enzymes (Allen et al., 2013). Exposure to the metal degreaser trichloroethylene (TCE) has been implicated as a risk factor for PD (De Miranda & Greenamyre, 2020; Dorsey et al., 2023). While dopaminergic toxicity could be the result of numerous mechanisms, including TCE/metabolite-mediated mitochondrial dysfunction, it is interesting to note that a major metabolite of TCE, trichloroacetaldehyde, is an ALDH inhibitor (Koppaka et al., 2012). In addition, Fitzmaurice et al. showed an association between human exposure to the fungicide benomyl and PD occurrence via an epidemiological study and further demonstrated that benomyl (likely via an active metabolite) potently inhibits ALDH enzymes, thereby disrupting DA homeostasis (Casida et al., 2014; Fitzmaurice et al., 2013, 2014).

Tyrosine hydroxylase (TH) catalyzes the hydroxylation of L-Tyr to L-3,4-Dihydroxyphenylalanin (L-DOPA), which represent the rate limiting step for the synthesis of DA (Eisenhofer et al., 2004). Decreased cellular/tissue expression of TH is used as a biomarker for loss of dopaminergic neurons in vitro and in vivo, and reduction of the enzyme in the SNpc is a hallmark feature of PD. Interestingly, dysregulation of TH is thought to contribute to PD pathogenesis beyond the decrease in DA synthesis. Previous studies have shown that L-DOPA exhibits a trophic role, yielding an increase in TH-positive neurons; therefore, loss of TH activity may be detrimental to dopaminergic health (Datla et al., 2001; Lipski et al., 2011). As discussed later in this article, TH is essential for the synthesis of neuromelanins, which have an important role in the homeostasis of DA neurons. In addition, a complex relationship between α-syn and TH has been hypothesized, based on several reports, indicating α-syn-mediated modulation of TH phosphorylation and activity and post-translational modification of α-syn catalyzed by TH (Jin et al., 2022; Perez et al., 2002; Wu et al., 2011).

Other environmental pollutants have emerged throughout the years as mediating catecholaminergic stress. Polychlorinated biphenyl compounds (PCBs) are persistent industrial chemicals and have been known to cause toxicity in multiple organ systems (Carlson et al., 2023). Despite the ban on commercial manufacturing of PCBs in the US since 1979, they still persist in the environment (Bannavti et al., 2021; Herkert et al., 2018; Herrick et al., 2004; Hua et al., 2023; Marek et al., 2017) and are also inadvertently produced as byproducts of paint and pigment manufacturing industries (Anezaki & Nakano, 2014; Bartlett et al., 2019; Hu & Hombuckle, 2010). The first evidence of neurotoxicity caused by PCBs came about after a mass-poisoning event in 1968 (Mitoma et al., 2015). Since then, various epidemiological, in vivo and in vitro studies have supported the evidence for PCB-induced neurotoxic outcomes, where, PCBs are implicated in neurodevelopmental disorders such as ADHD, ASD, and learning deficits, and, in neurodegenerative conditions such as PD (Bouchard et al., 2014; Bullert et al., n.d.; Caudle et al., 2006; Cheslack-Postava et al., 2013; Holland & Pessah, 2021; Keil et al., 2019; Rosenquist et al., 2017; Schantz et al., 2003). The mechanistic understanding of PCB-induced neurotoxicity, albeit largely neuron-centric, has demonstrated that PCBs impair DA neurotransmission in the brain (Seegal et al., 1990). Significantly reduced cellular DA concentrations were reported from an in vitro study conducted in PC-12 cells (Seegal et al., 1990). Successive in vivo studies also reported significantly reduced DA concentrations in striatal and substantia nigral tissues from rhesus monkeys (Macca mulatto) and a non-human primate (Macaca nemestrina), which were orally dosed with commercial PCB mixtures (Seegal et al., 1990). The adult male non-human primates (Macaca nemestrina) also showed sustained decrease in brain DA concentrations after discontinuation of the PCB exposure (Seegal, 1994). In Sprague-Dawley rats, in utero and lactational exposure to ortho-substituted PCBs (i.e., PCBs having chlorine atom substitutions at ortho positions of biphenyl structure) were shown to reduce brain DA concentrations in offsprings that sustained through adulthood (Seegal et al., 1997). Further, mice exposed to commercial PCB mixtures at concentrations relevant to human exposure showed significant dose-dependent reduction in DAT levels in the striatum (Caudle et al., 2006).

Of question, which cell type is the source of catecholaminergic stress and resulting injury relevant to PD? While neurons are an obvious choice given their function in neurotransmitter synthesis, trafficking and metabolism, astrocytes may also have a significant role in mediating catecholamine-related injury. Astrocytes are an abundant glial cell type of the CNS, attributed with various functions that bring about healthy and diseased states of the CNS (M. A. Anderson et al., 2016; C. O. Davis et al., 2014; Hayakawa et al., 2016; John Lin et al., 2017; Kamphuis et al., 2012; Tasdemir-Yilmaz & Freeman, 2014). Depending on the type of neurotoxic insult, astrocytes can either directly be damaged by the toxicant (e.g. oxidative stress, cell swelling induced by manganese toxicity) (Milatovic et al., 2007; Rao et al., 2007) or, they can become reactive and assume neuroprotective or neurotoxic functions, as in the case of traumatic brain injury (M. A. Anderson et al., 2016) and Herpes simplex virus 1 (HSV-1) encephalitis (Hensel et al., 2019), respectively.

Astrocytes express the central enzymes responsible for DA metabolism, including MAO B and COMT (Cahoy et al., 2008; Westlund et al., 1985). In addition, they express organic cation transporter 3, NET, DAT and possibly VMAT2 proteins required for monoamine transport and storage (Cahoy et al., 2008; Takeda et al., 2002). Recent evidence also demonstrates that cortical astrocytes heavily regulate DA homeostasis during postnatal development of prefrontal cortex, and the lack of this homeostasis by astrocytes significantly impairs synapse and spine formation and maturation, as well as cognitive functions (Petrelli et al., 2020). As noted above, overexpression of astrocytic MAO B yielded selective and progressive loss of DA neurons (Mallajosyula et al., 2008) . Taken together, astrocytes are central to overall neurodevelopment, cognitive function and brain health, and their structure and function is affected by dopaminergic neurotoxicants. It is critical that further studies make use of holistic approaches for better understanding of mechanisms of neurotoxicity, as the cellular processes in the brain are highly interconnected, interdependent, and regulated.

Outcomes of Catecholaminergic Stress

The auto-oxidation of DA within the cytosol produces both ROS and a highly reactive quinone (GRAHAM et al., 1978; Hastings, 2009; Hastings et al., 1996; Hastings & Zigmond, 1994). The resulting oxidative stress and generation of reactive species can overwhelm cellular defenses and serve as a source of injury. The DA quinone is a soft electrophile that can quickly react with proteins such as alpha-synuclein (α-syn), parkin, tyrosine hydroxylase (TH), and glutathione peroxidase 4 via cysteine or selenocysteine residues and cause mitochondrial dysfunction (Bisaglia et al., 2007; Hauser et al., 2013; Kuhn & Arthur Jr, 1999; Plotegher et al., 2017; Yu etal., 2015).

DA that does not undergo auto-oxidation may be metabolized to DOPAL via MAO. DOPAL is readily metabolized by ALDH and AR, as noted above; however, pesticides and products of oxidative stress can impair these enzymes yielding a build-up of the biogenic aldehyde (Jinsmaa et al., 2009). DOPAL is a unique molecule that possesses both a catechol (pro-oxidant) and aldehyde (electrophile) moiety, thought to contribute to its known toxicity (Rees et al., 2009). The catechol can oxidize to a semiquinone and subsequently to a quinone, via auto-oxidation or enzymatically by cyclooxygenase 2, producing a highly unstable species with unknown biological relevance (D. G. Anderson et al., 2011; Follmer et al., 2015). DOPAL readily reacts with proteins at physiologic pH via lysine yielding production of a ROS and stable adduct, rearranged from a Schiff base (D. G. Anderson et al., 2016; Coelho-Cerqueira et al., 2019; Jinsmaa et al., 2009). Further reactions may yield unique linkages, protein cross-linking and aggregation (Werner-Allen et al., 2018). Protein targets include TH, ALDH, glutathione-S-transferase and α-syn, the latter shown to oligomerize, pertinent to PD and related synucleinopathies (Crawford, Bowman, et al., 2021; Jinsmaa et al., 2018, 2020; Masato et al., 2023; Mexas et al., 2011; Plotegher et al., 2017; Vanle et al., 2017; Werner-Allen et al., 2016; Werner-Allen et al., 2018). Based on such evidence noted above, the “catecholaldehyde hypothesis” has emerged, which largely implicates DOPAL as a contributing factor to PD progression (Goldstein, 2020; Panneton et al., 2010). Additionally, loss of ALDH1A1 gene/protein expression is a proposed marker of PD further implicating DOPAL production (Grünblatt et al., 2018; Mandel et al., 2005; Molochnikov et al., 2012).

Pathogenic Markers of Catecholaminergic stress and Dopaminergic Dsyfunction

Given the potential for products of catecholaminergic dysregulation to cause damage to neurons and glia yielding later degeneration, there is interest in detecting and quantifying such toxic species in vitro and in vivo. Due to the high protein reactivity of reactive DA products, including the quinone and DOPAL, measurement of such species as an indicator of catecholaminergic stress is problematic. In addition, the isolation of modified proteins and confirmation of adducts is a laborious process, and it is uncertain whether or not a serum biomarker can be identified with sensitivity and accuracy. Recent work has demonstrated the in vitro production of L-Cys-DOPAL and carnosine-DOPAL conjugates that can be readily measured via LC/MS (Crawford, Gilardoni, et al., 2021; Nelson et al., 2019). Both L-Cys and carnosine serve as carbonyl scavengers and rapidly form conjugates with DOPAL. For L-Cys, the free amine and thiol react with the DOPAL-aldehyde to produce a thiazolidine that demonstrates a measure of stability, permitting its measurement in cell culture potential as a surrogate marker for aberrant DOPAL production. Of interest, the aldehyde metabolite of NE also forms a stable conjugate with carnosine, presenting opportunity to assess catecholaminergic stress via impaired NE metabolism and trafficking (Monroe & Anderson, 2021; Wanner et al., 2019). Such an approach has been used to probe the reactive aldehyde acrolein as a biomarker. Carnosine reacts with lipid peroxidation products, such as acrolein and 4-hydroxy-2-nonenal, and for acrolein, the carnosine-conjugates have been measured in urine and correlated with disease risk (Y. Liu et al., 2003; O’Toole et al., 2021). Future work is needed to demonstrate whether the conjugate can be measured in serum and if it undergoes further processing (e.g., methylation). In addition, it will be important to determine the role of peripheral versus central DA metabolism in the production of the conjugate.

In summary, there is accumulating evidence that catecholaminergic stress is an important factor in PD pathogenesis; however, our understanding of the underlying neurotoxic mechanism(s) for production of such stress (e.g., via environmental organochlorine pollutants) and its specific role in disease is still limited. Further research is needed to identify environmental mediators which disrupt catecholamine homeostasis as well as the cellular targets from the reactive intermediates produced, specifically, biogenic aldehydes. Moreover, elucidation of mechanisms for production of catecholaminergic stress and its role in disease presents unique opportunities for the determination of therapeutic targets and biomarkers for earlier PD diagnosis.

CALCIUM SIGNALING - CA_V1.3

Voltage-gated clcium channel (VGCC) overview

Voltage-gated calcium channels (VGCC) are a group of transmembrane proteins that mediate the flow of calcium across cell membranes which lead to the regulation of neurotransmission (including neurotransmitter release) and gene expression (M. R. Alam et al., 2022; Brown-Leung & Cannon, 2022; Zamponi et al., 2015). VGCCs are especially important in mediation of calcium entry after membrane depolarization during neuronal action potentials (Zamponi et al., 2015). The VGCCs expressed in CNS neurons include high VGCCs: L-type (Cav1.1, Cav1.2 and Cav1.3), P/Q-type (Cav2.1), N-type (Cav2.2), and R-type (Cav2.3); and low VGCC: T-type (Cav 3.1, 3.2, 3.3) (Brown-Leung & Cannon, 2022; Dolphin & Lee, 2020). The major VGCCs that drive the release of catecholamines (including DA) into the synaptic cleft are the L-type high-VGCCs (Cav1.2 and Cav1.3) (Brown-Leung & Cannon, 2022; Dolphin & Lee, 2020; Simms & Zamponi, 2014)..

L-type VGCCs

L-type channels have 4 isoforms: Cav1.1, Cav1.2, Cav1.3, and Cav1.4 (Dolphin & Lee, 2020). Cav1.3 and Cav1.2 are primarily found throughout the brain, whereas Cav1.1 and Cav1.4 are found in the skeletal muscle and retina, respectively, with very little expression in the brain (Hurley & Dexter, 2012). Cav1.2 and Cav1.3 as mentioned above are both L-type channels that are primarily expressed in the dopaminergic (DA) neurons (Ferron et al., 2021). Both play a crucial role in the regulation of calcium entry, (Pinggera & Striessnig, 2016) which is important for many neuronal functions, however, they differ in biophysical properties, expression patterns and their role in neuronal physiology. Cav1.2 regulates the neuronal excitability (Vierra et al., 2021) and synaptic plasticity (Sridharan et al., 2020), is involved in the shaping of action potential and release of neurotransmitters (X. Zhang et al., 2016) (Mueller et al., 2023). However, Cav1.3 on the other hand is predominantly expressed in specific neuronal populations including SNpc DA neurons (Verma & Ravindranath, 2020). It is responsible for pace making activity and calcium signaling in SNc DA neurons (Shin et al., 2022).

Cav1.3 is involved in functions such as cardiac pace making, hormone secretion, neuron firing and plasticity (Pinggera & Striessnig, 2016) . The influx of calcium through the Cav1.3 activates calmodulin (CAM), which is bound to the intracellular C-terminal domain of Cav1.3. The first process, i.e. calcium dependent inactivation protects the neurons from toxic intracellular calcium concentrations. The second process facilitates its open state probability. The third process activates CamKII and MAPK/ERK signaling cascade leading to the activation of cAMP response element-binding protein (CREB)-dependent transcription within the nucleus. The following channel is also needed for normal neuronal development, maturation of synapses as well as pruning.

Apart from the functions defined above with respect to Cav1.3, the following channel plays a distinct and critical role in the physiology and pathology of SNpc DA neurons. These neurons are involved in autonomous pace making activity that is regulated by Cav1.3 (Surmeier et al., 2012) along with a host of other ion channels, receptors, and transporters (Liss & Striessnig, 2019). DA release regulated by this autonomous pace making activity can increase in a phasic manner upon rewarding stimuli or decrease upon aversive signaling (Cataldi et al., 2022; Foehring et al., 2009). Since these processes involve a high influx of intracellular calcium levels (Surmeier & Schumacker, 2013) , the following neurons are put under an energetic burden which can lead to neurodegeneration since they have a moderate calcium buffering capacity (Liss & Striessnig, 2019).

A dysregulation in calcium homeostasis could lead to neuronal death, long term neurodegeneration in PD, Alzheimer’s diseases, and Huntington’s disease. Although Cav1.3 is much less abundant than the other isoforms mentioned above, it has been implicated in the pathogenesis of PD. PD is a movement disorder distinguished by locomotor deficits including resting tremor, bradykinesia, and postural instability. As discussed above dopaminergic neurons in the SNpc are distinguished by their pace making activity, which is regulated by Cav1.3. In particular, the striatal MSNs that are provided with an input from the nigral DA neurons have a distinctive physiology, including autonomous pacemaking activity modulated by the channel discussed above. The following pace making activity causes tonic release of dopamine in the striatum along with phasic increases in the presence of rewarding stimuli and decreased with respect to aversive signaling. The maintenance of the physiological ion gradients of these tonically active neurons creates a bio-energetic demand along with the challenge of increased Ca²⁺ influxes. Although this increase in Ca²⁺ levels contribute to the normal physiology of these cells, it is also the cause of neurodegeneration because of activation of Ca²⁺ dependent apoptotic enzymes in a population of neurons that have a modest Ca²⁺ buffering capacity. This sudden influx of Ca²⁺ is countered by adenosine triphosphate (ATP) mechanisms mediated by mitochondrial oxidative phosphorylation. This process eventually leads to the generation of ROS and their byproducts. This high metabolic demand leads to the conclusion that the following voltage gated Ca²⁺ channel makes the DA SNpc neurons at risk of degeneration due to an overload of calcium due to poor calcium buffering capacity and mitochondrial oxidant stress in aging and PD. The energetic burden due to the metabolic demand is further aggravated by the fact that SNpc neurons influence a significantly high number of striatal neurons (Bolam & Pissadaki, 2012) due to their highly branched and long axons (Surmeier & Schumacker, 2013).

Guzman et al. reported the sensitivity of DA neurons in SNpc due to mitochondrial stress by using a transgenic mice expressing a redox sensitive variant of green fluorescent protein (roGFP) with a mitochondrial-matrix targeting sequence (Mito-roGFP) (Guzman et al., 2010). When the basal oxidation of mito-roGFP was compared with ventral tegmental area (VTA) neurons it was found that it was significantly higher in neighboring SNc DA neurons. It was found that the calcium influx through plasma membrane was responsible for the mitochondrial oxidant stress, showing that it specifically renders SNc DA neurons vulnerable to oxidant stress (Guzman et al., 2010). Moreover, it has been reported that increased levels of pro-oxidant iron and reduced levels of glutathione contribute to the increase in susceptibility of dopaminergic neurons in SNpc to oxidative stress (Costas & Faro, 2022; Pradhan et al., 2020). They further showed how antagonizing L-type channels significantly lowered the extent of mito-roGFP oxidation but had no effect on the neighboring VTA DA neurons. Another study by the same group reported that mice pre-treated with Isradipine (L type channel antagonist) showed a reduced 6-hydroxydopamine (6-OHDA) induced TH fiber density. The pretreatment showed that higher concentration of the following drug results in a decrease in the available Cav1.3 channels, thereby increasing protection. Near 50 ng/ml (~135 nM), the dose-response curve for cell bodies plateaus, suggesting that approximately 95% of the Cav1.3 channels must be antagonistic for nearly full cell body protection (Ilijic et al., 2011).

Using the L-type VGCC inhibitor isradipine, Guzman et al. (2018) provided evidence on the mitochondrial stress induced by CAv1.3 by diminishing the effects of calcium dependent mitochondrial oxidant stress by systemic treatment of Isradipine in PD models of mice (Hurley & Dexter, 2012). In vivo tracking showed an increased level of mitophagy in SNc DA neurons than the ones treated with Isradipine. The following observation also led to the possibility of DA neurons in the SNc to be therapeutically tractable (Guzman et al., 2018). Given these encouraging findings from these studies; Isradipine has not yet proven to be a major breakthrough in the treatment of Parkinson’s Disease (PD) in clinical trials (Maiti & Perlmutter, 2020) although a few promising observations were made (Simuni, 2020; Venuto et al., 2021). A randomized trail testing Isradipine versus placebo in early Parkinsons disease reported no statistically significant difference between the isradipine and placebo groups. However, they mentioned that one of the limitations of the study was that the dose of Isradipine may not have been sufficient to engage the calcium channels associated with neuroprotective effects (Simuni, 2020). Another study aimed at analyzing the plasma samples from phase-III clinical trials of Isradipine to investigate the population pharmacokinetic modeling reported that even though exposure to Isradipine did not correlate with the primary clinical outcome(Venuto et al., 2021) it decreased the risk of needing antiparkinsonian treatment in comparison to placebo. However, the following was the case observed only in males and not females(Venuto et al., 2021). Another Study reported that that achievable brain levels of Isradipine could potentially decelerate the progression of early-stage Parkinson’s Disease (Surmeier et al., 2022).

A recent study published in 2020 explored the role of Cav1.3 L-type calcium channels, particularly the isoforms Cav1.3₄₂ and Cav1.3_42A (C terminally truncated splice variant), in SNpc DA neurons. They showed that the mRNA expressions for the splice variant Cav1.3_42A were higher in the mouse ventral midbrain and SNpc in comparison to cortex or striatum. This shows that the following isoform is expressed more in the midbrain DA neurons in SNpc. Moreover, the calcium current density through this isoform is known to be approximately 2.5 times higher than the full-length channel isoform. Given the spatial localization and higher influx of these channels, nigral DA neurons are more vulnerable to calcium overload through activation of this isoform since the pace making activity is accompanied by CaV1.3 channel. Moreover, upon the treatment of MPTP, the decrease of DA neurons was accompanied by a maintained expression of the two isoforms during degeneration in a parkinsonism model (Verma & Ravindranath, 2020).

The pathogenesis of PD also involves aggregation of α-synuclein which include Lewy bodies, Lewy neurites and glial cell infusions (Rocha et al., 2018). The following presynaptic protein has 3 main components consisting of the N terminus, non-amyloid-ß component (NAC) and the C terminus. The C terminus is negatively charged and is involved in calcium binding activity (Nielsen et al., 2001). There is an added pathogenic system of increased intracellular calcium, which is related to α-synuclein. This increase has not been observed in α-synuclein knockout cultures after MPP⁺ exposure, suggesting that neurons exposed to α-synuclein oligomers face an increased inflow of calcium due to stress conditions (Lieberman et al., 2017). Due to a change in these levels (Angelova et al., 2016), Ca²⁺ homeostasis is disrupted which leads to ATP taking control of the maintenance of electrochemical gradient across membranes, which results in an alteration of neurotransmitter release. Decreased calcium oscillation of dopamine neurons in SNpc trigger slow tonic firing, leading to oscillations in membrane potential. As a consequence of this, calcium enters mitochondria resulting in an increase of mitochondrial Ca²⁺ which further promotes α-synuclein aggregation. Pathological aggregation of α-synuclein leads to mitochondrial dysfunction and impairment of complex-1 function. Since DA neurons have higher requirements for mitochondrial activity, they are more vulnerable to cell death due to α-synuclein aggregation as a consequence of impaired cellular calcium influx. A study in 2018 showed the treatment of Cav1.3 blocker Isradipine significantly reduced the α-synuclein oligomerization, suggesting a link between α-synuclein and Cav1.3 is essential for their involvement in conveying dopaminergic toxicity, especially in SNpc since it exhibits calcium dependent pace making activity (Guzman et al., 2009) making them more vulnerable to this pathophysiology (Lautenschläger et al., 2018). Newer studies have also indicated that other calcium channels may be important for DA neuron vulnerability as well. One such study indicates that the auxiliary β2a and β2e splice variants of voltage-gated calcium channels enhance the gating characteristics of Cav2.3 channels (Siller et al., 2022). This enhancement leads to the continuous influx of calcium during pace making activities in dopaminergic neurons of the substantia nigra (Siller et al., 2022). The following influx of calcium might play a role in increased vulnerability of these neurons to neurodegeneration as observed in PD (Siller et al., 2022). Another study also indicated that T-type calcium channels can render dopaminergic neurons vulnerable to mitochondrial stress as seen in PD (Tabata et al., 2018). It was reported that the knockdown of T-type calcium channel or the use of antagonist to block these channels resulted in decreased the vulnerability to rotenone-induced stress in PD (Tabata et al., 2018).

Extensive neurobiology studies and studies using 6-OHDA strongly suggest Cav1.3 reliance of specific dopamine neurons would underlie both broad sensitivity in PD and toxicant specific sensitivity. Moreover, such sensitivity may arise from differential expression of individual calcium channels across dopaminergic nuclei such as the SNpc and VTA.

DNA DAMAGE AND DOPAMINE NEURONS

Genome integrity is critical for neurons

Accumulation of DNA damage (lesions) and DNA repair defects have been linked to several neurodegenerative diseases, with increasing evidence for DNA damage in PD pathophysiology (Gonzalez-Hunt & Sanders, 2021; Milanese et al., 2018; Rass et al., 2007; Sanders, McCoy, et al., 2014; Sanders & Greenamyre, 2013). Cells are continuously exposed to various endogenous and exogenous sources of DNA damage. Unrepaired or incorrectly repaired DNA lesions can interfere with fundamental cellular processes, which can lead to cellular dysfunction, genomic instability, senescence, and cell death (Aguilera & Gómez-González, 2008; Lombard et al., 2005; Yousefzadeh et al., 2021). To preserve genome integrity and protect against DNA damage, cells have developed complex and intricate sensing, signaling and repair mechanisms, which are collectively known as the DNA damage response. To date, DNA repair mechanisms have been primarily studied in replicating cells, in which DNA repair is coordinated with cell cycle machinery (Branzei & Foiani, 2008; Hustedt & Durocher, 2017). However, these processes are predicted to differ in neurons, which are terminally differentiated and post-mitotic (Shadfar et al., 2022).

Neurons have high metabolic activity, long lifespans, and are non-replicating; thus, they are particularly vulnerable to DNA damage and strategies to deal with this damage are vital (Welch & Tsai, 2022). Given their high energy demands and sustained rate of oxidative metabolism, neurons are especially susceptible to oxidative DNA damage (Shadfar et al., 2023). DNA damage (or lesion) is an alteration in the chemical structure of the DNA, that may or may not eventually result in a genetic mutation (ie change in the base sequence of the DNA). Accumulation of oxidative DNA damage has been associated with aged brains and individuals with neurodegenerative diseases, including PD ( Z. Alam et al., 1997). DNA lesions from oxidative damage can be repaired through base excision repair (BER), DNA double-strand break (DSB) repair, and in some contexts nucleotide excision repair (NER) (Cadet & Davies, 2017; Cooke et al., 2003). Repair pathways that are dependent upon a sister chromatid, such as homologous recombination (HR), are thought to be mainly absent in neurons (Fielder et al., 2017; Gonzalez-Hunt & Sanders, 2021). However, the DNA damage response in neurons has not been well characterized, and novel and non-canonical DNA repair pathways are being identified (Pollina et al., 2023; Welty et al., 2018).

Dopaminergic neurons within the substantia nigra are especially vulnerable to DNA damage and mutagenesis, during both the aging process and PD. Mitochondrial genome homeostasis in dopaminergic neurons in PD has been a particular focus of studies, in part due to the fact that the mitochondrial genome is highly susceptible to DNA damage given its proximity to the electron transport chain (ETC) (Yakes & Van Houten, 1997). Dopaminergic neurons from the substantia nigra exhibit an accumulation of abasic sites in mitochondrial DNA (mtDNA) in post-mortem PD brain tissue, which was not observed in cortical neurons (Sanders, McCoy, et al., 2014). Nigral neurons from post-mortem brains of patients with early PD or incidental Lewy body disease also have elevated levels of somatic point mutations in mtDNA (Lin et al., 2012). Furthermore, there are increased mtDNA deletions in the substantia nigra of aged human brains and PD patients compared to healthy controls (Bender et al., 2006; Kraytsberg et al., 2006). The precise mechanisms by which damaged or mutated mtDNA damage are generated has not been well characterized, and oxidative stress may be a major contributor. While persistent defects in mtDNA have been well established in PD, mtDNA repair has not been extensively studied in neurons - in particular dopaminergic neurons, and it is unclear whether these mechanisms may be conserved or not with cycling cells (Dölle et al., 2016; Gonzalez-Hunt & Sanders, 2021; Gu et al., 2002; Lin et al., 2012; Qi et al., 2023; Sanders et al., 2017; Sanders, Laganière, et al., 2014; Shadfar et al., 2023; Van Houten et al., 2016).

There is also evidence of nuclear DNA damage in dopaminergic neurons in idiopathic PD. Numerous studies have found oxidative DNA damage and increased levels of γH2AX in PD (Dias et al., 2013; El-Saadi et al., 2022; Guo et al., 2018; Z.-X. Wang et al., 2023). A previous study reported elevated levels of the oxidative DNA lesion, 8-hydroxy-2’-deoxyguanosine (8-OHdG) in PD brains and 8-hydroxyguanine (8-OHG) in the substantia nigra of PD patients compared to controls (Z. Alam et al., 1997). El-Saadi et al. confirmed that genotoxic stress is a prominent feature in human PD patient brains, indicated by increased γH2AX foci in TH-positive neurons in the substantia nigra and microglia compared to the age-matched healthy controls (El-Saadi et al., 2022). Higher numbers of DNA strand-breaks and altered DNA conformation in the substantia nigra of patients with PD were observed compared to controls (Hegde et al., 2006). Together, these studies demonstrate that in addition to mtDNA damage, nuclear genome integrity may also be altered in PD.

The basis for the selective accumulation of both mtDNA and nuclear DNA damage in nigral dopamine neurons is not well understood; however, dopaminergic neurons in the ventral midbrain are exposed to high levels of oxidative stress which may render them particularly vulnerable to DNA damage. Various neuronal processes can drive oxidative stress, including dopamine metabolism, low glutathione (GSH), high levels of iron and calcium in the substantia nigra, and high amounts of polyunsaturated fatty acids that can undergo lipid peroxidation (Dias et al., 2013). All of these neuronal processes coupled with dopaminergic neuronal unique neuroanatomy contribute to their vulnerability to oxidative stress and neurodegeneration (Dias et al., 2013; Guo et al., 2018).

LRRK2 associated PD

While the majority of PD cases are idiopathic with no identifiable cause, about 5-10 % of PD cases are monogenic and are caused by a single highly deleterious pathogenic variant (I. Martin et al., 2011; Pitz et al., 2024). One of the most common monogenic causes of PD are mutations in Leucine-rich repeat kinase 2 (LRRK2). LRRK2 is a large, multi-domain protein of which the PD-linked missense mutations cluster in the enzymatic core consisting of a Ras of complex followed by a C terminus of Roc (ROC-COR domain), guanosine triphosphatase (GTPase) and serine-threonine kinase domain (Berwick et al., 2019; Tolosa et al., 2020). The most prevalent LRRK2 pathogenic mutation is Gly2019Ser (G2019S); however there are over 100 LRRK2 variants linked to PD (Kalogeropulou et al., 2022). Though LRRK2 pathogenic mutations are autosomal dominant, penetrance of LRRK2 mutations are incomplete and influenced by age, environment, genetic and lifestyle factors, such as use of non-steroidal anti-inflammatory medicines and caffeine consumption (Crotty et al., 2020; San Luciano et al., 2020). Significant progress has been made towards expanding our knowledge on LRRK2 biology since the gene identification in 2004 (Taymans et al., 2023). Given its strong genetic link to PD, and the molecular interaction of LRRK2 with various other PD-linked genes, including but not limited to SNCA, GBA, and VPS35, efforts to further define the role of LRRK2 in cellular mechanisms underlying PD are being emphasized (Bieri et al., 2019; Mir et al., 2018; Pang et al., 2022). LRRK2 is an exciting therapeutic target and is the focus of several drug development efforts and on-going clinical trials, for which the field eagerly awaits the outcome of these trials (NCT05348785, NCT03976349, NCT05633745) (Jennings et al., 2023; Taymans et al., 2023).

Despite advancements in therapeutically targeting LRRK2, the normal physiological or pathological function of LRRK2 is still unclear. Clues to pathological functions of LRRK2 have been revealed by consistent and robustly reproduced findings that PD-linked LRRK2 variants drive increased LRRK2 kinase activity, with a subset of Rabs demonstrated to be physiological substrates of LRRK2 (Sheng et al., 2012; Steger et al., 2016; Vides et al., 2022; X. Wang et al., 2021; West et al., 2007). Yet, pathogenic LRRK2 variants have been shown to be involved in diverse cellular processes – from inflammation, vesicle trafficking, lysosome biology and ciliogenesis (Ahmadi Rastegar et al., 2022; Bonet-Ponce et al., 2020; Gomez-Suaga et al., 2014; Khan et al., 2021; Kozina et al., 2022; Ysselstein et al., 2019). There is debate whether LRRK2 is expressed in neurons across different brain regions including the cerebral cortex, striatum, hippocampus and dopamine neurons in the substantia nigra, and/or restricted to microglia and astrocytes (Iseki et al., 2023; Miklossy et al., 2006; Moehle et al., 2012). The notion that LRRK2 levels vary across different brain regions and cell types is suggestive that LRRK2 is tightly regulated and levels may be linked to various cellular processes in different cell types. Notably, recent studies have demonstrated detectable levels of LRRK2 gene expression in dopamingeric neurons in humans and mice (Gaertner et al., 2024; Kamath et al., 2022). While LRRK2 levels in neurons may be low, prior findings indicate LRRK2 may contribute to increased vulnerability in dopaminergic neurons. In line with studies that have shown impaired dopaminergic transmission, Gaertner et al. highlighted that dopaminergic neurons from Lrrk2 G2019S knock-in mice exhibited strong dysregulation of synaptic pathways, supporting the idea that LRRK2 impacts the molecular pathophysiology in dopaminergic neurons (Gaertner et al., 2024; Tozzi et al., 2018; Xenias et al., 2022). Yet, which underlying molecular mechanisms and cellular pathway dysfunction by which LRRK2 contributes to pathology and neurodegeneration in PD and cell type-specificity as it relates to LRRK2 levels are not completely understood and hotly debated (Tolosa et al., 2020). Roles for PD-linked LRRK2 variants are emerging for mitochondrial dysfunction and the regulation of genome integrity and homeostasis (Gonzalez-Hunt & Sanders, 2021; Howlett et al., 2017).

Additionally, some progress has been made towards understanding environmental risk factors of PD and identifying modifiers of LRRK2 penetrance. In vivo murine LRRK2 models have demonstrated increased susceptibility to PD-linked environmental toxicants, including MPTP, paraquat, and rotenone (Arbez et al., 2020; H.-F. Liu et al., 2017; Rudyk et a1., 2019). Exposure to rotenone has been also shown to result in increased LRRK2 kinase activity and cause endolysosomal deficits in rats, similar to those observed in dopaminergic neurons of human post-mortem brain tissue. LRRK2 kinase inhibition prevented the rotenone-induced neurodegeneration in vivo (Rocha et al., 2020). However, lack of mechanistic understanding of the interactions between LRRK2 and PD-linked environmental modifiers persists, and additional work focused on the PD metabolome and exposome may enable further delineation of these molecular interactions.

LRRK2 dependent mitochondrial DNA damage

While the underlying molecular mechanisms associated with LRRK2 that lead to pathology and neurodegeneration in PD are not completely understood, LRRK2 has been established to play an important role in mitochondrial DNA maintenance (Delcambre et al., 2020; Gonzalez-Hunt et al., 2020; Gonzalez-Hunt & Sanders, 2021; Howlett et al., 2017; Ouzren et al., 2019; Pena et al., 2022; Podlesniy et al., 2019; Qi et al., 2023; Sanders, Laganière, et al., 2014; Weindel et al., 2020). PD-linked LRRK2 G2019S causes increased mtDNA damage in PD patient-derived cells and in vitro neuronal models (Gonzalez-Hunt et al., 2020; Gonzalez-Hunt & Sanders, 2021; Howlett et al., 2017; Pena et al., 2024; Qi et al., 2023; Sanders, Laganière, et al., 2014). Interestingly, Lrrk2 G2019S knock-in transgenic mice demonstrate increased mtDNA damage, and LRRK2 deficiency decreased mtDNA lesions in the ventral midbrain compared to wild-type control animals, suggesting that LRRK2 may regulate mitochondrial genome homeostasis (Qi et al., 2023). This increase in mtDNA damage is dependent on LRRK2 kinase activity, and reversed upon LRRK2 kinase inhibition or gene correction (Gonzalez-Hunt et al., 2020; Howlett et al., 2017; Pena et al., 2022; Qi et al., 2023; Sanders, Laganière, et al., 2014). Future research is needed to better understand the underlying mechanisms of mtDNA damage accumulation in the central nervous system, including the localization of mtDNA damage in dopaminergic neurons (cell body vs processes) and how mtDNA damage either causes or is a consequence of other facets of mitochondrial dysfunction, pathology, and neuronal death in PD.

Nuclear genome instability and LRRK2

To better understand the normal or physiological function of LRRK2, LRRK2 deficient murine models have been informative – particularly since LRRK2 knockouts are viable (Hinkle et al., 2012). LRRK2 deficiency has been shown to modulate genomic stability and nuclear morphology in an age-dependent manner in dopaminoceptive striatal spiny projection neurons (SPNs), which express LRRK2 at higher levels than nigral dopaminergic neurons (X. Chen et al., 2020). Specifically, an increase in the percentage of SPNs with 10 or more γH2AX foci, a marker of DNA double-strand breaks, was observed in the nucleus (X. Chen et al., 2020). Alterations in molecular pathways involved in the regulation of nuclear assembly and neuronal excitability were observed in striatal neurons from aged LRRK2 knockout mice compared to age-matched controls (X. Chen et al., 2020). In addition, LRRK2 was shown to be involved in regulating nuclear and soma size in SPNs, with nuclear enlargement observed in LRRK2 deficient mice compared to age-matched controls (X. Chen et al., 2020). Nuclear hypertrophy is associated with increased biosynthetic activities such as DNA synthesis and repair, transcription and translation, and has been observed in pathological conditions (Jevtić et al., 2014; Koda et al., 2006). Similarly, LRRK2 G2019S caused nuclear hypertrophy in SPNs derived from Lrrk2 G2019S KI mice; however, this was not observed with SPNs derived from Lrrk2 R1441C KI mice, which instead exhibited altered nuclear shape (X. Chen et al., 2020). Sepe et al. demonstrated that fibroblasts derived from PD patients carrying LRRK2 G2019S or R1441C mutations had reduced DSB repair capacity, with persistence of γH2AX foci after gamma irradiation (Sepe et al., 2016). The patient-derived fibroblasts also exhibited decreased NER capacity, which was evaluated by assessing unscheduled DNA synthesis (UDS) (Sepe et al., 2016). These studies suggest that LRRK2 may be implicated in DNA repair processes, but the effect of pathogenic LRRK2 variants impact on nuclear DNA integrity and related biosynthetic processes has not been studied specifically in dopamine neurons. A recent study reported that immortalized lymphocytes and peripheral blood-derived lymphocytes from PD patients with LRRK2 G2019S and R1441C mutations exhibited cohesion deficits (Naaldijk et al., 2024). Given the importance of cohesion in mitosis, meiosis, DNA double-strand break repair, and gene expression, the investigation of LRRK2 function in these processes may provide further mechanistic insight into LRRK2’s role in genome integrity and PD neurodegeneration (Brooker & Berkowitz, 2014).

In summary, there is accumulating evidence that DNA damage contributes to PD pathogenesis; however, our understanding of these underlying mechanisms is limited. Further research is needed to understand causes and consequences of DNA damage, DNA damage response mechanisms in dopaminergic neurons and how they impact neuronal function, and the role of LRRK2 in the regulation of genome integrity. Moreover, additional studies are needed to elucidate which mechanisms contribute to degeneration of dopaminergic neurons, brain region specificity, and whether there may be intersection between mitochondrial and nuclear genome maintenance pathways.

DOWNSTREAM OXIDATIVE STRESS

Mitochondria is the major source of ATP in the body, including the brain and ATP is vital for neuronal functions and activity. The findings regarding exposure to mitochondrial complex I inhibitors, such as rotenone and MPP⁺, in rodent-based cultures leading to a significantly higher loss of SNpc dopaminergic neurons compared to dopaminergic neurons in the VTA highlight the differential vulnerability of these neuron populations to these inhibitors ((Jaumotte et al., 2016)). The selective vulnerability of substantia nigra pars compacta (SNpc) dopaminergic neurons has also been demonstrated in in vitro human studies using iPSC-based cell models ((Oosterveen et al., 2021). In vivo studies using 6-OHDA injections have similarly revealed a greater loss of substantia nigra pars compacta (SNpc) neurons compared to those in the ventral tegmental area (VTA) ((Tanguay et al., 2021). These findings collectively suggest that dopaminergic neurons in the SNpc are more susceptible or vulnerable to disruptions in oxidative phosphorylation (OXPHOS) compared to dopaminergic neurons in the VTA. The SNpc dopaminergic neurons have highly branched axonal endings as compared to those in the VTA, resulting in a greater demand for ATP to restore membrane potential after action potentials. Additionally, SNpc neurons exhibit higher expression levels of PGC1-α, a regulator of mitochondrial biogenesis. Furthermore, the basal oxygen consumption rate (OCR) of SNpc neurons is three times higher than that of VTA or olfactory bulb (OB) neurons, although their maximal OCR does not differ. This indicates that SNpc neurons operate at maximal capacity even under normal conditions, whereas VTA and OB neurons do not ((Pacelli et al., 2015)).

The expression of L-type Ca²⁺ channels is upregulated in SNpc dopaminergic neurons, driving pacemaking activity through calcium currents. In contrast, VTA neurons, while also containing L-type Ca²⁺ channels, rely on sodium channels for pacemaking (Striessnig et al., 2006). SNpc neurons heavily depend on Cav 1.3 pore-forming subunits of calcium channels for pacemaking, leading to a significant influx of calcium into the cytosol (Chan et al., 2007). This excess calcium is either pumped back across the plasma membrane or sequestered by the ER or mitochondria (Choi et al., 2006). However, this process overloads the cytosol with calcium, requiring a substantial amount of ATP for calcium extrusion, which in turn leads to ROS formation and oxidative stress. Additionally, the overload of calcium sequestration by mitochondria also contributes to oxidative stress, ultimately triggering apoptosis.

Elevated cytosolic dopamine levels in cultured midbrain neurons measured by intracellular patch electrochemistry have been linked to neurotoxicity. In SN dopaminergic neurons, L-DOPA increases cytosolic dopamine nearly three times higher than in VTA neurons, a process reliant on sensitive calcium channels ((Mosharov et al., 2009)). This renders SN neurons more susceptible to L-DOPA-induced neurotoxicity, associated with increased oxidative stress due to cytosolic dopamine buildup ((L. Chen et al., 2008)). Interestingly, cytosolic dopamine levels remain unchanged by alpha-synuclein (α-syn) deletion, but dopaminergic neurons lacking α-syn show resistance to L-DOPA-induced cell death. These findings highlight the selective vulnerability of dopaminergic neurons to multiple factors, including high cytoplasmic calcium levels and elevated cytosolic dopamine leading to oxidative stress.

IRON SUSCEPTIBILITY

The role of iron is essential in various biochemical processes in the brain, including synthesis of neurotransmitters, myelin, transport of molecules, etc. The substantia nigra is one of the brain regions in which levels of iron have been found to be the highest (Ramos et al., 2014), while its overload or accumulation has been reported in the SNpc of PD patients (Graham et al., 2000; K. R. Li et al., 2022) and elevated intracellular iron levels in nigral dopaminergic neurons (Friedrich et al., 2021). The iron, which is in the form of ferrous iron in the substantia nigra is mostly sequestered by NM in the dopaminergic neurons, or by ferritin proteins in the glial cells of the nigral region (Zecca et al., 2001).

The resultant elevated intracellular iron interacts with α-syn and initiate a cascade of vicious cycle of toxicity that makes the dopaminergic neurons vulnerable. Both ionic forms of iron have the ability to bind to the C-terminus of α-syn, and promote its fibrillar aggregation (Binolfi et al., 2006; Davies et al., 2011; Golts et al., 2002; Jinsmaa et al., 2014; W. Li et al., 2011).

The activated glial cells in the substantia nigra regulate the iron levels through expressing a iron-import protein divalent metal transporter 1 (DMT1) and reducing levels of ferroportin, which accumulates the metal in the microglia (Rathnasamy et al., 2011) . Gliosis occurring in PD conditions results in the ferric ions being free into the cytosol, or extracellular space, making this iron available for oxidation and reduction reactions leading to production of harmful reactive toxic species.

The role of NM as an iron-binding molecule in SNpc dopaminergic neurons has been extensively described in the previous section, implying its crucial role in intraneuronal iron homeostasis. The loss of NM-containing neurons in Parkinson’s Disease (PD) indicates reduced iron binding to NM compared to the normal brain, leading to elevated levels of intraneuronal free iron (Zecca et al., 2001; Zucca et al., 2017) . This increase in free iron levels contributes to oxidative damage and eventual cell death observed in PD.

It is observed that treatment with rotenone in rats and monkeys leads to iron accumulation in SNpc dopaminergic neurons. However, globular iron staining of microglia indicates that the iron content in microglia originates from the phagocytosis of iron-laden dopamine neurons (Mastroberardino et al., 2009) . This phenomenon is also evident in the MPTP mouse model of PD (Salazar et al., 2008) . The iron accumulated in nigral neurons is bound to transferrin (Tf), which also accumulates in nigral dopamine neurons in PD patients and rotenone rat models of PD, primarily localizing to the mitochondria (Mastroberardino et al., 2009) . The generation of ROS leads to the oxidation of Tf, releasing reactive ferrous iron. Normally, in cases of cellular iron overload, iron import mechanisms, like TfR1 are halted or downregulated to regulate import. However, in PD, treatment with rotenone in nigral neurons maintains the expression of transferrin receptor 2 (TfR2), which is another import machinery. This results in the toxic import of iron to mitochondria via the TfR2 pathway, contributing specifically to the cell death of dopaminergic neurons (Mastroberardino et al., 2009).

NEUROANATOMICAL FEATURES THAT RENDER DOPAMINERGIC NEURONS VULNERABLE IN PD

Analyzing brains from individuals diagnosed with clinical PD at different time points after diagnosis and comparing them with the brains of individuals without neurological symptoms has given rise to a hypothesis, according to which, in early PD, Lewy pathology (LP) initially emerges in either the olfactory bulb or the dorsal motor nucleus of the vagus (DMV) in the caudal medulla. These two brain regions have axons that extend to the body surface (Braak et al., 2004). Braak’s hypothesis suggests the retrograde spread of a pathological agent through synaptically connected networks, leading to the development of LP and subsequent neuronal death in clinical PD. While some research strongly supports the Braak model, there are other observations that do not align with this proposition. These conflicting observations argue that there is a significant involvement of internal cell-autonomous factors, emphasizing their crucial role not only in shaping the LP pattern in clinical PD but, more importantly, in determining the specific pattern of neuronal loss that is undeniably linked to symptomatic manifestations. There have been opinions that the spread of LP over time is staged accordingly from particular starting points. According to the Braak hypothesis, the LP evolution is divided into roughly six stages about the spread of LP: the pre-symptomatic stage (1-2), early symptomatic stage (3-4) and late symptomatic stage (5-6) (G. Halliday et al., 2012). This discovery suggests that although the brain regions prone to LP are clearly defined, the sequence and degree to which they exhibit LP are not definitively established.

A study had investigated whether there is neuronal cell death in the patients in the presymptomatic/the Braak stage 1-2 phase of PD. In the brains of these patients, the LP was contained in the medulla and pons, however there was still a 10-20% neuronal loss of ventral tier of SNpc dopaminergic neurons, but not any other regions, even if that region exhibited LP (Milber et al., 2012). Later, only in the early symptomatic phase, almost all dopaminergic neurons of SNpc are lost, and slowly neuronal loss becomes apparent in some other regions of the brain as well, for instance in the cholinergic (but not glutaminergic) neurons of the pedunculopontine nucleus (PPN) (G. M. Halliday et al., 1990). Hence, many studies have shown that neuronal death in clinical PD does not follow the same pattern as LP, and there are other factors as well that are involved in portraying the vulnerability of neurons in PD.

In the brain, dopaminergic neurons are found in the mesencephalon, diencephalon, and the olfactory bulb, which are functionally and anatomically heterogenous in nature. The SNpc is considered to be a “harsh” region of the brain because the dopaminergic neurons in this region are rich in dopamine containing high levels of iron and redox available neuromelanin (Chinta & Andersen, 2005). This dopaminergic cell group lies in the mesencephalic system, which is further divided into nigrostriatal system, where the cell bodies from SNpc project into the dorsal striatum. The more medial to this is the mesocorticolimbic system, which arise from the dopaminergic neurons of the VTA (Wise, 2004). It is to be noted that the various clusters of the dopaminergic neurons have different kinds of projections and anatomical positions that allow them to play different vital roles and cellular functions. The only thing that all these clusters have in common is the synthesis and storage of the neurotransmitter DA. Apart from that, there are massive differences in their cell numbers, branching patterns, and phenotypes.

A detailed study of pathological features of a human brain with PD has shown that the terminals of the dopaminergic neurons are lost prior to the cell bodies (Kordower et al., 2013). This has concluded that the terminals of the dopaminergic neurons are more vulnerable as compared to the cell bodies and the neurodegeneration begins at the DA terminals (Garcia-Reitböck et al., 2010; Volpicelli-Daley et al., 2011). Moreover, the SNpc neurons are known to have highly branched axons of considerable length that are either poorly or altogether not myelinated. This is so that a single dopaminergic neuron of the SNpc has over 100,000 synapse endings in the striatum and can extend up to 40cm in length (Matsuda et al., 2009). The dopaminergic neurons lying in the VTA region on the other hand don’t have as much of extensive branching as the substantia nigra (Loughlin & Fallon, 1984). The axonal domain structure also contributes to the increased vulnerability of SNpc neurons. A study comparing vulnerable and resilient neurons in PD demonstrated that vulnerable neurons, characterized by longer length and more complex axonal arborization, were more susceptible to oxidative stress induced by hydrogen peroxide. Additionally, these vulnerable neurons exhibited elevated levels of synaptogamin-1 (Syt-1) (Burke & Trudeau, 2022).

The maintenance of such a huge terminal field, with an unmyelinated axon no less is bioenergetically expensive, and most likely to create a proteostatic as well as a metabolic burden on the cell body of the dopaminergic neuron (Bolam & Pissadaki, 2012). Additionally, the trafficking of newly synthesized proteins, organelles, and other cellular components from the soma to the axonal endings back and forth demands a physical maintenance that adds to the vulnerability of the dopaminergic neuron. All of the endosomal trafficking in the neuron requires ATP and funneling it through one single unmyelinated axon adds to the susceptibility (Ashrafi et al., 2014). A study has shown that the expression of motor proteins required for axonal trafficking is reduced in the SNpc of early PD patients (Chu et al., 2012). Hence, the axonal transport of mitochondria is also greatly reduced due to spatial limitations across the axonal tree. Another factor is that the SNpc DA neurons have axons that possess at least 200,000 vesicular release sites, which is likely to extraordinarily increase the expression of α-syn, which is a major synaptic protein making it vulnerable to α-syn pathology (Matsuda et al., 2009; Zharikov et al., 2015). This leads to the notion that the vast DA release sites of the dopaminergic axons of the nigrostriatal pathway also are rendered vulnerable due to the variations in synaptic transmission and vesicular trafficking.

A recent study has screened genes to identify the dopaminergic neuron loss during PD in an attempt to find out why the dopaminergic neurons are particularly vulnerable to PD. The study identified top-gene candidates with single-nucleotide polymorphisms (SNPs) in D. melanogaster that render dopaminergic neurons susceptible (J. Davis et al., 2021). In the top-associated gene list, genes associated with neuroanatomy alterations were identified that made dopaminergic neurons vulnerable. Examples of the genes are tow (target of wingless), that is crucial for axon targeting in Drosophila, and Trf2 (TATA box binding protein-related factor 2) which regulates neurite formation.

The role of transcription factors is well-known in SNpc neurons for differentiation and maintenance throughout the development stages of the brain. SNpc neurons are well-defined by the combination of transcription factors that play a role in development. A recent study has interestingly revealed that undifferentiated, yet committed cells giving rise to SNpc neurons exhibit higher baseline OXPHOS and ATP levels compared to developmentally comparable and isogenic forebrain cells. This suggests that SNpc cells are inherently programmed by transcription factors to a higher energy state even before differentiation, contributing to the morphological characteristics unique to SNpc neurons (Bell et al., 2021).

There is a need to translate the above findings on devising a novel therapeutic strategy to slow down or hinder the PD pathology and progression. However, making changes to a neurotransmitter phenotype of the DA neurons or to modify its axonal branching is not a viable option, although therapeutic strategies to amend the downstream effects of these factors could be considered, once they are well-defined.

SUMMARY AND FUTURE WORK

PD is a complex disease, with widespread brain and systemic pathology that underlie both the motor and nonmotor phenotypes. Despite the breadth of pathology, the experimental studies have longed focused on dopaminergic neurons of the substantia nigra due to the severity of the pathology and that such pathology produces much of the motor phenotype. Dopaminergic neurons clearly do not die in isolation; many extracellular factors and cell-cell interactions (both neuronal and non-neuronal) are critical to pathogenesis. However, there are clearly intracellular factors that drive cell-type sensitivity to both endogenous and exogenous insults (Figure 1). These factors are especially critical in neuronal subtype sensitivity to environmental toxicants. Biochemical, molecular, and anatomical features have been identified in humans that drive such sensitivity (Figure 1). It is critical to consider such factors in drawing conclusions on primary mechanisms of pathogenesis, environmentally induced dopaminergic neurotoxicity, and the design of laboratory studies. Here, model systems that replicate human relevant dopaminergic neurobiology are needed to increase the predictive translational value.

Highlights.

• Inherent neurobiology of nigral dopamine neurons increases toxicant vulnerability

• Dopamine neurobiology and toxicant interactions are primary toxicity mechanisms

• Translational Parkinson’s disease models must replicate human neurobiology