Perspective: Low Risk of Parkinson's Disease in Quasi-Vegan Cultures May Reflect GCN2-Mediated Upregulation of Parkin

ABSTRACT

Mitochondrial dysfunction in dopaminergic neurons of the substantia nigra (SN) appears to be a key mediating feature of Parkinson's disease (PD), a complex neurodegenerative disorder of still unknown etiology. Parkin is an E3 ubiquitin ligase that promotes mitophagy of damaged depolarized mitochondria while also boosting mitochondrial biogenesis—thereby helping to maintain efficient mitochondrial function. Boosting Parkin expression in the SN with viral vectors is protective in multiple rodent models of PD. Conversely, homozygosity for inactivating mutations of Parkin results in early-onset PD. Moderate protein plant-based diets relatively low in certain essential amino acids have the potential to boost Parkin expression by activating the kinase GCN2, which in turn boosts the expression of ATF4, a factor that drives transcription of the Parkin gene. Protein-restricted diets also upregulate the expression of PINK1, a protein that binds to the outer membrane of depolarized mitochondria and then recruits and activates Parkin. This effect of protein restriction is mediated by the downregulation of the kinase activity of mammalian target of rapamycin complex 1; the latter suppresses PINK1 expression at the transcriptional level. During the 20th century, cultures in East Asia and sub-Sahara Africa consuming quasi-vegan diets were found to be at notably decreased risk of PD compared with the USA or Europe. It is proposed that such diets may provide protection from PD by boosting Parkin and PINK1 expression in the SN. Other measures that might be expected to upregulate protective mitophagy include supplemental N-acetylcysteine (precursor for hydrogen sulfide) and a diet rich in spermidine—a polyamine notably high in corn.

Introduction

Parkinson's disease in a nutshell

The second most common neurodegenerative disorder following Alzheimer's disease, Parkinson's disease (PD), affects 1–2 per 1000 of the population; its prevalence increases with age, affecting ∼1% of the population aged above 60 y (1). Males are more often affected than females, with a ratio of ∼3:2, and it is less prevalent in those of African or Asian ancestry. PD is classified as a movement disorder in which the most prominent signs are bradykinesia, tremor, rigidity, and postural instability. However, nonmotor symptoms such as depression, anxiety, hyposmia, sleep disorders, cognitive impairments, and constipation are increasingly gaining attention, and are now included among the clinical diagnostic supportive criteria (2). They can typically precede the motoric dysfunctions, sometimes by several years, but the neuroanatomical pathways responsible for these symptoms remain to be elucidated (3).

Pathophysiologically, PD is characterized by disruptions to protein folding attributable to various factors such as mutations, errors in transcription/translation, yet-unidentified environmental factors, or age-related decline, which leads to failures in protein quality control (proteostasis) in susceptible dopaminergic neurons (4). This results in aggregation of misfolded proteins and their organization into larger deposited structures—a hallmark process in neurodegenerative conditions in general and in PD in particular. PD is characterized by the presence in the substantia nigra (SN) of Lewy bodies (LBs), protein aggregates whose primary constituent is misfolded α-synuclein (ASN), a protein of 140 amino acids. The accumulation of these protease-resistant ASN-rich LBs is associated with a marked loss in dopaminergic neurons of the SN, leading to reduced facilitation of voluntary movements, manifested as movement abnormality.

Several susceptibility genes and environmental factors are associated with increased risk of PD. In an umbrella review of meta-analyses examining environmental influences on PD risk, the only clearly established associations were with constipation and physical activity (5). Notably, many other factors showed suggestive links, including microbial or viral infections (6–8) and nutritional deficiencies (9). Recently, food-based strategies for preventing or controlling PD have been suggested, such as ingestion of phospholipid membrane precursors and microbiota-directed therapy (10). Vegetarian or vegan diets have been linked to decreased risk of numerous chronic conditions, including ischemic heart disease, cardiovascular disease, obesity, hyperlipidemia, diabetes, metabolic syndrome, diverticular disease, cancer, and eye cataract (11–14). However, the possible impact of such diets on PD risk has so far received minimal attention.

Quasi-vegan diet

Quasi-vegan diets have also been referred to as flexitarian, meat-reduced, semi- or demi- or partial vegetarian and reducetarian diets (15, 16). These terms are often used interchangeably in the nutritional, clinical, or scientific literature. Quasi-vegan diets refer to diets that, although primarily plant-based, may include the occasional consumption of poultry, fish, eggs, and dairy products—but typically avoid red meat. The main menu of quasi-vegans includes vegetables, fruits, whole grains, legumes, seeds, nuts, eggs, and dairy foods, with occasional poultry and fish (17). The people who avoid animal products other than fish are called pesco-vegetarians or pescatarians. Lacto-ovo-vegetarian diets allow the consumption of dairy products and eggs but avoid flesh foods. The reasons that people adopt a quasi-vegan diet include purported health benefits, religious convictions (Hinduism, Buddhism, and Jainism), animal welfare concerns prompted by inhumane housing and slaughtering of animals, ecological concerns owing to the high carbon footprint and extravagant use of limited resources associated with meat production, and household budgetary concerns reflecting the relatively high cost of meat. Traditional diets in some less-wealthy societies have been quasi-vegan simply owing to the relatively high cost of animal rearing.

Since meat is a rich source of proteins, quasi-vegan consumers need to take into consideration their daily protein intake, to ensure that they fulfill the RDA of 0.8 g of protein per kilogram of weight (18). For those quasi-vegetarians who avoid meat or fish, adequate protein intakes (10–35% of total daily calories) can be achieved from numerous sources: green peas, chickpeas, nuts, beans, quinoa, grains, tempeh and tofu, leafy greens, sunflower, sesame and poppy seeds, eggs, milk, and dairy products. Fruits, alcoholic beverages, added oils, and refined sugars and starches are very low in protein, and should not be the predominant components of a quasi-vegan diet. Those eating wholly or primarily plant-based diets need to supplement with vitamin B-12 and might also be well advised to supplement with the long-chain ω-3 fatty acids found in marine fish.

In 2001, McCarty, making note of pertinent ecologic epidemiology, suggested that the consumption of a quasi-vegan diet might help to prevent or control PD by slowing the loss of dopaminergic neurons, and might aid the efficacy of L-dopa therapy by minimizing the impact of variations of plasma amino acid concentrations on the blood-brain barrier transport of L-dopa (19). Notably, the impacts of quasi-vegan diets on neurodegenerative disease in general, and PD in particular, have received minimal consideration. As summarized below, some geoepidemiological data points to a notably lower age-adjusted risk of PD in societies in East Asia and sub-Sahara Africa whose traditional diets are quasi-vegan (20, 21). In the present communication, we are proposing a mechanism whereby the long-term consumption of quasi-vegan diets might indeed lessen PD risk.

Mitochondrial dysfunction in PD

Despite the unidentified etiology of sporadic PD, the crucial pathophysiological place of mitochondrial dysfunction in this disorder is well established and was recently summarized (22). This is 1 of 3 major cellular pathways that are affected in PD, the other 2 being proteostasis and oxidative/nitroxidative stress; but all 3 are interconnected, reinforcing each other in multiple ways (23). Several endogenous and exogenous factors can inhibit mitochondrial functional integrity: heavy metals, agricultural pesticides and herbicides, paraquat, rotenone, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine, high concentrations of nitric oxide, the dopamine metabolite aminochrome, and others have also been reported. With respect to familial PD, this has been linked to genetic polymorphisms of a number of genes coding for proteins that influence mitochondrial structure and function. These genes include ATP13A2, PINK1, CHCHD2, LRRK2, SNCA,GBA and also Parkin, a key focus of this review (24).

Polymerized ASN, the hallmark of PD, induces mitochondrial dysfunction. By binding to mitochondrial outer membrane proteins such as voltage-dependent anion-selective channel 1 (VDAC1), translocase of outer membrane (TOM) 40 and TOM20, it mediates mitochondrial dysfunction (25). Oxidative stress is tightly related to PD mitochondrial dysfunction. The accumulation of unfolded ASN in mitochondria inhibits complex I activity, thereby driving mitochondrial reactive oxygen species production (26). The accumulation of iron within the SN enhances oxidative stress and promotes local ASN aggregation, thereby exacerbating mitochondrial dysfunction (27).

The mitochondria are extremely dynamic intracellular organelles; a proper balance between mitochondrial biogenesis and mitophagy ensures that cells have adequate numbers of functionally efficient mitochondria. The failure of mitochondrial quality control mechanisms is an important contribution to PD development. Dysfunction of the mitochondrial electron transport chain plays a central role in the pathogenesis of PD and is associated with oxidative stress and alterations to the mitochondrial genome (22).

The role of impaired mitophagy in the genesis of PD deserves some emphasis. Malfunctions of Parkin that are inherited (in some familial PD), or acquired, result in excessive expression of Drp1, which when phosphorylated and translocated to mitochondria, catalyzes mitochondrial fission (28). Excessive concentrations of nitric oxide can cause nitrosylation of Parkin, which suppresses its ubiquitin ligase activity, boosting Drp1 concentrations, and resulting in mitochondrial hyperfragmentation (29). It seems that all of the above PD-associated mitochondrial abnormalities interact and are crossconnected, hence, establishing the pathogenic basis of PD synucleinopathy.

Parkin Protects against PD by Preserving Efficient Mitochondrial Function

Individuals who are homozygous for loss of function of the E3 ubiquitin ligase Parkin develop early-onset PD (30, 31). This may reflect a key role for Parkin in promoting mitophagy of depolarized mitochondria, and in stimulating mitochondrial biogenesis (32, 33). PINK1 binds to the outer membrane of depolarized mitochondria and recruits Parkin, which promotes proteasomal degradation of outer membrane proteins that tether mitochondria to other structures, and mark the outer membrane with polyubiquitin chains that promote incorporation of mitochondria into developing autophagosomes (32). Concurrently, Parkin promotes mitochondrial biogenesis by inducing proteasomal degradation of PARIS, a protein that represses the transcription of PGC1α, a key driver of mitochondrial biogenesis (34). Nitrosylation of Parkin suppresses its activity, whereas persulfidation boosts it—phenomena thought to contribute to the pathogenic role of excess nitric oxide, and the protective role of hydrogen sulfide (H₂S), respectively, in PD (35, 36). The maintenance of efficient mitochondrial bioenergetics is highly important to dopaminergic neurons in the SN, as their long, highly arborized axons make over a million synaptic connections; maintaining the proper function of these connections imposes a heavy energy burden that can only be met by rapid mitochondrial ATP generation (37). It is notable that mutations of other proteins required for efficient mitophagy—such as PINK1 and LRRK2—have likewise been linked to early-onset PD (33).

The overexpression of Parkin in the SN via lentiviral vectors protects mice from PD induced with 6-hydroxy-dopamine, MTPT, or mutant ASN (38–40). This suggests that more clinically practical measures for enhancing striatal Parkin expression might aid prevention or control of PD. The ATF4 transcription factor has been shown to promote transcription of the Parkin gene (41, 42). The protein expression of ATF4 is increased when, under certain stress conditions, eIF2α kinase is activated, leading to phosphorylation of eIF2α that selectively boosts translation of ATF4 mRNA (43). Intriguingly, drugs that amplify the activation of this kinase—salubrinal and guanabenz—have been shown to be protective in rodent models of PD (44, 45).

Parkin Expression Can be Boosted by Dietary Restriction of Essential Amino Acids

GCN2 is a kinase that very selectively phosphorylates and thereby activates eIF2α kinase (46). GCN2, in turn, functions as a detector of essential amino acid deficiency; uncharged tRNAs that accumulate when cellular concentrations of 1 or more essential amino acids are low can bind to and directly activate GCN2 (46–48). Protein-restricted diets in mice lead to an increase in hepatic eIF2α phosphorylation and a marked increase in hepatic production of fibroblast growth factor-21 (FGF21), whose transcription is driven by ATF4; this effect is absent in GCN2 knockout mice (49). Upregulation of FGF21, by reducing the responsiveness of hepatocytes to growth hormone, may be responsible for the decrease in hepatic IGF-I production associated with low-protein diets (50, 51). Protein-restricted diets in humans have likewise been shown to increase plasma FGF21 and decrease plasma insulin-like growth factor (IGF-I) (49, 51). A recent cross-sectional study found that plasma concentrations of FGF21 tend to be about 3-fold higher in vegans than in omnivores (P < 0.01) (52). Vegan diets of modest protein content tend to be relatively low in certain essential amino acids, notably methionine and lysine (53, 54). Furthermore, vegans are known to have relatively low plasma IGF-I concentrations (54, 55).

Low-Protein Diets Also Promote Efficient Mitophagy by Downregulating Mammalian Target of Rapamycin Complex 1 Activity

Amino acid status regulates not only GCN2 activity, but also that of the crucial regulatory kinase mammalian target of rapamycin complex 1 (mTORC1). Increased cellular concentrations of leucine, arginine, and of the methionine metabolite S-adenosylmethionine boost mTORC1 activity by suppressing mechanisms that turn off this activity (56–62). Crucially, mTORC1 acts to inhibit the expression of PINK1 at the transcriptional level; this effect might be mediated, in part, by transcriptional repression of FOXO1a, which binds to the promoter of the PINK1 gene and promotes its transcription (63–65). Since PINK1 is required for the recruitment of Parkin to damaged depolarized mitochondria, mTORC1 functions to suppress mitophagy (63–65). Conversely, low-protein diets can be expected to boost mitophagy by the concurrent activation of GCN2 and deactivation of mTORC1. As might be expected, certain genetic variants of PINK1 have been linked to autosomal recessive early-onset PD (66).

Quasi-Vegan Diets Have Been Associated with Lower Parkinson's Risk

During the 20th century, epidemiological analyses found that age-adjusted PD rates in Europe and the Americas tended to relatively uniform, whereas those in sub-Sahara Africa, rural China, and Japan tended to be notably lower (18, 20, 21). Investigations conducted by Schoenberg and colleagues comparing the age-adjusted prevalence of PD in the USA, rural China, and Nigeria found that this prevalence was notably lower in the latter 2 (67). Although more recent studies suggest that the prevalence of PD in black Americans is lower than that of whites or Asians, Schoenberg's door-to-door community-based studies, using identical assessment methods, found that age-adjusted PD prevalence was 5 times higher among blacks in Mississippi than those in Nigeria (67–70). These observations suggest that environmental factors may be largely responsible for the observed discrepancies in risk. Although black Americans often carry genes of non-African origin, it seems unlikely that this alone could account for disparities in PD risk of this magnitude. It is notable that the traditional diets of the low-risk areas tend to be relatively low in protein and are quasi-vegan. We, therefore, suggest that the relatively low risk of PD observed in sub-Saharan Africanand East Asian cultures may reflect, in part, a modest intake of protein and essential amino acids. This could be expected to decrease PD risk by upregulating the striatal production of Parkin via increased GCN2 and ATF4 activity, and by upregulating PINK1 synthesis via the downregulation of mTORC1 activity—thereby boosting mitophagy. We do not rule out the possibility that a lower daily calorie intake, increased physical activity, or other factors might also contribute to lower PD risk in these cultures.

Conceivably, vegan diets might also confer protection from PD owing to a decreased content of fat-soluble chemical contaminants that tend to accumulate in animal fat—some of which are potentially toxic to dopaminergic neurons (71–73). Moreover, whole-food plant-based diets tend to be rich in phytochemicals, many of which have phase 2-inductive activity, increasing the expression of various antioxidant enzymes and boosting glutathione synthesis (71, 74, 75). Prospective epidemiology has linked increased dietary flavonoid intakes to a lower risk of PD (75).

The impact of protein restriction on animal models of PD appears to have received little study. Protein-restricted or plant-based diets have been studied clinically in short-term studies as a strategy for improving the response to levodopa therapy – as postprandial increases in plasma branched-chain amino acids competitively inhibit the transport of levodopa through the blood-brain barrier—but not as a measure for preventing or slowing the progression of PD (76–78). A handful of anecdotal reports in journals or the internet are consistent with the possibility that plant-based diets might slow PD progression (19, 79). It would be of interest to determine how low-protein diets or diets restricted in single essential amino acids influence Parkin and PINK1 expression in the striatum of rodents, and whether such diets might be protective in rodent PD models.

Adjunctive Measures for Upregulating Mitophagy—Spermidine and Dietary Corn

Upregulation of mitophagy may be viewed as a key strategy for the prevention of PD. Both autophagy and mitophagy can be promoted by various measures, which prevent the lysine acetylation of certain proteins that are key mediators of autophagy. In particular, agents that prevent or reverse the activity of the histone acetylase EP300—dubbed “caloric restriction mimetics”—have been shown to have this effect (80–85). Of particular interest in this regard is the polyamine spermidine, which competitively inhibits the binding of EP300 to its substrate acetyl-CoA (85). Dietary spermidine is well absorbed, and spermidine-enriched diets have been shown to promote both autophagy and mitophagy in rats and other organisms (86–89). Moreover, the administration of spermidine has been shown to protect rats from rotenone-induced PD, and nematodes and fruit flies from a Parkinsonian syndrome induced by ASN overexpression (90, 91). The cerebrospinal fluid concentrations of spermidine in PD patients were found to be lower than those of controls (92). Commercial spermidine nutraceuticals are not yet available, but certain foods are relatively rich in this compound. The food which stands out most in this regard is corn; a cup (250 g) of canned corn contains about 60 mg of spermidine (whereas the median daily intake of spermidine in American postmenopausal women has been estimated to be ∼10 mg) (93, 94). A search for correlations between corn consumption and PD risk revealed a Japanese study in which corn-raising regions in Japan were found to be at decreased risk of PD (95, 96). Perhaps a vegan diet high in corn might be of particular merit for PD protection; the notably low tryptophan and lysine content of the chief corn protein, zein, may aid GCN2 activation in such diets (97). Consistent with a role for autophagy/mitophagy in the promotion of lifespan/health span, dietary spermidine intake has been found to correlate inversely with total mortality in a prospective epidemiological investigation (98, 99).

Although it is an amino acid, cysteine may upregulate mitophagy by serving as a substrate for the production of H₂S (100). Via persulfidation, this boosts the activity of USP8, a deubiquitinase that aids mitophagy by removing K6-linked polyubiquitin chains from Parkin; for reasons that remain unclear, the self-induced polyubiquitination of Parkin hinders mitophagy (101, 102). Furthermore, as noted above, persulfidation of Parkin itself boosts its catalytic activity (36). H₂S can also promote both autophagy and mitophagy by activating AMP-activated kinase (AMPK) via calmodulin-activated kinase kinase-β; how H₂S activates the latter has not been determined (103, 104). AMPK promotes macroautophagy by conferring inhibitory phosphorylation on mTORC1 as well as activating phosphorylation on ULK1, a key initial mediator in the initial formation of autophagosomes (105, 106). Not surprisingly, supplemental N-acetylcysteine has been found to be protective in rodent models of PD (107, 108).

However, increased dietary cysteine also has the potential to decrease mitophagy by diminishing the need for mitophagy. Cysteine is the rate-limiting substrate for glutathione synthesis; intramitochondrial glutathione provides antioxidant protection for the mitochondrial respiratory chain, and hence may tend to prevent the decline in mitochondrial potential that triggers mitophagy (109). Glutathione concentrations in the SN of PD patients have been found to be low relative to control concentrations (110, 111). Especially in the elderly, in whom tissue concentrations of both cysteine and glutathione tend to decline, supplementation of low-protein diets with N-acetylcysteine may be advisable, for the promotion of both glutathione and H₂S synthesis (112, 113).

Broader Health Ramifications of Mitophagy Upregulation

A more general thesis that might be derived from these considerations is that vegan diets of modest protein content may help to minimize oxidative stress of mitochondrial origin by upregulating mitophagy. Hence, such diets may be protective in the range of disorders in which mitochondrial oxidative stress plays a pathogenic role—PD being a notable case in point. Intriguingly, Parkin can protect the heart from ischemia-reperfusion damage by promoting ubiquitination and degradation of cyclophilin-D; the latter is required for the formation of mitochondrial permeability transition pores, which precipitates cardiomyocyte necrosis during cardiac ischemia-reperfusion injury (114). Moreover, studies employing the phosphatase inhibitor salubrinal, which upregulates the phosphorylation of eIF2α, find that this drug provides protection in cardiac and cerebral ischemia-reperfusion, cardiac hypertrophy, hypoxic pulmonary hypertension, and brain trauma (115–123); the extent to which ATF4 and/or Parkin contribute to these benefits remains to be evaluated. In any case, these results suggest that moderately protein vegan diets, via activation of GCN2 and consequent phosphorylation of eIF2α, may be protective in these syndromes.

ATF4 also promotes transcription of the gene coding for nrf2, the mediator of phase 2 induction of cytoprotective/antioxidant enzymes—suggesting an additional mechanism whereby low-protein vegan diets may aid control of oxidative stress (124, 125). Indeed, a low-protein diet has been shown to increase the expression of nrf2 mRNA in the monocytes of patients with chronic renal disease (126). This effect could be expected to interact synergistically with the administration of clinically effective phase 2-inductive nutraceuticals, such as lipoic or ferulic acids (127, 128).

Conclusions

Efficient mitochondrial function is of great importance for the proper function and survival of dopaminergic neurons of the SN, which, owing to their vastly ramified neural connections, have an extremely high requirement for mitochondrially generated ATP. A failure of well-coordinated mitophagy and compensatory mitochondrial biogenesis in these neurons during the evolution of PD results in the accumulation of mitochondria that are dysfunctional, hyperfragmented, and a major source of oxidative stress. The proteins Parkin and PINK1 are key mediators of mitophagy and mitochondrial biogenesis, and upregulation of their expression is notably protective in rodent models of PD. Diets relatively low in protein and certain essential amino acids tend to boost the activity of GCN2—leading to an ATF4-dependent transcriptional upregulation of Parkin expression—and a decline in mTORC1 activity, resulting in enhanced transcription of the gene coding for PINK1. These phenomena may rationalize, at least in part, epidemiology pointing to relatively low age-adjusted risk of PD in African and East Asian populations whose traditional diets have been quasi-vegan. The influence of quantity and quality of dietary protein on SN expression of Parkin/PINK1 in rodents, and on PD risk in rodent models of PD, should be examined. Other dietary measures that upregulate mitophagy/autophagy, such as increased dietary spermidine, may also have potential for PD prevention; further epidemiology should evaluate a possible correlation between dietary spermidine content—or corn consumption—and PD risk.

Notes

The authors reported no funding received for this study.

Author disclosures: The authors report no conflicts of interest.

Perspective articles allow authors to take a position on a topic of current major importance or controversy in the field of nutrition. As such, these articles could include statements based on author opinions or point of view. Opinions expressed in Perspective articles are those of the author and are not attributable to the funder(s) or the sponsor(s) or the publisher, Editor, or Editorial Board of Advances in Nutrition. Individuals with different positions on the topic of a Perspective are invited to submit their comments in the form of a Perspectives article or in a Letter to the Editor.

AL is retired.

Abbreviations used: AMPK, AMP-activated kinase; ASN, α-synuclein; FGF21, fibroblast growth factor-21; H₂S, hydrogen sulfide; LB, Lewy bodies; mTORC1, mammalian target of rapamycin complex 1; PD, Parkinson's disease; SN, substantia nigra; TOM, translocase of outer membrane.