Research investigating pesticides and autophagy is still accumulating, however, several pesticides have been shown to inuence the process21. For example, the cotton cluster contains sodium cacodylate, an organic arsenic compound used as an herbicide. Arsenic-containing pesticides have previously been related to PD14,22. Furthermore, studies have demonstrated arsenic promotes protein aggregation23, can induce the accumulation of alpha-synuclein24, and inhibits autophagic ux25. We have also previously shown using the Comparative Toxicogenomics Database that the chemicalgene network linked to sodium cacodylate is enriched for autophagy-related gene sets determined through gene ontology26. Beyond arsenic-containing pesticides, several of the cotton cluster pesticides have been linked to mitochondrial dysfunction and elevated reactive oxygen species, which can induce autophagy. Triuralin, for instance, was found to reduce the spare capacity of mitochondria in PD-patient-derived dopamine

Prometryn exposure led to mitochondrial and proteasome dysfunction27, while phorate induced oxidative stress and DNA damage 28.

Therefore, as some pesticides may alter autophagy, a 2-hit model to the lysosomal system where exposure would add to the genetic variants impact is plausible2933. A similar gene-environment interaction has been seen with aldehyde dehydrogenase gene (ALDH) variants and ALDH-inhibiting pesticides34. Dysfunction of autophagy has also been implicated in other neurodegenerative diseases such as Alzheimers disease35, and the added toxicity from pesticide exposure coupled with a potent genetic variant affecting autophagy might very well cause a Parkinsonian disorder. Supporting this, other genes with multiple variants scoring highly on disease severity and pesticide exposure observed in this study (Table 2, Supplementary Table 5) also show links to autophagy, including HTT36, LAMP137, npj Parkinsons Disease | (2024) 10:87 Article and MAPT38, or mitophagy, such as FBXO739. Another gene with multiple observed enriched variants, BAG6, modies autophagy via modulation of EP300 40.

Only three variants (3/36, 8%) were observed to be enriched in an independent PD cohort from the Parkinsons Progression Marker Initiative (PPMI), likely without pesticide exposure comparable to the PEG cohort, suggesting that the majority of variants we identied (11/14 variants, 79% of variants seen in both PEG and PPMI cohorts) contribute most to PD risk when individuals are pesticide exposed (Supplementary Table 5). It is also notable that 2 of the PEG variants also enriched in the PPMI cohort (in the GALC and TSC1 genes) were observed in multiple PEG subjects (Supplementary Table 5). This may reect that some variants exhibit a baseline PD risk that can be augmented by pesticide exposure. This is further supported by the observation of 10 prioritized variants (10/36, 28%) being observed in genes with a known association to PD risk. GAK, HIP1R, and MAPT each had 2 variants apiece (2/10, 20%). GBA1, the most common PD risk gene, had one variant (1/10, 10%), as did ACMSD, LAMP3, and M