{"paper_id":"346a1a4c-2b60-414e-8ba7-a0993f693d60","body_text":"1 \nTitle: Aggregation-prone alpha -synuclein proteoforms and dysregulated \nmolecular signatures in the vermiform appendix of synucleinopathy patients \n \nAuthors: Ehraz Anis1☨, Saima Zameer1☨, Joshua Wierenga1, Peipei Li1☨, Jacek W. Sikora2, Juozas \nGordevicius1,3, Meghan Schilthuis1, Richard D . LeDuc2,4, Jeffery H. Kordower1,5, Michelle \nPinho6, Sandra Pritzkow6, Claudio Soto6, Patrik Brundin1,7, Lena Brundin1☨*, Killinger BA1,8☨ \nAffiliations: \n1Center for Neurodegenerative Science, Van Andel Research Institute ; Grand Rapids, MI, USA  \n2Proteomics Center of Excellence, Northwestern University; Evanston, IL, USA \n \n3Vugene LLC; Grand Rapids, MI, USA \n \n4Department of Pharmacology-Physiology, Faculty of Medicine and Health Sciences, Institut de  \nPharmacologie de Sherbrooke (IPS), Université de Sherbrooke; Sherbrooke, Québec, Canada \n \n5ASU-Banner Neurodegenerative Disease Research Center, Arizona State University; Tucson, \nArizona, USA \n \n6Mitchell Center for Alzheimer’s Disease and Related Brain Disorders, Department of Neurology, \nMcGovern Medical School, The University of Texas Health Science Center at Houston; Houston, \nTexas, USA \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n2 \n2 \n7Roche Pharma Research and Early Development (pRED), Neuroscience and Rare Diseases, \nRoche Innovation Center Basel; Basel, Switzerland \n \n8Department of Neurological Sciences , Division of Movement Disorders, Rush University;  \nChicago, IL, USA \n \n*Corresponding author. Email: lena.brundin@vai.org \n☨These authors contributed equally to this work \nOne Sentence Summary : Appendixes from synucleinopathy patients show altered gene \nexpression, unique α-syn proteoforms, and higher aggregation propensity than substantia nigra. \nAbstract: Synucleinopathies, including Parkinson’s disease , are neurodegenerative diseases \ncharacterized by intracellular inclusions containing the amyloidogenic protein alpha -synuclein. \nWhile classically considered to be brain disorders, increasing evidence suggest s involvement of \nthe gut, with alpha-synuclein aggregates potentially propagating to the brain via the vagus nerve. \nEvidence also suggests that the vermiform appendix is particularly susceptible to alpha-synuclein \naggregation, and appendectomy impacts the onset of P arkinson’s disease . However, the \nmechanisms underlying the aggregation of alpha-synuclein in the vermiform appendix remains \npoorly understood. To explore this, we assessed aggregation properties in postmortem appendix \ntissues from healthy controls and synucleinopathy patients using the alpha -synuclein seed \namplification assay (alpha -synuclein-SAA) and performed total RNA sequencing alongside \ndifferential bisulfite-hybridization-based DNA methylation analysis in the same tissues  to \ninvestigate the  molecular underpinnings. Moreover, we determined alpha -synuclein cleavage \npatterns by cataloging soluble alpha-synuclein proteoforms from postmortem substantia nigra and \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n3 \n3 \npost-surgical appendix tissues using top -down mass spectrometry (TD -MS). Alpha-synuclein-\nSAA was positive in appendix samples for 68.75% of synucleinopathy patients and 6.6% of \ncontrols. Genomic profiling  revealed dysregulated expression of genes linked to protein \nfolding/degradation, immune/inflammatory responses , and ciliary dynamics in synucleinopathy \nappendix tissues. TD-MS identified 65 distinct alpha-synuclein proteoforms in the substantia nigra \nand appendix, with 9 unique to the appendix. Further, in silico  modeling revealed higher \naggregation propensity of alpha -synuclein proteoforms in the appendix versus substantia nigra.  \nTogether, our findings suggest that a tissue environment of alpha-synuclein dysproteostasis in the \nappendix has the potential to contribute to the development of synucleinopathies.  \n \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n4 \n4 \nMain Text: \nINTRODUCTION \nSynucleinopathies, including Lewy body diseases such as Parkinson’s disease (PD) , are \ncommon neurodegenerative diseases characterized by accumulation of aggregated alpha-synuclein \nin the central and peripheral nervous systems . Alpha-synuclein is an intrinsically disordered \nprotein that misfolds into beta-sheet-rich structures that are detergent-insoluble, protease resistant, \nand act via a prion -like aggregation mechanism (1, 2). The N -terminal domain (1 -60 residues) \nfolds into alpha-helical structures when interacting with lipid membranes (3) and the C-terminus \n(96-140 residues) remains largely disordered. Both the N - and C- terminus shield core residues, \ntermed the non -amyloid component (NAC) domain (61 -95 residues) , which is  required for \naggregation. \nSynucleinopathies are  classically characterized by central nervous system (CNS)  \npathology, but increasing evidence highlights a role for alpha-synuclein in peripheral tissues. In \nparticular, the gastrointestinal tract has been proposed as a potential origin site for PD pathology  \n(4-6). Gastrointestinal dysfunction is a common non-motor symptoms of PD, with about 80% of \npatients reporting constipation years before the onset of motor symptoms (7). Moreover, misfolded \nalpha-synuclein is present in the enteric nervous system in PD, and is suggested to propagate from \nthe gastrointestinal tract to the brain via the vagus nerve , among other potential neural pathways \n(8).  \nThe vermiform appendix, a lymphoid-rich organ in the gastrointestinal tract that is highly \ninnervated by enteric neurons, harbors monomeric and aggregated alpha-synuclein in both healthy \nindividuals and in patients with PD (9). The presence of insoluble/aggregated alpha-synuclein in \nhealthy appendix suggests that it does not necessarily have to trigger a synucleinopathy in the \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n5 \n5 \nCNS; other factors likely contribute to the accumulation and propagation of pathogenic alpha-\nsynuclein to the brain. Epidemiological findings have previously suggested that removing the \nappendix early in life might reduce the risk of PD and delay the onset of PD late in life (9). In \nanother study, chronic appendicitis -like lesions were detected  using multislice spiral computed \ntomography (MSSCT) in 53% of PD patients, compared to 4% in matched controls  (10). This \nsuggests that chronic inflammation of the appendix is associated with PD in a subset of patients. \nAberrant gene expression and epigenetic modifications have been reported to contribute to \nsynucleinopathy through alterations in cellular homeostasis, stress response  pathways, and \ncompromising neuronal survival (11, 12). An earlier study from our team demonstrated epigenetic \nand gene expression abnormalities specifically in the autophagy -lysosomal pathway in the PD \nappendix (13). Additionally, perturbations in other proteostasis machinery responsible for the \ndegradation of misfolded proteins , like the  ubiquitin-proteasome system or chaperone-mediated \nautophagy (14), could contribute to the accumulation of aggregated alpha -synuclein in the \nappendix. Numerous studies have also consistently shown an association between inflammation \nand alpha-synuclein aggregation, although the underlying mechanisms remain to be fully \nelucidated  (15, 16). Notably, growing evidence from clinical and pre -clinical studies supports a \nsignificant link betwe en chronic gut inflammatory conditions like inflammatory bowel disease \n(IBD) and PD (17, 18). \nSpecific N- and C-terminus truncation enhances alpha-synuclein aggregation propensity \nand tendency to propagate from one neuron to another  (19), provided the alpha-synuclein NAC \ndomain (6 1-95 residues) remains intact. Several truncated proteoforms are enriched in Lewy \npathology and show enhanced aggregation kinetics in vitro (20-22). Past efforts have been focused \non identifying  post-translational modifications ( PTMs) of insoluble alpha-synuclein, and \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n6 \n6 \nconsequently, less is known about PTMs of the soluble alpha-synuclein pool in synucleinopathies. \nThis is despite evidence that smaller soluble oligomeric structures are key players in toxicity (23) \nand pathology spread (24). Moreover, the truncation of alpha-synuclein in peripheral tissues and \nits functional significance in disease pathogenesis is poorly understood.  \nWhile previous research has established a link between the appendix and PD, the complex, \nmulti-layered disease-related changes within the appendix that could potentially explain this link  \nremain unclear . We have taken  an integrated approach,  combining alpha-synuclein-seed \namplification assay (alpha -synuclein-SAA), transcriptomic (RNA sequencing), and epigenomic \n(DNA methylation profiling) analyses from the same appendix tissues, which allows for a holistic \nunderstanding of disease-associated molecular changes . Additionally, we  comprehensively \ncataloged alpha-synuclein proteoforms in the post-surgical appendix and substantia nigra (SN) of \npatients with synucleinopathy  and neurologically intact controls using top -down mass \nspectrometry (TD-MS), the gold -standard approach for identifying intact and truncated alpha-\nsynuclein proteoforms (25, 26). Based on our previous findings and other recent literature linking \nalpha-synuclein aggregation propensity (27), aberrant gene regulation (28), and truncation (29) to \npropagation of pathology, we hypothesized that aberrant gene regulation in appendices of certain \nindividuals facilitates accumulation of pathogenic alpha-synuclein and thereby increases risk of \ndeveloping synucleinopathy.  \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n7 \n7 \n \nRESULTS \nCatalog of alpha-synuclein proteoforms in the human appendix and SN \nTo determine whether the appendix contains alpha -synuclein proteoforms relevant to \ndisease pathology, we profiled the alpha -synuclein proteoforms in the appendix of \nsynucleinopathy patients and healthy individuals and performed comparisons to those present in \nthe SN. In total, 65 distinct alpha -synuclein proteoforms were identified in the human appendix \n(Figure 1A), ranging from the full-length N-terminally acetylated proteoform (alpha-synucleinac1-\n140) to truncated species 31 amino acids in length (alpha -synuclein78-109). There were truncated \nspecies with N - or C -terminal cleavage only (n= 4 and 13 alpha -synuclein proteoforms, \nrespectively). However, the large majority, 46 alpha-synuclein proteoforms, were cleaved at both \nends: 23 with N - and C -terminal truncations and 23 with intra -NAC domain and C -terminal \ntruncations. In total, we identified 27 unique cleavage locations for alpha -synuclein proteoforms \nin the appendix, of which 6 o ccurred within the N -terminus, 6 were in the NAC domain, and 15 \nwere in the C-terminus. The most common truncation site was in the C -terminus at position 114, \nwhich has been found to promote Lewy pathology formation in neurons (30), and other frequently \ncleaved sites included positions 40, 73, 109 and 115 of alpha -synuclein (number of proteoforms \nwith cleavage site: 13, 9, 9, 8, 8, respectively). Of the 65 alpha -synuclein proteoforms identified \nin the appendix, 56 were also detected in the SN. All proteoforms detected in the substantia nigra \nwere present in the appendix. We found 9 alpha -synuclein proteoforms that were present only in \nthe appendix, with most of these cleaved in both the NAC domain and C -terminus (alpha -\nsynuclein66-114, alpha -synuclein68-114, alpha -synuclein73-114, alpha -synuclein66-132, alpha -\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n8 \n8 \nsynuclein68-109, alpha-synuclein73-122, and alpha -synuclein73-125). Thus, the appendix contained a \ndiverse and distinct pool of alpha-synuclein truncation products. \nWe next compared the abundance of alpha-synuclein proteoforms in the appendix and SN \n(Figure 1B, 1C). First, we determined whether the appendix and SN had quantitative differences \nin common proteoforms (present in ≥50 samples). We found that alpha -synuclein proteoforms \nwith truncations at the C -terminus position 114/115 were significantly more common i n the \nappendix than in the SN ( q<0.05 for alpha -synuclein68-115, alpha -synuclein1-114 and alpha -\nsynuclein1-115, linear regression). In contrast, the appendix had lower levels of other C -terminal \ntruncations, including at positions 103, 119, and 122, relative to the SN ( q<0.05 for alpha -\nsynuclein1-103, alpha-synuclein1-119 and alpha-synuclein1-122, linear regression). To analyze the rare \nproteoforms, we determined whether each alpha -synuclein proteoform occurred more frequently \nin the appendix or SN (Fig. 1C). We found that in the appendix several rare proteoforms were \nmore abundant (alpha -synucleinac1-125, alpha -synuclein40-115, alpha -synucleinac1-129, alpha -\nsynucleinac1-114) and conversely, several rare proteoforms were more abundant in the SN (alpha -\nsynucleinac1-121, alpha -synuclein58-101, alpha -synuclein54-115, alpha -synuclein58-104, alpha -\nsynuclein76-104, alpha-synuclein68-119, alpha-synuclein73-119). We also found that levels of reliably \ndetected proteoforms alpha -synucleinac1-122 and alpha-synuclein78-113 were significantly depleted \nin synucleinopathy cases relative to controls (q<0.05, linear regression; Supplementary Figure S1). \nWe also detected instances of syn proteoforms that were unacetylated at the N-terminus for alpha-\nsynuclein1-140 and alpha -synuclein1-114, in addition to their abundant N -terminal acetylated \ncounterparts in both the appendix and SN (Fig. 1D). Therefore, both common and rare proteoforms \ndiffered in relative abundance between appendix and SN. Through classifier model training, we \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n9 \n9 \npredicted with an accuracy above 80% that alpha -synucleinac1-122 may be a highly relevant \npredictor of synucleinopathy in both appendix and SN (Supplementary Figure S1). \nWe further investigated age-dependent changes in the alpha-synuclein proteoforms present \nin the appendix and SN and found that synucleinopathy appendix had a significant gain in full -\nlength synuclein (alpha -synucleinac1-140) with age, which was not observed in the appendix of \nunaffected controls (q<0.05, linear regression; Supplementary Figure S2). Finally, sample -wise \nPearson correlation showed that synuclein proteoforms in the appendix are more similar to those \nin synucleinopathy SN than control SN (Supplementary Figure S3).  \n \nThe aggregation propensity of alpha-synuclein in the appendix \nGiven that alpha -synuclein aggregation plays a central role in the development of \nsynucleinopathies, we sought to characterize the aggregation propensity of the proteoforms we \nidentified in the appendix and SN. We modelled the aggregation potential of each  alpha-syuclein \nproteoform using a computational approach that determines the energy requirements for amyloid \nfibril formation (31). We then determined the alpha -synuclein aggregation propensity score for \neach sample based on proteoform abundances and their aggregation potential. We found that the \naggregation propensity of alpha-synuclein in the appendix was significantly higher than in the SN \n(p<10-10, linear regression; Figure 2A). Overall, this signifies that the appendix contains an \nabundance of aggregation-prone alpha-synuclein proteoforms (32-35). \nIn parallel, we employed an enhanced alpha-synuclein-SAA to detect alpha -synuclein \nseeding activity in postmortem appendix tissues from synucleinopathy patients compared to \nmatched controls. Using ThT fluorescence (AU) to measure seeding activity, we found 11 out of \n16 (68.75%) Synucleinopathy appendices exhibited increased seeding activity (Figure 2D) as \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n10 \n10 \nevidenced by increased ThT fluorescence (Figure 2B). In contrast, only 1 out of 15 controls (6.6%) \ndemonstrated detectable seeding activity. Over a period of 24 h, synucleinopathy appendices \nshowed a progressive increase in ThT fluorescence, whereas fluorescence in controls remained \nlower (Figure 2B). Furthermore, the average Fmax recorded at the plateau was significantly higher \nin synucleinopathy appendix as compared to controls (p=10 -4) (Figure 2C). These results show \nsignificantly higher overall alpha-synuclein seeding activity in synucleinopathy appendix. \nIn the appendix and SN, we detected full -length alpha -synuclein lacking N -terminal \nacetylation, which has been shown to substantially increase its aggregation potential while \nreducing its lipid and synaptic vesicle binding affinity (32-35)(p<10-4; Figure 1D). Indeed, full -\nlength alpha-synuclein that lacked N-terminal acetylation comprised 17.9% of the total full-length \nalpha-synuclein in the appendix and 3.8% in the SN. The only other N -terminal unacetylated \nalpha-synuclein proteoform detected in our study was unAc -alpha-synuclein1-114, which was \nprimarily found in the appendix; present in 50.9% of appendix samples compared to only 21.1% \nof SN samples (unAc -alpha-synuclein1-114 present in 27 of 53 appendix samples and 4 of 19 SN \nsamples; Figure 1D). Thus, delay in N-terminal acetylation in the appendix may accelerate alpha-\nsynuclein aggregation in this tissue.  No differences in N -terminal unacetylated alpha -synuclein \nwere observed between synucleinopathy cases and controls (Supplementary Figure S4). \n \nDifferential gene expression in disease \nA total of 22029 genes were assessed, out of which 137 genes were significantly \nupregulated, and 113 genes were significantly downregulated in synucleinopathy appendix \ncompared to healthy appendix (p-adj<0.05) (Figure 3A) (supplementary table S1). We identified \ndysregulation of genes related to heat shock and chaperone proteins associated with protein folding \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n11 \n11 \nand degradation, immune response , as well as genes associated with cili a assembly and \norganization (Table 1).  \nGene Set Enrichment Analysis (GSEA) was performed to assess whether predefined gene \nsets exhibit statistically significant differences between synucleinopathy and control samples. \nDifferentially expressed ( DE) genes in synucleinopathy appendix samples were found to be \nenriched in a total of 384 pathways, categorized as follows: 303 in GO Biological Processes (GO \nBP), 46 in GO Molecular Functions (GO MF), 31 in GO Cellular Components (GO CC), and 4 in \nKEGG pathways (supplementary data files S2-S5). The analysis highlighted several key processes \ninvolved in protein homeostasis and immune signaling that were significantly upregulated , and \nsome related to ciliary dynamics that were downregulated (Figure 3B, 3C, 3D). \n \nDifferential DNA methylation in disease \nThe overall distribution of differentially methylated regions (DMRs) across gene regions, \nCpG regions, and the total methylation profile suggests a predominantly hypomethylated status in \nsynucleinopathy appendix samples compared to healthy controls (Figure 4B, 4C, 4D). Most CpG \nsites were at promoter region transcription start site 1500 ( TSS1500)(67.29%), followed by  5’ \nuntranslated region (5’UTR)(12.87%), transcription start site 200 ( TSS200)(10.99%), gene body \n(6.43%), intergenic (1.61%) and 1st Exon (0.80%). In terms of CpG regions, the highest number \nof CpG sites was at islands (68.90%), followed by shores (22.52%), open sea (6.97%), and shelves \n(1.61%). \nWe found 46 unique genes in the vicinity of 45 DMRs- 39 genes were hypomethylated, \nwhile 7 genes were hypermethylated (supplementary table S6). Methylation differences between \nsynucleinopathy and control appendices were subtle but aligned with our transcriptomic analysis; \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n12 \n12 \nwe found differentially methylated genes related to protein degradation, folding and homeostasis, \nimmune response, and ciliary and cytoskeletal dynamics  (Table 2) . Additionally, we found \ndifferential methylation in some genes associated with oxidative stress and mitochondrial function, \nboth of which have been implicated in neurodegeneration. GSEA revealed neuronal signaling and \ndevelopment pathways that were enriched for hypomethylated genes, and stress response and cell \nsignaling pathways that were enriche d for hypermethylated genes in synucleinopathy appendix \n(Figure 5) (supplementary tables S7-S10). \n \nFunctional clustering based on protein interactions in disease \nDifferentially expressed and methylated genes (p -adj<0.05) from RNA -seq and DNA \nmethylation analyses were analyzed using the STRING database to create a network map (Figure \n6). Functional clusters identified through Markov Cluster Algorithm (MCL) aligned with key \npathways highlighted in GSEA analysis, including protein homeostasis (Unfolded protein binding \nand Chaperone complex), immune response (Autoimmune  disease), and ciliary structure \n(Axonemal microtubule) networks.  \n \nDISCUSSION \nThe vermiform appendix has emerged as a potential peripheral site where increased levels \nof aggregated alpha-synuclein potentially appear early in the course of synucleinopathies (9). In \nthe presence of impaired autophagy-lysosomal function in the same tissue (13), aggregated alpha-\nsynuclein might propagate from enteric nerves to the CNS, making the appendix one possible site \nfor initiation of the first steps in the pathogenesis of synucleinopathies. Despite growing evidence, \nthe mechanisms causing alpha-synuclein aggregation in the appendix are poorly understood. To \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n13 \n13 \naddress this knowledge gap, we employed proteomic, transcriptomic, and epigenomic profiling of \nappendices from synucleinopathy patients and healthy controls. Our findings reveal a distinct, \naggregation-prone pool of alpha -synuclein proteoforms in the synucleinopathy appendix, and \nsignificant dysregulation in protein homeostasis, immune response , and ciliary function and \nprocesses in the synucleinopathy appendix. These findings provide a novel molecular insight into \nthe vermiform appendix as a peripheral site where alpha-synuclein aggregates in the earliest stages \nof the disease process. \nTo characterize the molecular features underlying synuclein pathology , we investigated \nalpha-synuclein aggregation and truncation in appendix tissues and SN. When we modeled \naggregation propensity for the total pool of alpha-synuclein proteoforms identified using TD-MS, \nwe observed a higher aggregation propensity in the appendix in comparison to SN. This suggests \nthe unique proteolytic processing in the vermifo rm appendix “primes” the endogenous alpha -\nsynuclein pool for pathological aggregation, which may also be true of other peripheral tissues \nwhere pathology has also been documented (36, 37). We then investigated this experimentally \nusing alpha-synuclein-SAA, where we detected seeding activity in 68.75% of synucleinopathy \nappendix samples compared to 6.6% of healthy appendix samples. Importantly, fluorescence \nkinetics revealed higher seeding activity, reflected by greater ThT signal and  Fmax, in \nsynucleinopathy appendix relative to controls. These findings are consistent with previous studies \nshowing that gastrointestinal tissues from synucleinopathy patients exhibit more consistent \nseeding activity compared to controls (38). Notably, our finding that 68.75% of synucleinopathy \nappendices are SAA-positive aligns well with the proposed “body-first subtype” of Lewy body \ndiseases (39), in which pathology is hypothesized to begin in the enteric nervous system and ascend \nto the brain via the vagus and sympathetic nerves. According to this model, approximately half of \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n14 \n14 \nthe Lewy body disease patients belong to the “body-first” subgroup, and the remaining fall into a \n“brain-first” category, with around 70% cases  of dementia with Lewy bodies (DLB)  and only \n~30% of PD cases exhibiting a body-first pattern (40). Since our synucleinopathy cohort was DLB-\nrich (n=12 of 16), the observation that 68.75% of our  synucleinopathy appendix samples were \nSAA-positive is consistent with the expected proportion of body-first cases. Conversely, 31.25% \nof synucleinopathy appendices were negative for seeding activity in alpha -synuclein-SAA, \nsuggesting a brain-first trajectory while exhibiting less peripheral involvement, demonstrating the \nheterogeneity of alpha-synuclein pathology within synucleinopathies.  \nWe previously identified the appendix as a reservoir of intraneuronal alpha -synuclein \naggregates that are enriched with truncated proteoforms (9), which raised questions about the \nproteolytic turnover of alpha -synuclein in the appendix.  Here we addressed those questions by \ncataloging all detectable alpha -synuclein proteoforms using a TD -MS approach, similar to what \nhas been done in brain tissue (26). This study provides one of the most comprehensive catalogs of \nalpha-synuclein truncation in human tissues  to date. We compared alpha -synuclein proteoforms \nbetween tissues and found proteoform  abundance varied significantly between the appendix and \nSN. Unexpectedly, 9 proteoforms were unique to the appendix and were not detected in any of the \nSN samples. Conversely, two proteoforms, namely alpha -synuclein78-113 and alpha-synucleinac1-\n122, were decreased in the synucleinopathy appendix compared to controls. The decrease in these \nproteoforms can be attributed to either abnormal proteolytic turnover of alpha -synuclein in the \nsynucleinopathy appendix or possibly indicative of aggregation, as proteof orms that form \ninsoluble inclusions would not be detected with our methodology.  \nSeveral of the cleavage sites and proteoforms we identified have been described previously \n(9, 21, 26), but we also characterize d proteoforms not yet described in literature. As truncated \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n15 \n15 \nforms of alpha -synuclein are a substantial component of inclusions (41), characterization of the \nproteoforms present in the appendix and SN may improve  our understanding of the origins of \nsynucleinopathies. Truncation and N -terminal acetylation were the only detectable PTMs. Like \nmany eukaryotic proteins, the N -terminal methionine of mature alpha -synuclein is permanently \nacetylated. Here, we made the interesting observation that full -length and some C -terminally \ntruncated alpha -synuclein species  were unacetylated. Although the levels of N -terminally \nunacetylated alpha -synuclein were relatively low in the brain and appendix, several appendix \nsamples showed a surprising abundance of this particular proteoform (~1:2). Furthermore, \nunacetylated alpha-synuclein was the third most abundant proteoform we detected regardless of \nsynucleinopathy or tissue type. Unacetylated alpha-synuclein likely represents immature or newly \nsynthesized protein that has not yet been enzymatically modified. T hus, in some instances alpha-\nsynuclein was cleaved prior to becoming a mat ure protein. This raises the possibility that alpha -\nsynuclein truncation can be an early event in the protein's natural lifecycle. Of note, we did not \ndetect any phosphorylated proteoforms, including the disease -associated phosphoserine 129, \nwhich is in agreement with prev ious studies showing very low levels of this modification in the \nnormal soluble alpha-synuclein pool (26). Furthermore, alpha-synuclein from the human appendix \nwas more abundant in truncated prote oforms when compared to the brain . Combined with our \nprevious observation of an impaired lysosomal autophagy system in the Parkinson’s disease \nappendix (13), this might suggest that the turnover of alpha-synuclein was impaired, leading to an \naccumulation of truncated proteoforms and the formation of proteopathic seeds that could initiate \nthe first steps of synucleinopathy. \nTogether, our results suggest that proteolytic degradation (i.e. turnover) of alpha-synuclein \nis tissue-specific, and therefore some tissues, like the vermiform appendix, may be more likely to \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n16 \n16 \naccumulate alpha-synuclein, including rare truncated proteoforms. Some important considerations \nshould be taken when interpreting our findings. First, immunopurification of alpha-synuclein was \nconducted using clone 42 anti -alpha-synuclein antibody (BD Transduction Laboratories) which \nmaps to residues 91 -99 of human alpha -synuclein (42), and thus, all proteoforms detected \nexpectedly contain this epitope, and any proteoforms  lacking this epitope would not be detected \nhere. The total alpha -synuclein signal intensity was similar between appendix and SN samples \ndespite relatively low expression of alpha -synuclein in the appendix, indicating that \nimmunopurification w as conducted at near -saturating conditions. As a result, our data are not \nindicative of changes in total alpha-synuclein content of the tissues but instead are an indicator of \nthe proteoform signature of the samples. Because our analyses were performed on the triton x100 \nsoluble fraction of alpha-synuclein, it is limited to proteoforms found outside of inclusion bodies. \nConceptually, proteoforms identified here are “pre -pathological” and provide insight into the \nnormal turnover of this protein in peripheral and CNS tissues. \nBuilding on our TD-MS and seed amplification assay results, which indicate a heightened \naggregation potential in synucleinopathy appendix, we turned to multi -omics profiling to \ninvestigate the transcriptional and epigenetic underpinnings of this phenomenon. We identified a \nprominent alteration in genes involved in proteostasis and quality control, which suggests cellular \nstress and activated mechanisms to counteract misfolded protein burden  in the synucleinopathy \nappendix. Activated pathways associated with molecular assemblies such as intracellular protein-\ncontaining complex, transferase complex and ribonucleoprotein complex hint towards changes in \nprotein turnover, potentially instigated by aggregation of misfolded alpha-synuclein in the \nsynucleinopathy appendix. Molecular function pathways relevant to synuclein pathology such as \nprotein phosphorylated amino acid binding, phosphoprotein binding , and phosphotransferase \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n17 \n17 \nactivity were also upregulated. In the context of PD, these pathways are central to the recognition \nof phosphorylated alpha-synuclein, and binding to 14-3-3 domain proteins to produce Lewy body \npathology (43), further highlighting a dysregulated proteostasis network. Simultaneously, protein \nfolding chaperone, and HSP/protein folding chaperone  binding pathways were upregulated in \nsynucleinopathy appendix, suggesting a cellular attempt to attenuate misfolding and aggregation \nof alpha-synuclein. Activation of pathways associated with lysosomal membrane and lytic vacuole \nmembrane indicates an increase in compensatory autophagic clearance. DNA methylation pathway \nenrichment revealed hypomethylated  (may indicate potentially  increased activity)  genes \nassociated with cation channel complex, as well as d ifferential methylation of calcium signaling \npathway genes  (both hypo - and hyper -enriched), reflecting altered regulation of processes  \nimplicated in alpha-synuclein aggregation (44). Additionally, hypermethylation ( may indicate \npotentially decreased activity) of genes associated with Wnt signaling pathway and negative \nregulation of autophagy is consistent with our previous observations at transcriptomic level. At the \ngene level, these changes  were reflected in  the observed upregulation of  several molecular \nchaperones and co-chaperones of heat shock proteins (HSPs) such as AHA1, STIP1, METTL21A, \nHSPA1L, and HSPA4  (45). The chaperones HSP70 and HSP90  play a crucial role in mitigating \ncellular stress such as accumulation of misfolded proteins (46), degradation of toxic protein \naggregates (47), and modulating alpha -synuclein dynamics (48). We also observed \nhypomethylation in promoter regions of several genes implicated in autophagy, lysosomal \ntrafficking, and stress signaling, such as CDK16, SH3GLB1, and ABCB9 in the synucleinopathy \nappendix (49) (50) (51), and hypermethylation of DCUN1D1, which was previously identified as \na novel candidate gene in protein interactome-based PD studies (52). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n18 \n18 \nImmune response -related pathways such as T cell receptor signaling, lymphocyte \ndifferentiation and activation, and adaptive immune response , were activated in transcriptomic \nenrichment analysis, indicating a sustained state of immune activation in the synucleinopathy \nappendix. Interestingly, immune response regulating pathways were also activated, potentially \nsuggesting a compensatory attempt to maintain immune balance in response to proteostatic stress. \nDifferential methylation of the stress and immune regulatory MAPK signaling pathway (both \nhypo- and hyper -enriched), and hyper-enrichment of Toll -like receptor signaling also suggest \nimmune remodeling at the epigenetic level. Our observations at gene level, such as upregulation \nof chemokine and interferon response genes CCL2, IRF7, IL7R, and the immunomodulatory \ncheckpoint gene VSIR, further reflect the complex immune environment within synucleinopathy \nappendixes. Particularly, in the context of neurodegenerative disease, CCL2 and VSIR have been \nassociated with microglial activation and neuroinflammation (53, 54). We also found \nhypomethylation in the promoter region of long noncoding RNA H19, the overexpression of which \nhas been associated with downregulation of tight junction proteins  ZO-1 and occludin (55). This \nmay, in turn, compromise  the integrity of the epithelial barrier in the synucleinopathy appendix \nand further increase vulnerability to inflammation , while simultaneously allowing for microbial \nproducts inside the appendix  to circulate systemically . This so -called ‘leaky gut’  state, in \ncombination with the perturbed  local immune environment, may contribute to systemic \ninflammation and potentially influenc e alpha-synuclein pathology (56). These results are also \nconsistent with another recent study, where transcriptomic and epigenomic profiling of the colonic \nmucosa revealed shared immune dysregulation in PD and inflammatory bowel disease patients \n(57). Their findings also suggest that chronic peripheral inflammation can increase the risk for PD \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n19 \n19 \nin inflammatory bowel disease  patients, further strengthening the link between gut immune \ndysregulation and neurodegeneration. \nOur data also revealed consistent dysregulation of genes related to primary and motile \nciliary structure and function. Primary cilia are essential cellular organelles that are present on the \nepithelial layer, as well as in the enteric neurons embedded inside the appendix  (58). These \nstructures are involved in modulating the Sonic Hedgehog (Shh) pathway, which is crucial for \nneurogenesis, and has been linked to  PD (59, 60). Motile cilia, on the other hand,  are found on \nspecialized cells and facilitate the movement of fluid over epithelial surfaces (61). In pathway \nanalysis, we observed significant disruption of cilia -related processes as well as  other structural \ncomponents, as reflected by suppressed pathways associated with axoneme assembly, cilium \nassembly and organization,  ciliary plasm, mobile cilium, ciliary transition zone,  cilium- or \nflagellum-dependent cell motility,  epithelial morphogenesis, and extracellular matrix . Beyond \ndysfunctional signaling, disruption of these components would also have implications for barrier \nintegrity, corroborating our earlier observation. The hypo-enrichment of KEGG terms such as \nregulation of actin cytoskeleton and focal adhesion may indicate compensatory activity of the \nepigenetic machinery, which appears to mitigate  transcriptomic suppression. The observed \ndownregulation of primary cilia stability gene  SAXO1, and motile cilia function genes LCA5L, \nCFAP45, and DNAI7, as well as hypomethylation of the downstream cellular regulators GLI4 and \nGAS7, would further hint towards downstream disruption of Shh signaling  and possibly, GI \nmotility, in synucleinopathy patients. Interestingly,  GLI4 and GAS7 have been associated with \nPD in previous studies (62, 63). Additionally, LRRK2 mutations, which are common in PD, have \nalso been found to interfere with cilia formation and Shh signaling, further suggesting a potential \nmechanistic link between ciliary dysfunction and PD development (64). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n20 \n20 \nAlthough we observed no direct overlap between differentially methylated genes and \ndifferentially expressed genes in our data after FDR correction, both datasets overlap substantially \nin functional themes. This  suggests that even subtle methylation changes in the synucleinopathy \nappendix observed here could potentially influence gene expression, and the divergence could be \nbecause of additional regulatory layers such as histone modifications or non-coding RNAs.  \nThe present study provides new insights into the molecular landscape of the vermiform \nappendix in PD, highlighting it as a potential  contributor to pathology. The interplay between \nimpaired protein homeostasis , immune dysregulation, and ciliary dysfunction may prove \nconducive to aggregation of alpha-synuclein, which could then lead to propagation of pathology \nto the central nervous system. Further research into the identified dysregulated genes and \npathways, as well as aberrant  alpha-synuclein processing in the appendix, may provide critical \ninsights and inform the development of therapeutics for synucleinopathies.  \n \nMATERIALS AND METHODS \nStudy design \nThis study was designed to investigate the molecular mechanism underlying alpha -synuclein \naggregation in vermiform appendix. We conducted TD-MS to characterize alpha -synuclein \nproteoforms in post-surgical appendix (24 synucleinopathy cases and 31 controls) and postmortem \nSN (10 synucleinopathy cases and 10 controls) tissue samples (20 -100 mg). Next, we performed \nin silico modeling of aggregation propensity for each proteoform characterized in the previous \nexperiment. We also investigated this in vitro  utilizing alpha -synuclein-SAA to differentiate \nseeding activity exhibited by synucleinopathy and healthy control appendixes  \n(synucleinopathy=16, control=15, 30 -40mg each) . Excluding the samples without genomic \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n21 \n21 \nprofiling permissions, the same appendix samples were  utilized to explore gene regulation \nabnormalities through RNA seq and DNA methylation profiling (synucleinopathy=11, control=15, \n30-40mg each) .  After initial QC, one healthy control appendix sample was excluded from \nsubsequent analyses (final control n=14). Network analysis was performed to integrate findings \nfrom transcriptomic and epigenomic analyses. Due to sample collection constraints, no statistical \nmethods were used to predetermine sample size. Number of technical replicates and positiv ity \ncriteria for alpha-synuclein-SAA experiment are detailed below. \n \nAppendix sample processing for enhanced alpha-synuclein seed amplification assay (alpha-\nsynuclein-SAA) \nPostmortem human vermiform appendix samples  were cryopulverized using the CP02 \nAutomated Dry Pulverizer (Covaris LLC, Woburn, MA, USA) and prepared at a 3% (w/v) \nconcentration using a homogenization buffer containing PBS and 1x EDTA-free protease inhibitor \nfrom Roche. Homogenization was performed with Precellys Lysing Kit using the Bertin Precellys-\n24 Dual homogenizer for two 30 -second cycles  at 6500 rpm. Following homogenization, the \nappendix samples were centrifuged at 800 × g for 1 minute at 4°C. The resulting supernatant was \ncollected, vortexed, aliquoted, snap-frozen, and stored at -80°C until further use. \n \n Alpha-synuclein-SAA procedure \nThe enhanced alpha-synuclein-SAA used here is a modification of a previously described \nassay (65), and was performed following a protocol described by Ma et al. (66). Briefly, the \nreaction mixture contained 100mM PIPES pH 6.50, 500mM NaCl, 10µM thioflavin T (ThT), 0.1% \nsarkosyl, and 0.3mg/mL recombinant alpha-synuclein. The assay was performed in 96-well clear-\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n22 \n22 \nbottom plates in the presence of two 3.2mm Si3N4 beads per well. Plates were incubated at 42°C \nwith intermittent orbital shaking (800rpm) for 1 min every 15 mins and fluorescence was read at \nevery cycle (15 min), using FLUOstar Omega (BMG LABTECH, Ortenberg, Germany). Samples \nwere tested in triplicate during a 24h period and were considered positive when three replicates \nexhibited a fluorescence > 1000 AU. Samples with 0 or 1 positive replicates were designated as \nnegative. Samples with 2 positive replicates were considered inconclusive and were tested again. \nThe maximum fluorescence (Fmax) of each sample was determined by averaging Fmax from all \nthree replicates. \n \nAlpha-synuclein immunoprecipitation for mass spectrometry \nAppendix and SN tissue samples were ground on liquid nitrogen with a mortar and pestle. \nThe resulting powdered tissue was suspended in lysis buffer (PBS, 1% Triton X-100, 2mM EDTA, \n2mM PMSF, 1X Roche protease inhibitor cocktail) and sonicated at 50% intensity with 2s pulses \nfor 30 pulses total, on ice using Q55 model sonicator (QSonica). Samples were then vortexed and \nthen incubated on ice for 30 min. Samples were centrifuged at 22,000 -x g for 20 min, and the \nsupernatant was retained. Protein content was determined by BCA assay (Thermo Fisher) and \nadjusted to 2 mg protein / mL lysis buffer. \nFor each sample, 50 µl of the tosylactivated beads (see supplementary methods for details \non preparation of beads)  was collected on a magnetic stand and combined with 200 µl of tissue \nlysate. Samples were then incubated overnight at 4°C with nutation. The sample -exposed beads \nwere magnetically collected and washed three times with ice-cold lysis buffer (1 mL of PBS, 1 % \nTriton X-100, 2mM EDTA, 2mM PMSF, 1 X Roche protease inhibitor cocktail), with each wash \nperformed at 4°C with nutation for 30 min. Next, th e samples were washed with 1 mL PBS pH \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n23 \n23 \n7.4, three times at 4°C with nutation for 15 min and then rinsed three times with 1 mL ultrapure \nwater (Thermo Fisher). To elute alpha -synuclein from the beads, samples were suspended in 10 \nµL of 1% formic acid in ultrapure water and mixed vigorously for 5  min at room temperature. \nSamples were then centrifuged for 5 s, placed on the magnet, and the eluent collected into a protein \nLoBind Eppendorf tube (Sigma-Aldrich). The elution procedure was repeated with another 10 µL \nof 1% formic acid, and both eluents were pooled. The presence of alpha-synuclein was verified in \nan aliquot (2 µL) of the eluent by western blotting, and the remaining eluent was immediately snap \nfrozen in liquid nitrogen and stored at -80°C until top-down mass spectrometry analysis. \n \nLiquid chromatography-mass spectrometry \nAfter immunoprecipitation, samples eluted in high acid were loaded in a nanocapillary \nliquid chromatography (LC) (Dionex RSLCnano) setup for proteoform separation. A trap column \n(2 cm × 150 μm i.d.) and analytical (20 cm × 75 μm) column were used, both packed with PLRP-\nS stationary phase (5 μm particle size, Agilent, Santa Clara, CA) and proteins were eluted over a \n40 min gradient of 95:5 water/acetonitrile with 0.2% formic acid to 5:95 water/acetonitrile into a \ncustom electrospray ionization (ESI) source t erminated with a PicoTip spray emitter (New \nObjective, Woburn, MA). Proteoforms were then analyzed online on a LTQ Velos Orbitrap Elite \nmass spectrometer (Thermo Fisher Scientific) operated in a data -dependent mode, using \nestablished instrument methods (67). Two LC-MS replicates of each sample were analyzed by LC-\nMS, and sample replicates were run in full random order to mitigate the effect of instrumental drift \nin differential proteoform quantification. The ensuing raw data was then searched for target alpha-\nsynuclein proteoforms using a previously described custom MS1 -based isotopic fitting software  \n(see supplementary methods for details on data processing) (68).  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n24 \n24 \n \nRNA sequencing (RNA-seq)  \nTotal RNA was extracted from previously mentioned appendix samples utilized for alpha-\nsynuclein-SAA using the KAPA RNA HyperPrep Kit (Kapa Biosystems, Wilmington, MA USA). \nLibraries were prepared by the Van Andel Genomics Core from 445 ng of  the total RNA. \nRibosomal RNA material was reduced using the QIAseq FastSelect –rRNA HMR Kit (Qiagen, \nGermantown, MD, USA). RNA was sheared to 300 -400 bp. Prior to PCR amplification, cDNA \nfragments were ligated to IDT for Illumina TruSeq UD Indexed adapters (Illumina Inc, San Diego \nCA, USA). Quality and quantity of the finished libraries were assessed using a combination of \nAgilent DNA High Sensi tivity chip (Agilent Technologies, Inc.), QuantiFluor® dsDNA System \n(Promega Corp., Madison, WI, USA), and Kapa Illumina Library Quantification qPCR assays \n(Kapa Biosystems).  Individually indexed libraries were pooled and 100 bp paired-end sequencing \nwas performed on an Illumina NovaSeq6000 sequencer to an average depth of 60M raw paired -\nreads per transcriptome. Base calling was done by Illumina RTA3 and output of NCS was \ndemultiplexed and converted to FastQ format with Illumina Bcl2fastq v2.20.0.  Differential gene \nexpression (DGE) analysis was performed using DESeq2 (69) based on gene counts generated \nfrom STAR. Further details on data processing have been provided in supplementary methods. \n \nDNA methylation analysis \nDNA extracted from previously mentioned cryopulverized  appendix samples was \nquantified by Qubit fluorometry (Life Technologies), and 250 ng of DNA from each sample was \nbisulfite converted using the Zymo EZ DNA Methylation Kit (Zymo Research, Irvine, CA, USA) \nfollowing the manufacturer’s protocol with the spec ified modifications for the Illumina Infinium \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n25 \n25 \nMethylation Assay. After conversion, all bisulfite reactions were cleaned using the Zymo -Spin \ncolumns and eluted in 12 μL of Tris buffer. Following elution, BS converted DNA was processed \nthrough the Infinium MethylationEPIC array v2.0 protocol (Illumina I nc., San Diego CA, USA).  \nThe EPIC array v2.0 contains >930K probes querying methylation sites including CpG islands and \nnon-island regions, RefSeq genes, ENCODE open chromatin, ENCODE transcription factor \nbinding sites, and FANTOM5 enhancers. To perform t he assay, 4uL of converted DNA was \ndenatured with 4ul 0.1N sodium hydroxide.  DNA was then amplified, hybridized to the EPIC \nbead chip, and an extension reaction was performed using fluorophore-labeled nucleotides per the \nmanufacturer’s protocol.  Array beadchips were scanned on the Illumina iScan platform and probe-\nspecific calls were made using Illumina Genome Studio software.  The R package SeSAMe (70) \nwas used to process IDAT files generated from the Infinium MethylationEPIC v2.0 array and for \ndownstream differential methylation locus (DML) and differential methylation region (DMR) \nanalysis. Further details on data processing have been provided in supplementary methods. \n \nNetwork analysis \nSignificant genes (p -adj. 0.05) identified from RNA -seq and methylation analysis were \nanalyzed using the STRING database. MCL clustering was used to organize nodes, and the top \nstatistically significant functional enrichment (i.e., lowest FDR value) was displayed above each \ncluster. TP53 was removed from the g ene list because it was identified as a disease -associated \nvariant, rather than a methylation site.   \n \nStatistical Analysis \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n26 \n26 \nThe alpha-synuclein-SAA data (ThT fluorescence and Fmax) collected from appendices \nwere analyzed using the Mann -Whitney U test. For common proteoforms identified via LC -MS, \nproteoforms with more than 50% missing values were removed. Raw proteoform counts were \nlog10 transformed, and samples were scaled to 0 mean and a unit standard deviation. Values were \nthen averaged across the replicates. For rare proteoforms, proteoforms with more than 50% \nmissing values were retained. Rare proteoforms were detected if raw counts existed in any of the \nreplicates. For aggregation propensity, proteoform raw counts for each sample were first divided \nby its total, then multiplied by the aggregation propensity score calculated by P ASTA 2.0 (31). \nValues were then averaged across the replicates. We used Limma’s robust linear regression to \nidentify tissue differences, adjusting for disease group and age, and empirical Bayes was used to \nadjust coefficient estimates. P -values were corrected for multi ple testing using FDR. Group \ndifferences were tested separately for the appendix and SN. For proteoform abundance, aging \ndifference was tested separately for appendix and SN. Limma’s robust linear regression was fitted \non a model that includes group and age interaction. Then contrast.fit with contrast.matrix was used \nto perform aging difference on control and synucleinopathy. The p -values were corrected for \nmultiple testing using FDR. For aggregation propensity, robust linear regression from Limma was \nfirst fitted on a model that includes tissue and age interaction, adjusting for group. Then contrast.fit \nwith contrast.matrix was used to perform aging difference on the appendix and SN. p-values were \ncorrected for multiple testing using FDR. \nCommon proteoforms were used for clustering analysis. For the remaining proteoforms the \nmissing values were imputed with the MICE package (71). Then enhanced hierarchical clustering \nwas performed by R. The clustering result was plotted by factoextra package (72). The following \ntissue and group difference analysis for each cluster was based on the average values of \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n27 \n27 \nproteoforms inside the cluster. Common proteoforms were used for diagnosis prediction using the \nR package caret (73). Proteoforms that have missing values were imputed using k-nearest neighbor \nimputation (k=5). The resulting data matrix was used to train a classifier using 10 -fold repeated \ncross-validation (3 repeat iterations). Within each cross -validation step, the da ta were further \npreprocessed by removing near -zero variable proteoforms and then scaling and centering each \nproteoform to a 0 mean and unit standard deviation. In order to assess variable importance, a \nclassifier was trained on all the data using the parameters learned in the cross-validation step.  \n \nList of Supplementary Materials \nSupplementary Methods \nFig S1 to S4 \nData files S1 to S10 \n \nReferences and Notes \n \n1. H. Miake, H. Mizusawa, T. Iwatsubo, M. Hasegawa, Biochemical characterization of the \ncore structure of alpha-synuclein filaments. J. Biol. Chem. 277, 19213-19219 (2002). \n2. R. K. Leak, M. P. Frosch, T. G. Beach, G. M. Halliday, Alpha-synuclein: prion or prion-\nlike? Acta Neuropathol. 138, 509-514 (2019). \n3. N. Bisi, L. Feni, K. Peqini, H. Pérez-Peña, S. Ongeri, S. Pieraccini, S. Pellegrino, α-\nSynuclein: An All-Inclusive Trip Around its Structure, Influencing Factors and Applied \nTechniques. Front Chem 9, 666585 (2021). \n4. H. Braak, K. Del Tredici, U. Rüb, R. A. de Vos, E. N. Jansen Steur, E. Braak, Staging of \nbrain pathology related to sporadic Parkinson's disease. Neurobiol. Aging 24, 197-211 \n(2003). \n5. H. Braak, K. Del Tredici, Potential Pathways of Abnormal Tau and alpha-Synuclein \nDissemination in Sporadic Alzheimer's and Parkinson's Diseases. Cold Spring Harb \nPerspect Biol 8,  (2016). \n6. H. Braak, U. Rub, W. P. Gai, K. Del Tredici, Idiopathic Parkinson's disease: possible \nroutes by which vulnerable neuronal types may be subject to neuroinvasion by an \nunknown pathogen. J Neural Transm (Vienna) 110, 517-536 (2003). \n7. M. N. Han, D. I. Finkelstein, R. M. McQuade, S. Diwakarla, Gastrointestinal Dysfunction \nin Parkinson's Disease: Current and Potential Therapeutics. J Pers Med 12,  (2022). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n28 \n28 \n8. M. Chen, D. E. Mor, Gut-to-Brain α-Synuclein Transmission in Parkinson's Disease: \nEvidence for Prion-like Mechanisms. Int. J. Mol. Sci. 24,  (2023). \n9. B. A. Killinger, Z. Madaj, J. W. Sikora, N. Rey, A. J. Haas, Y. Vepa, D. Lindqvist, H. \nChen, P. M. Thomas, P. Brundin, L. Brundin, V. Labrie, The vermiform appendix impacts \nthe risk of developing Parkinson's disease. Sci. Transl. Med. 10,  (2018). \n10. Y. Chen, W. Wu, S. Zhao, X. Lv, J. Hu, C. Han, G. Wang, S. Wang, P. Bo, J. Zhang, W. \nGui, Q. Tang, Q. Liu, S. Zhu, F. Yu, Increased Accumulation of α-Synuclein in Inflamed \nAppendices of Parkinson's Disease Patients. Mov. Disord. 36, 1911-1918 (2021). \n11. H. Song, J. Chen, J. Huang, P. Sun, Y. Liu, L. Xu, C. Wei, X. Mu, X. Lu, W. Wang, N. \nZhang, M. Shang, M. Mo, W. Zhang, H. Zhao, F. Han, Epigenetic modification in \nParkinson's disease. Front Cell Dev Biol 11, 1123621 (2023). \n12. M. Tsalenchuk, S. M. Gentleman, S. J. Marzi, Linking environmental risk factors with \nepigenetic mechanisms in Parkinson's disease. NPJ Parkinsons Dis 9, 123 (2023). \n13. J. Gordevicius, P. Li, L. L. Marshall, B. A. Killinger, S. Lang, E. Ensink, N. C. Kuhn, W. \nCui, N. Maroof, R. Lauria, C. Rueb, J. Siebourg-Polster, P. Maliver, J. Lamp, I. Vega, F. \nP. Manfredsson, M. Britschgi, V. Labrie, Epigenetic inactivation of the autophagy-\nlysosomal system in appendix in Parkinson's disease. Nat Commun 12, 5134 (2021). \n14. D. Ebrahimi-Fakhari, P. J. McLean, V. K. Unni, Alpha-synuclein's degradation in vivo: \nopening a new (cranial) window on the roles of degradation pathways in Parkinson \ndisease. Autophagy 8, 281-283 (2012). \n15. C. Yildirim-Balatan, A. Fenyi, P. Besnault, L. Gomez, J. E. Sepulveda-Diaz, P. P. Michel, \nR. Melki, S. Hunot, Parkinson's disease-derived α-synuclein assemblies combined with \nchronic-type inflammatory cues promote a neurotoxic microglial phenotype. J. \nNeuroinflammation 21, 54 (2024). \n16. G. Forloni, Alpha Synuclein: Neurodegeneration and Inflammation. Int. J. Mol. Sci. 24,  \n(2023). \n17. M. Villumsen, S. Aznar, B. Pakkenberg, T. Jess, T. Brudek, Inflammatory bowel disease \nincreases the risk of Parkinson's disease: a Danish nationwide cohort study 1977-2014. \nGut 68, 18-24 (2019). \n18. S. Grathwohl, E. Quansah, N. Maroof, J. A. Steiner, L. Spycher, F. Benmansour, G. \nDuran-Pacheco, J. Siebourg-Polster, K. Oroszlan-Szovik, H. Remy, M. Haenggi, M. \nStawiski, M. Selhausen, P. Mailver, A. Wolfert, T. Emrich, Z. Madaj, A. Su, M. L. \nEscobar Galvis, C. Mueller, A. Herrmann, P. Brundin, M. Britschgi, Specific immune \nmodulation of experimental colitis drives enteric alpha-synuclein accumulation and \ntriggers age-related Parkinson-like brain pathology. Free Neuropathology 2, 13 (2021). \n19. M. Terada, G. Suzuki, T. Nonaka, F. Kametani, A. Tamaoka, M. Hasegawa, The effect \nof truncation on prion-like properties of alpha-synuclein. The Journal of biological \nchemistry 293, 13910-13920 (2018). \n20. M. Baba, S. Nakajo, P. H. Tu, T. Tomita, K. Nakaya, V. M. Lee, J. Q. Trojanowski, T. \nIwatsubo, Aggregation of alpha-synuclein in Lewy bodies of sporadic Parkinson's \ndisease and dementia with Lewy bodies. The American journal of pathology 152, 879-\n884 (1998). \n21. W. Li, N. West, E. Colla, O. Pletnikova, J. C. Troncoso, L. Marsh, T. M. Dawson, P. \nJäkälä, T. Hartmann, D. L. Price, M. K. Lee, Aggregation promoting C-terminal \ntruncation of alpha-synuclein is a normal cellular process and is enhanced by the familial \nParkinson's disease-linked mutations. Proc. Natl. Acad. Sci. U. S. A. 102, 2162-2167 \n(2005). \n22. T. Chakroun, V. Evsyukov, N. P. Nykanen, M. Hollerhage, A. Schmidt, F. Kamp, V. C. \nRuf, W. Wurst, T. W. Rosler, G. U. Hoglinger, Alpha-synuclein fragments trigger distinct \naggregation pathways. Cell death & disease 11, 84 (2020). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n29 \n29 \n23. G. Fusco, S. W. Chen, P. T. F. Williamson, R. Cascella, M. Perni, J. A. Jarvis, C. Cecchi, \nM. Vendruscolo, F. Chiti, N. Cremades, L. Ying, C. M. Dobson, A. De Simone, Structural \nbasis of membrane disruption and cellular toxicity by alpha-synuclein oligomers. Science \n358, 1440-1443 (2017). \n24. M. Helwig, M. Klinkenberg, R. Rusconi, R. E. Musgrove, N. K. Majbour, O. M. El-Agnaf, \nA. Ulusoy, D. A. Di Monte, Brain propagation of transduced alpha-synuclein involves \nnon-fibrillar protein species and is enhanced in alpha-synuclein null mice. Brain : a \njournal of neurology 139, 856-870 (2016). \n25. N. Siuti, N. L. Kelleher, Decoding protein modifications using top-down mass \nspectrometry. Nature methods 4, 817-821 (2007). \n26. J. F. Kellie, R. E. Higgs, J. W. Ryder, A. Major, T. G. Beach, C. H. Adler, K. Merchant, \nM. D. Knierman, Quantitative measurement of intact alpha-synuclein proteoforms from \npost-mortem control and Parkinson's disease brain tissue by intact protein mass \nspectrometry. Sci. Rep. 4, 5797 (2014). \n27. G. K. Tofaris, Initiation and progression of α-synuclein pathology in Parkinson's disease. \nCell Mol Life Sci 79, 210 (2022). \n28. B. Cao, X. Chen, Y. Li, T. Zhou, N. Chen, Y. Guo, M. Zhao, C. Guo, Y. Shi, Q. Wang, X. \nDu, L. Zhang, Y. Li, PDCD4 triggers α-synuclein accumulation and motor deficits via co-\nsuppressing TFE3 and TFEB translation in a model of Parkinson's disease. NPJ \nParkinsons Dis 10, 146 (2024). \n29. G. M. Lloyd, B. Long, S. Quintin, Z. A. Sorrentino, K. M. Gorion, B. M. Bell, D. Carrillo, P. \nSullivan, D. Borchelt, B. I. Giasson, Carboxyl truncation of α-synuclein occurs early and \nis influenced by human APOE genotype in transgenic mouse models of α-synuclein \npathogenesis. Acta Neuropathol Commun 11, 119 (2023). \n30. A. L. Mahul-Mellier, J. Burtscher, N. Maharjan, L. Weerens, M. Croisier, F. Kuttler, M. \nLeleu, G. W. Knott, H. A. Lashuel, The process of Lewy body formation, rather than \nsimply α-synuclein fibrillization, is one of the major drivers of neurodegeneration. Proc. \nNatl. Acad. Sci. U. S. A. 117, 4971-4982 (2020). \n31. I. Walsh, F. Seno, S. C. Tosatto, A. Trovato, PASTA 2.0: an improved server for protein \naggregation prediction. Nucleic Acids Res. 42, W301-307 (2014). \n32. T. Bartels, N. C. Kim, E. S. Luth, D. J. Selkoe, N-alpha-acetylation of alpha-synuclein \nincreases its helical folding propensity, GM1 binding specificity and resistance to \naggregation. PloS one 9, e103727 (2014). \n33. M. D. Watson, J. C. Lee, N-Terminal Acetylation Affects alpha-Synuclein Fibril \nPolymorphism. Biochemistry 58, 3630-3633 (2019). \n34. B. Bu, X. Tong, D. Li, Y. Hu, W. He, C. Zhao, R. Hu, X. Li, Y. Shao, C. Liu, Q. Zhao, B. \nJi, J. Diao, N-Terminal Acetylation Preserves alpha-Synuclein from Oligomerization by \nBlocking Intermolecular Hydrogen Bonds. ACS chemical neuroscience 8, 2145-2151 \n(2017). \n35. M. Runfola, A. De Simone, M. Vendruscolo, C. M. Dobson, G. Fusco, The N-terminal \nAcetylation of alpha-Synuclein Changes the Affinity for Lipid Membranes but not the \nStructural Properties of the Bound State. Scientific reports 10, 204 (2020). \n36. T. G. Beach, A. G. Corbillé, F. Letournel, J. H. Kordower, T. Kremer, D. G. Munoz, A. \nIntorcia, J. Hentz, C. H. Adler, L. I. Sue, J. Walker, G. Serrano, P. Derkinderen, \nMulticenter Assessment of Immunohistochemical Methods for Pathological Alpha-\nSynuclein in Sigmoid Colon of Autopsied Parkinson's Disease and Control Subjects. J. \nParkinsons Dis. 6, 761-770 (2016). \n37. T. G. Beach, C. H. Adler, L. I. Sue, L. Vedders, L. Lue, C. L. White Iii, H. Akiyama, J. N. \nCaviness, H. A. Shill, M. N. Sabbagh, D. G. Walker, Multi-organ distribution of \nphosphorylated alpha-synuclein histopathology in subjects with Lewy body disorders. \nActa Neuropathol. 119, 689-702 (2010). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n30 \n30 \n38. C. Shin, J. Y. Han, S. I. Kim, S. H. Park, H. K. Yang, H. J. Lee, S. H. Kong, Y. S. Suh, H. \nJ. Kim, Y. P. Choi, B. Jeon, In vivo and autopsy validation of alpha-synuclein seeding \nactivity using RT-QuIC assay in the gastrointestinal tract of patients with Parkinson's \ndisease. Parkinsonism Relat. Disord. 103, 23-28 (2022). \n39. K. Knudsen, T. D. Fedorova, J. Horsager, K. B. Andersen, C. Skjærbæk, D. Berg, E. \nSchaeffer, D. J. Brooks, N. Pavese, N. Van Den Berge, P. Borghammer, Asymmetric \nDopaminergic Dysfunction in Brain-First versus Body-First Parkinson's Disease \nSubtypes. J. Parkinsons Dis. 11, 1677-1687 (2021). \n40. P. Borghammer, The brain-first vs. body-first model of Parkinson's disease with \ncomparison to alternative models. J Neural Transm (Vienna) 130, 737-753 (2023). \n41. C. W. Liu, B. I. Giasson, K. A. Lewis, V. M. Lee, G. N. Demartino, P. J. Thomas, A \nprecipitating role for truncated alpha-synuclein and the proteasome in alpha-synuclein \naggregation: implications for pathogenesis of Parkinson disease. J. Biol. Chem. 280, \n22670-22678 (2005). \n42. R. J. Perrin, J. E. Payton, D. H. Barnett, C. L. Wraight, W. S. Woods, L. Ye, J. M. \nGeorge, Epitope mapping and specificity of the anti-alpha-synuclein monoclonal \nantibody Syn-1 in mouse brain and cultured cell lines. Neurosci. Lett. 349, 133-135 \n(2003). \n43. E. Giusto, T. A. Yacoubian, E. Greggio, L. Civiero, Pathways to Parkinson's disease: a \nspotlight on 14-3-3 proteins. NPJ Parkinsons Dis 7, 85 (2021). \n44. X. Shi, Y. Sun, P. Wang, L. Gu, L. Wang, H. Yang, Q. Wei, Z. Li, J. Luo, The interaction \nbetween calcineurin and α-synuclein is regulated by calcium and calmodulin. Biochem \nBiophys Res Commun 496, 1109-1114 (2018). \n45. M. E. Jakobsson, A. Moen, L. Bousset, W. Egge-Jacobsen, S. Kernstock, R. Melki, P. \nFalnes, Identification and characterization of a novel human methyltransferase \nmodulating Hsp70 protein function through lysine methylation. J. Biol. Chem. 288, \n27752-27763 (2013). \n46. R. Gomez-Pastor, E. T. Burchfiel, D. J. Thiele, Regulation of heat shock transcription \nfactors and their roles in physiology and disease. Nat. Rev. Mol. Cell Biol. 19, 4-19 \n(2018). \n47. S. H. Park, N. Bolender, F. Eisele, Z. Kostova, J. Takeuchi, P. Coffino, D. H. Wolf, The \ncytoplasmic Hsp70 chaperone machinery subjects misfolded and endoplasmic reticulum \nimport-incompetent proteins to degradation via the ubiquitin-proteasome system. Mol. \nBiol. Cell 18, 153-165 (2007). \n48. J. Klucken, Y. Shin, E. Masliah, B. T. Hyman, P. J. McLean, Hsp70 Reduces alpha-\nSynuclein Aggregation and Toxicity. J. Biol. Chem. 279, 25497-25502 (2004). \n49. M. Dohmen, S. Krieg, G. Agalaridis, X. Zhu, S. N. Shehata, E. Pfeiffenberger, J. \nAmelang, M. Bütepage, E. Buerova, C. M. Pfaff, D. Chanda, S. Geley, C. Preisinger, K. \nSakamoto, B. Lüscher, D. Neumann, J. Vervoorts, AMPK-dependent activation of the \nCyclin Y/CDK16 complex controls autophagy. Nat Commun 11, 1032 (2020). \n50. Y. Chun, J. Kim, Autophagy: An Essential Degradation Program for Cellular \nHomeostasis and Life. Cells 7,  (2018). \n51. Y. H. Chuang, K. C. Paul, J. M. Bronstein, Y. Bordelon, S. Horvath, B. Ritz, Parkinson's \ndisease is associated with DNA methylation levels in human blood and saliva. Genome \nMed. 9, 76 (2017). \n52. R. Ferrari, D. A. Kia, J. E. Tomkins, J. Hardy, N. W. Wood, R. C. Lovering, P. A. Lewis, \nC. Manzoni, Stratification of candidate genes for Parkinson's disease using weighted \nprotein-protein interaction network analysis. BMC Genomics 19, 452 (2018). \n53. M. Borggrewe, C. Grit, W. F. A. Den Dunnen, S. M. Burm, J. J. Bajramovic, R. J. Noelle, \nB. J. L. Eggen, J. D. Laman, VISTA expression by microglia decreases during \ninflammation and is differentially regulated in CNS diseases. Glia 66, 2645-2658 (2018). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n31 \n31 \n54. R. Shen, S. Lin, L. He, X. Zhu, Z. Zhou, S. Chen, Y. Wang, J. Ding, Association of Two \nPolymorphisms in CCL2 With Parkinson's Disease: A Case-Control Study. Front. Neurol. \n10, 35 (2019). \n55. S. W. Chen, P. Y. Wang, Y. C. Liu, L. Sun, J. Zhu, S. Zuo, J. Ma, T. Y. Li, J. L. Zhang, \nG. W. Chen, X. Wang, Q. R. Zhu, Y. W. Zheng, Z. Y. Chen, Z. H. Yao, Y. S. Pan, Effect \nof Long Noncoding RNA H19 Overexpression on Intestinal Barrier Function and Its \nPotential Role in the Pathogenesis of Ulcerative Colitis. Inflamm. Bowel Dis. 22, 2582-\n2592 (2016). \n56. V. Metta, V. Leta, K. R. Mrudula, L. K. Prashanth, V. Goyal, R. Borgohain, G. Chung-\nFaye, K. R. Chaudhuri, Gastrointestinal dysfunction in Parkinson's disease: molecular \npathology and implications of gut microbiome, probiotics, and fecal microbiota \ntransplantation. J. Neurol. 269, 1154-1163 (2022). \n57. M. L. Bolen, M. Buendia, J. Shi, H. Staley, J. M. Kachergus, P. A. Efron, G. Park, R. \nNagpal, S. D. Alvarez, Q. S. Xue, N. R. McFarland, E. M. Zimmermann, C. E. Forsmark, \nK. B. Menees, A. Salas, E. A. Thompson, M. G. Tansey, Single-cell multiomics identifies \nboth shared and unique features of immune dysfunction in Parkinson's disease and \ninflammatory bowel disease colon, plasma and stool. bioRxiv,  (2025). \n58. M. J. Luesma, I. Cantarero, T. Castiella, M. Soriano, J. M. Garcia-Verdugo, C. Junquera, \nEnteric neurons show a primary cilium. J Cell Mol Med 17, 147-153 (2013). \n59. C. I. Ugbode, I. Smith, B. J. Whalley, W. D. Hirst, M. Rattray, Sonic hedgehog signalling \nmediates astrocyte crosstalk with neurons to confer neuroprotection. J. Neurochem. 142, \n429-443 (2017). \n60. R. Ma, N. A. Kutchy, L. Chen, D. D. Meigs, G. Hu, Primary cilia and ciliary signaling \npathways in aging and age-related brain disorders. Neurobiol. Dis. 163, 105607 (2022). \n61. L. E. Ostrowski, S. K. Dutcher, C. W. Lo, Cilia and models for studying structure and \nfunction. Proc. Am. Thorac. Soc. 8, 423-429 (2011). \n62. B. R. Corradini, P. Iamashita, E. Tampellini, J. M. Farfel, L. T. Grinberg, C. A. Moreira-\nFilho, Complex network-driven view of genomic mechanisms underlying Parkinson's \ndisease: analyses in dorsal motor vagal nucleus, locus coeruleus, and substantia nigra. \nBiomed Res Int 2014, 543673 (2014). \n63. E. Quansah, N. Vatsa, E. Ensink, J. Brown, T. Cave, M. Aguileta, E. Schulz, A. \nLindquist, C. Gilliland, J. A. Steiner, M. L. Escobar Galvis, M. Milčiūtė, M. X. Henderson, \nP. Brundin, L. Brundin, L. L. Marshall, J. Gordevicius, Tet2 loss and enhanced \nciliogenesis suppress α-synuclein pathology. Acta Neuropathol Commun 13, 71 (2025). \n64. Y. E. Lin, E. Jaimon, F. Tonelli, S. R. Pfeffer, Pathogenic LRRK2 mutations cause loss of \nprimary cilia and Neurturin in striatal parvalbumin interneurons. Life Sci Alliance 8,  \n(2025). \n65. L. Concha-Marambio, S. Pritzkow, M. Shahnawaz, C. M. Farris, C. Soto, Seed \namplification assay for the detection of pathologic alpha-synuclein aggregates in \ncerebrospinal fluid. Nat. Protoc. 18, 1179-1196 (2023). \n66. Y. Ma, C. M. Farris, S. Weber, S. Schade, H. Nguyen, A. Pérez-Soriano, D. M. Giraldo, \nM. Fernández, M. Soto, A. Cámara, C. Painous, E. Muñoz, F. Valldeoriola, M. J. Martí, \nJ. Clarimon, P. Kallunki, T. C. Ma, R. N. Alcalay, B. F. Gomes, K. Blennow, H. \nZetterberg, J. Constantinescu, D. Mengel, V. Kadam, P. Parchi, K. Brockmann, T. F. \nTropea, A. Siderowf, M. Synofzik, U. J. Kang, Y. Compta, P. Svenningsson, B. \nMollenhauer, L. Concha-Marambio, Sensitivity and specificity of a seed amplification \nassay for diagnosis of multiple system atrophy: a multicentre cohort study. Lancet \nNeurol. 23, 1225-1237 (2024). \n67. I. Ntai, T. K. Toby, R. D. LeDuc, N. L. Kelleher, A Method for Label-Free, Differential \nTop-Down Proteomics. Methods in molecular biology 1410, 121-133 (2016). \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n32 \n32 \n68. H. D. S. Seckler, L. Fornelli, R. K. Mutharasan, C. S. Thaxton, R. Fellers, M. Daviglus, A. \nSniderman, D. Rader, N. L. Kelleher, D. M. Lloyd-Jones, P. D. Compton, J. T. Wilkins, A \nTargeted, Differential Top-Down Proteomic Methodology for Comparison of ApoA-I \nProteoforms in Individuals with High and Low HDL Efflux Capacity. Journal of proteome \nresearch 17, 2156-2164 (2018). \n69. M. I. Love, W. Huber, S. Anders, Moderated estimation of fold change and dispersion for \nRNA-seq data with DESeq2. Genome Biol. 15, 550 (2014). \n70. W. Zhou, T. J. Triche, Jr., P. W. Laird, H. Shen, SeSAMe: reducing artifactual detection \nof DNA methylation by Infinium BeadChips in genomic deletions. Nucleic Acids Res. 46, \ne123 (2018). \n71. S. van Buuren, K. Groothuis-Oudshoorn, mice: Multivariate Imputation by Chained \nEquations in R. Journal of Statistical Software 45, 1 - 67 (2011). \n72. F. M. Alboukadel Kassambara. (R package, 2020). \n73. M. Kuhn, Building Predictive Models in R Using the caret Package. Journal of Statistical \nSoftware 28, 1 - 26 (2008). \n \nAcknowledgements: We thank the Van Andel Institute’s Pathology and Biorepository, Genomics, \nBioinformatics and Biostatistics Cores, and the Michigan State University Genomics Core for \nassistance with experimentation. We are grateful to the Oregon Brain Bank, Parkinson’s UK Brain \nBank, the NIH NeuroBioBank, LifeNet Health and Corewell Health for providing the tissues used \nin this study. We would like to acknowledge Paul M. Thomas and Eva Boonen for their \ncontributions and support. Finally, we gratefully acknowledge the contributions of the late Dr . \nViviane Labrie, whose scientific vision and passion for advancing neurodegenerative disease \nresearch significantly shaped this work.  \n \nFunding: This work was supported by the NIH -National Institute of Neurological Disorders and \nStroke grant R01NS114409 awarded to LB . The SAA studies were partially funded by \nU24AG079685 grant to CS and a grant from Michael J. Fox Foundation to SP. \n \nAuthor contributions: \nConceptualization: PB, LB \nMethodology: CS, PB, LB, JHK, RDL \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n33 \n33 \nInvestigation: EA, SZ, JWS, MS, PL, MP, SP, BAK \nVisualization: PL, JWS, JG, MP, SP, BAK \nFunding acquisition: PB, LB \nProject administration: CS, PB, LB \nSupervision: RDL, JHK, SP, CS, PB, LB \nWriting – original draft: EA, SZ, JW, BAK \nWriting – review & editing: EA, SZ, JW, PB, LB, BAK \n \nCompeting interests:  \nCS is a Founder, Chief Scientific Officer, consultant and shareholder of Amprion  Inc., a \nbiotechnology company that focuses on the commercial use of seed amplification assays for high-\nsensitivity detection of misfolded protein aggregates involved in various neurodegenerative \ndiseases. Sandra Pritzkow also has a conflict in relation to  Amprion. The University of Texas \nHealth Science Center at Houston has licensed patents and patent applications to Amprion. During \nthe time that this study was being conducted, PB became an employee of F. Hoffmann-La Roche \nand obtained stock in the company  (one of the data were generated by F. Hoffmann -La Roche). \nHe also has ownership interests in Acousort AB, Axial Therapeutics, Enterin Inc and Kenai \nTherapeutics. \n \nData and materials a vailability: Custom code for statistical analysis is available at \nhttps://github.com/lipeipei0611/SynProteoforms. Due to the current U.S. government shutdown, \ndata deposition in publicly accessible repositories has been temporarily delayed. All datasets \nsupporting the findings of this study will be made publicly available in the designated repositories \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n34 \n34 \nas soon as repository services are restored. Data are available from the corresponding author upon \nreasonable request until the repositories reopen. \n \nFigure captions \n \nFigure 1. Alpha -synuclein proteoforms in the appendix and substantia nigra (SN)  (A) Bar \nplot showing 65 alpha-synuclein proteoforms identified in the human appendix and SN. Acetylated \nproteoforms are colored dark blue; unacetylated proteoforms are colored light blue. Nine \nproteoforms, which were only found in the appendix, are denoted with *. (B) Bar plot showing \nalpha-synuclein proteoforms distribution according to their abundance in the human appendix and \nSN separately. X-axis shows the percentage of the abundance for each existing proteoform. Top 5 \nmost abundant proteoforms are listed with thei r color codes. (C) Common alpha-synuclein and \nrare proteoforms showing tissue differences. For common alpha -synuclein proteoforms, robust \nlinear regression was used for statistical analysis adjusting for diagnosis (synucleinopathy/control) \nand age, with proteoform abundance as dependent variable (Y) and tissue as independent variable \n(X).  Proteoforms showing significant tissue differences after FDR q<0.05 are denoted with *. For \nrare alpha-synuclein proteoforms robust linear regression was used for statistical analysis adjusting \nfor diagnosis (synucleinopathy/control) and age, with proteoform detection (TRUE or FALSE) as \ndependent variable (Y) and tissue as independent variable (X).  Only rare proteoforms showing \ntissue difference after FDR q<0.05 are shown and denoted with *. (More in appendix in color dark \ngreen; more in SN in color dark blue). (D) unAc1-140 increased in the appendix, unAc1 -114 \nincreased in the SN. Robust linear regression was used for statistical analysis, adjusting for \ndiagnosis (synucleinopathy/control) and age. \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n35 \n35 \n \nFigure 2. Aggregation propensity and seeding activity of alpha -synuclein. (A) High \naggregation propensity of alpha -synuclein in the synucleinopathy (Syn) appendix compared to \nsubstantia nigra (SN). Aggregation propensity for each proteoform was estimated using P ASTA \n2.0. Robust linear regression was used for statistical analysis adjusting for diagnosis \n(synucleinopathy/control) and age, with aggregation propensity as dependent variable and tissue \nas independent variable. (B) Aggregation kinetics as measured  by the assay demonstrated \nsignificantly increased average ThT fluorescence (AU) in synucleinopathy appendixes (n=16) over \ntime compared to controls (n=15). This suggests the higher seeding activity of misfolded a lpha-\nsynuclein in synucleinopathy appendixes versus controls with no detectable seeding activity. (C) \nThe maximum fluorescence (Fmax) recorded at the plateau of aggregation was significantly higher \nin synucleinopathy appendix as compared to controls (p<0.05).  The graph includes data from all \n3 replicates in each sample. Since positive/negative scores are defined by the number of replicates \npositive, this explains why four individuals have higher average fluorescence, but only one is \nconsidered positive. (D) In enhanced alpha-synuclein-SAA, positive seeding activity was detected \nin 11 synucleinopathy and 1 control appendix, while 5 synucleinopathy and 14 control appendices \nwere negative for seeding activity. Mann -Whitney test was used to analyze average ThT \nfluorescence, Fmax SAA positive activity. Data are presented as mean ± SEM. \n \nFigure 3. Dysregulation of genes and pathways in synucleinopathy patients’ appendices. (A) \nVolcano plot depicting differentially expressed genes in the appendix of synucleinopathy patients \n(n=11) as compared to controls (n=13). The upregulated (red) and downregulated (blue) genes \nwith log2fold-change >2 (p-adj <0.05) are highlighted. The genes with no significant (NS) change \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n36 \n36 \nare shown as grey dots. Gene Set Enrichment Analysis (GSEA) of Gene Ontology (GO) terms \nidentified pathways affected by the DE genes in PD appendices vs controls. (B) GO-Biological \nProcess, (C) GO-Cellular Components, and (D) GO-Molecular Function show activated (left) and \nsuppressed (right) pathways in the PD appendix. Most of the DE genes and dysregulated pathways \nin synucleinopathy appendix contribute to immune responses, protein folding and cilia -related \nprocesses.  \n \nFigure 4. Epigenetic dysregulations in synucleinopathy patients’ appendices. (A) The volcano \nplot displays differential methylation between synucleinopathy appendices (n=11) and controls, \n(n=13) with the x -axis showing beta value differences (synucleinopathy – control). Blue dots \nrepresent observations that were significantly differen t in both p -value and beta value. (B) The \ntotal number of hypomethylated and hypermethylated CpG sites exhibiting more hypomethylation \nin synucleinopathy appendix as compared to controls. (C) The distribution of number of CpG sites \nacross various gene regions in synucleinopathy appendices compared to controls. (D) The \ndistribution of number of CpG sites across CpG regions in synucleinopathy appendices compared \nto controls. \n \nFigure 5. Gene Set Enrichment Analysis (GSEA) of Gene Ontology (GO) terms identified \npathways affected by the DM genes in the synucleinopathy appendices vs controls. (A)  \nHypomethylated pathways mapped to GO database, (B) hypermethylated pathways mapped to GO \ndatabase. (C) Hypomethylated pathways mapped to KEGG database, (D) hypermethylated \npathways mapped to KEGG database. \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n37 \n37 \nFigure 6. Protein –protein interaction network of significantly altered genes in \nsynucleinopathy appendix. Significant genes (p-adj. 0.05) identified from RNA -Seq and DNA \nmethylation analysis were analyzed using the STRING database. The plot excludes genes that lack \nSTRING-annotated functional connectivity. Nodes are colored according to whether the gene was \nidentified via RNA or DNA methylation analysis. Markov Cluster Algorithm (MCL) was used to \norganize nodes, and the top statistically significant functional enrichment (i.e., lowest FDR value) \nwas displayed above each cluster.  \n \nTable 1. Subset of differentially expressed genes linked to synucleinopathy -associated processes \nin appendix samples from synucleinopathy patients vs. controls \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n38 \n38 \nEnsembl ID Gene Name Base \nMean \nlog2Fold \nChange \nSE p value p adj \nENSG00000132141.14 CCT6B 31.181 -2.0081 0.39032 2.68E-07 0.0014 \nENSG00000145911.6 N4BP3 69.835 1.2763 0.29598 1.62E-05 0.0100 \nENSG00000100591.8 AHSA1 866.744 1.4395 0.33509 1.74E-05 0.0100 \nENSG00000110172.12 CHORDC1 3898.817 1.9982 0.47763 2.87E-05 0.0128 \nENSG00000168439.16 STIP1 1651.987 1.5961 0.39216 4.70E-05 0.0169 \nENSG00000144401.14 METTL21A 186.311 1.2672 0.317 6.40E-05 0.0198 \nENSG00000173530.6 TNFRSF10D 272.469 1.1524 0.29374 8.73E-05 0.0227 \nENSG00000213085.10 CFAP45 15.921 -2.9731 0.76333 9.82E-05 0.0239 \nENSG00000204390.10 HSPA1L 77.397 1.5926 0.40962 1.01E-04 0.0239 \nENSG00000117586.11 TNFSF4 59.429 -0.8860 0.22893 1.09E-04 0.0242 \nENSG00000214279.13 SCART1 59.208 -1.2312 0.321 1.25E-04 0.0262 \nENSG00000185507.21 IRF7 96.706 0.8967 0.23455 1.32E-04 0.0263 \nENSG00000118307.19 DNAI7 40.386 -1.8829 0.49214 1.30E-04 0.0263 \nENSG00000137441.8 FGFBP2 23.606 1.6575 0.4342 1.35E-04 0.0265 \nENSG00000168685.15 IL7R 1761.234 1.9616 0.51572 1.43E-04 0.0275 \nENSG00000004478.8 FKBP4 1658.254 1.8554 0.4877 1.42E-04 0.0275 \nENSG00000143401.15 ANP32E 937.318 -0.5640 0.14873 1.49E-04 0.0285 \nENSG00000157578.13 LCA5L 16.816 -1.7741 0.47723 2.01E-04 0.0308 \nENSG00000155875.15 SAXO1 9.332 -2.1946 0.60949 3.17E-04 0.0389 \nENSG00000108691.9 CCL2 790.439 -1.8356 0.51245 3.41E-04 0.0392 \nENSG00000163468.15 CCT3 1434.118 0.8230 0.23049 0.000356 0.0399 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n39 \n39 \n \nTable 2. Subset of differentially methylated genes linked to synucleinopathy-associated processes \nin appendix samples from synucleinopathy patients vs. controls. \ntpCombine Seg Chrm Seg Estimate p value p adj Relation to \nGene \nCategory \nMIP chr12 -0.056460768 7.8464E-07 0.02640 5'UTR \nH19 chr11 -0.045129478 5.1411E-12 2.2613E-06 TSS1500 \nGLI4 chr8 -0.040098331 1.5587E-06 0.02981 TSS1500, \nBody \nGAS7 chr17 -0.031342502 2.5995E-06 0.0420 TSS1500, \n5'UTR \nCDK16 chrX -0.030810321 1.1623E-09 2.5563E-04 TSS1500, \nTSS200 \nTBL1X chrX -0.029706597 2.7826E-08 0.0041 TSS1500, \nTSS200, \nBody \nENSG00000170606.15 HSPA4 1702.360 1.1588 0.32499 3.63E-04 0.0405 \nENSG00000107738.20 VSIR 516.399 0.6449 0.18304 4.26E-04 0.0444 \nENSG00000251595.7 ABCA11P 28.998 -1.1861 0.33727 4.36E-04 0.0451 \nENSG00000150753.12 CCT5 928.511 0.7997 0.22922 4.85E-04 0.0481 \nENSG00000178927.18 CYBC1 379.918 0.7616 0.21822 4.82E-04 0.0481 \nENSG00000086061.16 DNAJA1 2138.092 1.3378 0.38485 5.08E-04 0.0491 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n40 \n40 \nTMC4 chr19 -0.02341605 2.9657E-06 0.0432 TSS200 \nEIF1AX chrX -0.011564424 1.4196E-06 0.0284 5'UTR \nWBP11 chr12 -0.00582025 5.0605E-06 0.0495 TSS1500 \nNR4A1 chr12 -0.005543784 4.9188E-06 0.0495 Body \nMRPL41 chr9 -0.005405291 4.7806E-07 0.0264 1stExon, \nBody \nTEX22 chr14 -0.005314138 3.9308E-07 0.0264 TSS1500 \nMTA1 chr14 -0.005314138 3.9308E-07 0.0264 TSS1500 \nSH3GLB1 chr1 -0.005202031 2.0330E-06 0.0358 TSS1500, \nBody \nAKR7A2 chr1 -0.005023715 4.0374E-06 0.0487 TSS1500 \nSLC66A1 chr1 -0.005023715 4.0374E-06 0.0487 TSS1500 \nANXA5 chr4 -0.004686528 3.4229E-06 0.0470 TSS200, \n5'UTR \nZMYND11 chr10 -0.004486106 6.0076E-07 0.0264 TSS1500, \nTSS200, \nBody \nDPH3 chr3 -0.004301509 1.1084E-06 0.0284 TSS1500 \nOXNAD1 chr3 -0.004301509 1.1084E-06 0.0284 TSS1500 \nUBTF chr17 -0.003862467 2.7495E-06 0.0420 TSS1500 \nNAP1L1 chr12 -0.003581127 8.3028E-07 0.0264 TSS1500, \nTSS200, \n5'UTR \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n41 \n41 \nACBD3 chr1 -0.003069002 3.9148E-06 0.0487 5'UTR \nABCB9 chr12 -0.003039661 8.9398E-08 0.0079 TSS1500 \nLSM2 chr6 -0.002908889 4.2098E-06 0.0487 TSS1500, \nTSS200, \n5'UTR \nDHX29 chr5 -0.002533829 1.0731E-06 0.0284 TSS1500 \nMTREX chr5 -0.002533829 1.0731E-06 0.0284 TSS1500 \nTRIM33 chr1 -0.002213311 1.3577E-06 0.0284 5'UTR \nTOR3A chr1 -0.00137959 3.8652E-06 0.0487 TSS200, \nBody \nDNMT3A chr2 -1.27370184049748E-\n04 \n4.1074E-06 0.0487 TSS1500, \n5'UTR \nTRNP1 chr1 -5.70366549201819E-\n05 \n4.7327E-06 0.0495 TSS1500, \nTSS200, 1st \nExon, Body \nPPP2R2B chr5 0.019478798 2.3741E-06 0.0402 TSS1500, \nTSS200, \n5'UTR \nR3HDM2 chr12 0.033485722 5.0267E-06 0.0495 5'UTR \nDCUN1D1 chr3 0.071012065 1.2839E-06 0.0284 5'UTR \nMEI1 chr22 0.351120112 6.7254E-07 0.0264 Body \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint \n\n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted October 7, 2025. ; https://doi.org/10.1101/2025.10.07.680938doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}