α-Synuclein-specific T cell reactivity is associated with preclinical and early Parkinson’s disease

The study investigates whether α-synuclein-specific T cell reactivity is linked to the preclinical and early motor phases of Parkinson’s disease (PD). PD is characterized by a prolonged prodromal period with accumulating dopaminergic neuron loss in the substantia nigra pars compacta before motor symptoms lead to diagnosis. Prodromal non-motor symptoms (hyposmia, constipation, mood changes, REM sleep behavior disorder) can precede diagnosis by years to decades but lack specificity. Disease-modifying treatments may fail when initiated after substantial neuronal loss, underscoring the need for early biomarkers. Prior work by the authors demonstrated α-syn-specific T cells in some PD patients, suggesting autoimmune features. The central hypothesis is that α-syn-specific T cell responses emerge in the preclinical phase, peak around the time of motor diagnosis, and decline thereafter, potentially serving as an early disease indicator and informing pathogenesis.

Evidence supports extensive dopaminergic neuron loss in early motor PD, with stereological analyses showing 50–90% loss of SNpc neurons within the first four years after diagnosis and marked reductions in neuromelanin-pigmented cells and dopaminergic markers, consistent with a preceding prodromal phase. Prodromal symptoms are common but non-specific. Inflammation is implicated in PD pathogenesis, and classical autoimmune diseases show dynamic antigen-specific T cell responses over disease course. The authors’ prior study identified α-synuclein epitope-specific T cells in PD, raising the possibility of autoimmunity. The current work builds on this by examining temporal dynamics of α-syn-specific T cells, comparing PD to healthy controls and to Alzheimer’s disease (as a neurodegenerative control), and assessing clinical correlates (age, sex, HLA, cognition, motor severity, medication).

Design: Two-pronged approach: (1) a longitudinal case study of a single individual with PD who had banked peripheral blood mononuclear cell (PBMC) samples spanning years before and after motor PD diagnosis; (2) cross-sectional analyses of PD patients and age-matched healthy controls (HC) from multiple sites, with an independent PD validation cohort, plus an Alzheimer’s disease (AD) cohort with matched controls. Participants and cohorts: PD and HC were recruited at UC San Diego (UCSD), Rush University Medical Center (RUMC/Chicago), University of Alabama at Birmingham (UAB), and La Jolla Institute (LJI). Cross-sectional comparisons included, for certain analyses, PD (n ≈ 77) and HC (n ≈ 69), and broader demographics reported PD (n = 97) and HC (n = 68). Time-since-diagnosis analyses used two cohorts: Cohort 1 (n = 76 PD) and Cohort 2 (n = 26 PD), combined n = 96 PD. An AD cohort (n = 38) and age-matched HC (n = 41) were used to assess disease specificity. Clinical data included age, sex, HLA type, MoCA, UPDRS (including Part III), Hoehn & Yahr stage (limited spread), and levodopa equivalent dose (LED). Sample processing and assays: PBMCs were cultured for 14 days with an α-synuclein peptide epitope pool (previously defined set; typically 10–12 peptides; 5 µg/ml). After expansion, cells were restimulated and cytokine responses measured by triple-color FluoroSpot (IFNγ, IL-5, IL-10), expressed as spot-forming cells (SFC) per 10^6 cultured PBMC. Positive pools were deconvoluted to identify reactive peptides where cell numbers allowed. Intracellular cytokine staining (ICS) with flow cytometry characterized responding cell subsets (markers included CD3, CD4, CD8, CD14, CD19, CD56; cytokines IFNγ, IL-4, IL-10). Regulatory T cell markers CD25 and CD127 were used to assess Treg involvement among IL-10–producing cells. The longitudinal case study used similar 14-day expansion and epitope mapping across multiple timepoints before and after diagnosis. HLA typing: Performed by an ASHI-accredited lab at Murdoch University, with multi-locus genotyping (A, B, C, DRB1) and downstream analysis. Outcomes and thresholds: Primary outcome was magnitude of α-syn-specific T cell responses (sum of IFNγ, IL-5, IL-10 SFC per 10^6 PBMC). An a priori arbitrary responder threshold of 250 SFC was applied for some analyses. Temporal association with years since PD diagnosis (<10 vs ≥10 years) and relationships with age, sex, HLA, cognitive/motor scores, and LED (<1000 vs ≥1000 mg/day) were evaluated. Statistics: Nonparametric tests were used given non-normal distributions. Mann–Whitney tests compared groups; Fisher’s exact tests compared responder frequencies; Spearman correlations assessed associations (e.g., time since diagnosis vs LED). Significance defined as p < 0.05.

- Longitudinal case study: Elevated α-syn-specific T cell responses were detectable years prior to motor PD diagnosis and declined after diagnosis. Responses mapped to a specific α-syn epitope (predicted to bind HLA-DRB1*0101 at high affinity, ~71 nM). Cytokine production was mediated predominantly by CD4+ T cells (IFNγ and IL-4). - PD vs HC: α-syn-specific T cell reactivity was significantly higher in PD than HC when considering peptide- and participant-specific combinations (p < 0.0001). Using a pooled-peptide response and a 250 SFC cutoff, PD had a higher proportion of responders (p = 0.02, Fisher’s exact). - Disease specificity: No significant difference in α-syn-specific T cell reactivity between AD patients and their controls (p = 0.15 Mann–Whitney; p = 0.21 Fisher’s exact at 250 SFC), suggesting the signal is not a general feature of neurodegeneration. - Temporal dynamics: Reactivity was inversely related to time since diagnosis, being highest near the time of diagnosis and waning thereafter. Across combined cohorts, 39.7% (29/73) of PD donors diagnosed <10 years responded (>250 SFC) vs 8.6% (3/23) diagnosed ≥10 years (p = 0.022). - Cytokine-specific patterns: IFNγ responses showed a significant temporal association: 26/73 responders <10 years vs 1/23 ≥10 years (p = 0.036). IL-5 and IL-10 showed similar trends (17/73 and 12/73 responders <10 years, respectively, vs 1/23 each ≥10 years) but were not statistically significant (p = 0.283 and p = 0.179). - Cellular sources: IFNγ and IL-4 production arose mainly from CD4+ T cells; IL-10 was produced by both CD4+ and CD8+ T cells. IL-10 production was not associated with CD25+CD127− Tregs; instead, a CD25−CD127− T cell population accounted for α-syn-specific IL-10. - Clinical correlates: Age positively correlated with α-syn reactivity in PD (patients >70 years responded more frequently; p = 0.0177) but not in HC (p = 0.51). A non-significant trend suggested higher responses in males (38%) than females (23%) (p = 0.2279). No correlations were found with MoCA or UPDRS scores. No HLA allele association with PD or responder status survived multiple-testing correction. - Medication and synergy with disease duration: Low LED (<1000 mg/day) significantly associated with α-syn-specific responses; LED correlated with time since diagnosis (Spearman r = 0.335, p = 0.016). Responders were largely confined to those with both recent diagnosis (<10 years) and low LED: 25/28 responders met both criteria, whereas 33/58 nonresponders did not (p = 0.0029). - Classification: Combining age, LED, and time since diagnosis yielded a classification accuracy of nearly 70% for distinguishing PD from non-PD in this dataset. Overall, α-syn-reactive T cells are most abundant in preclinical and early motor PD and decline with advancing disease duration.

The findings support a temporal model in which α-synuclein-specific T cell responses emerge before the onset of motor symptoms, peak around diagnosis, and diminish as PD progresses. This pattern aligns with autoimmune paradigms in which antigen-specific T cells may contribute to early tissue damage but become less detectable later, potentially due to immune regulation, antigen availability changes, or immunosenescence. The absence of elevated α-syn reactivity in AD suggests disease specificity rather than a generic neurodegeneration-associated response. The predominance of CD4+ T cell-derived IFNγ and IL-4 indicates classical helper T-cell involvement, while IL-10 production from non-Treg CD4+ and CD8+ T cells suggests concurrent regulatory or compensatory responses. Clinical correlations (increased responses with age in PD, association with low LED, and strongest responses in recent diagnoses) further indicate that immune reactivity is tied to early disease stage rather than disease severity per se (no associations with MoCA or UPDRS). Lack of robust HLA associations in this diverse, multi-site cohort suggests that α-syn-reactive T cell responses may not be strongly restricted to specific HLA alleles at the population level, though larger studies may be required. These results underscore the relevance of adaptive immunity in PD pathogenesis and highlight α-syn-specific T cell reactivity as a candidate biomarker for early or preclinical PD.

This study confirms that α-synuclein-reactive T cells are associated with PD and are most prevalent immediately before and after motor diagnosis, declining with time from diagnosis. A longitudinal case demonstrated pre-diagnostic α-syn-specific responses, while cross-sectional cohorts showed higher reactivity in early PD relative to controls and AD. Reactivity correlated with older age in PD and lower LED, and was primarily mediated by CD4+ T cells (IFNγ/IL-4), with IL-10 arising from non-Treg CD4+ and CD8+ cells. These insights support a role for adaptive immunity in early PD and suggest that monitoring α-syn-specific T cell responses could aid early identification of at-risk individuals (e.g., those with REM sleep behavior disorder or PD-associated genetic variants) and inform immunotherapeutic strategies. Future work should include longitudinal, multi-timepoint studies from prodromal stages, integration of broader antigen panels, larger HLA- and TCR-repertoire analyses, and validation of predictive models combining immune and clinical variables.

- Cross-sectional design for most analyses limits causal inference and precise temporal mapping; only one individual had true longitudinal pre- and post-diagnosis samples. - Potential misdiagnosis of early PD (estimated 15–29%) could confound associations. - Sex imbalance between PD and control cohorts may affect comparisons. - Limited PBMC availability prevented full epitope deconvolution for all donors; an arbitrary 250 SFC cutoff was used for responder classification. - Time since motor diagnosis may not reflect true disease onset or prodromal duration. - Heterogeneity across recruitment sites and assays may introduce variability; some methodological descriptions indicate complexity and potential batch effects. - HLA association analyses may be underpowered for detecting weak effects after multiple-testing correction. - Findings in AD do not exclude potential reactivity in other neurodegenerative diseases; broader disease controls were not assessed.