rs199347 — GPNMB

Deep inside the cells of your brain, a protein called GPNMB (glycoprotein nonmetastatic melanoma protein B) is quietly managing one of the most critical housekeeping operations in the nervous system: keeping lysosomes functional. Lysosomes are the cellular recycling centers that break down damaged proteins, worn-out organelles, and cellular debris. In neurons — cells that can live for a century and cannot simply divide to replace themselves — lysosomal health is not optional. It is the difference between a neuron that ages gracefully and one that accumulates toxic protein aggregates until it dies.

The rs199347 variant sits in an intron of GPNMB on chromosome 7 and acts as a potent [expression quantitative trait locus (eQTL) | An eQTL is a genetic variant that controls how much of a nearby gene is transcribed into mRNA, without changing the protein sequence itself] in the brain. People carrying the common A allele produce measurably more GPNMB mRNA in the cerebral cortex and putamen — and this increase in GPNMB expression is directly linked to Parkinson's disease risk at genome-wide significance, confirmed in studies encompassing over 400,000 participants.

GPNMB plays a dual role in aging biology that has only recently come into focus. First, it is a lysosomal integrity protein: GPNMB physically binds ATP6V1A, a component of the vacuolar ATPase proton pump that maintains the acidic environment lysosomes need to function. When GPNMB is absent or malfunctioning, the V0 and V1 domains of this pump dissociate, lysosomal acidity is compromised, and protein degradation fails. Suda et al. 2022¹¹ Suda et al. 2022
Glycoprotein nonmetastatic melanoma protein B regulates lysosomal integrity and lifespan of senescent cells. Sci Rep, 2022 demonstrated that in senescent cells — which accumulate with age throughout the body — GPNMB is upregulated through the TFEB/MITF transcription factor axis as a protective response to lysosomal stress. Cells that cannot make GPNMB senesce faster; cells with extra GPNMB resist stress-induced senescence. This makes GPNMB a survival factor for senescent cells, which has a double-edged implication for aging: while it protects individual cells, it may also help abnormal senescent cells persist when they should be cleared.

Second, GPNMB is the molecular doorman for alpha-synuclein (aSyn) in neurons. Diaz-Ortiz et al. 2022²² Diaz-Ortiz et al. 2022
GPNMB confers risk for Parkinson's disease through interaction with alpha-synuclein. Science, 2022 showed that GPNMB physically binds to aSyn — the protein that misfolds and aggregates into the toxic Lewy bodies that define Parkinson's disease — and that neurons require GPNMB to internalize aSyn fibrils from their environment. When GPNMB is genetically eliminated in iPSC-derived neurons, the cells lose the ability to take up aSyn fibrils entirely. The implication is that elevated GPNMB in A-allele carriers creates neurons that are more efficient at importing aSyn, which may accelerate the cell-to-cell spreading of aSyn pathology that drives Parkinson's disease progression.

Adding to this picture, GPNMB is tightly linked to progranulin (PGRN), another lysosomal protein encoded by GRN. When progranulin levels fall — as occurs in frontotemporal dementia-causing GRN mutations — GPNMB expression in macrophages surges. GPNMB is upregulated as a compensatory response to lysosomal dysfunction, making it both a biomarker of lysosomal stress and a participant in the neuroinflammatory response that characterizes multiple neurodegenerative diseases.

The genetic case for rs199347 as a Parkinson's disease risk variant is exceptionally well-powered. Chang et al. 2017³³ Chang et al. 2017
A meta-analysis of genome-wide association studies identifies 17 new Parkinson's disease risk loci. Nature Genetics, 2017 identified rs199347-A in the largest PD GWAS conducted to that point: 26,035 cases and 403,190 controls across discovery and replication phases. The association reached p=4×10⁻¹⁸ (far beyond genome-wide significance at p<5×10⁻⁸), with OR=1.10 (95% CI 1.08–1.12). This was independently confirmed in Nalls et al. 2014⁴⁴ Nalls et al. 2014
Large-scale meta-analysis of genome-wide association data identifies six new risk loci for Parkinson's disease. Nature Genetics, 2014 (13,708 cases, 95,282 controls; OR=1.11, p=1×10⁻¹²).

The mechanistic link was established by Murthy et al. 2017⁵⁵ Murthy et al. 2017
Increased brain expression of GPNMB is associated with genome wide significant risk for Parkinson's disease on chromosome 7p15.3. Neurogenetics, 2017, which used four independent brain eQTL datasets (Braineac, CAGEseq, GTEx, PheGenI) and 134 brain samples to confirm that the major A allele at rs199347 consistently drives higher GPNMB mRNA levels across all datasets, most prominently in cortical regions and the putamen. GPNMB expression in temporal cortex was approximately 2.4-fold higher than in cerebellum, reflecting strong regional specificity relevant to Parkinson's pathology. The authors concluded that elevated GPNMB expression, rather than protein sequence change, is the causative link between this locus and disease risk.

Proteogenomic evidence from Kaiser et al. 2023⁶⁶ Kaiser et al. 2023
A proteogenomic view of Parkinson's disease causality and heterogeneity. NPJ Parkinson's Disease, 2023 independently nominated GPNMB as the single top causal protein for PD neuroinflammatory pathology via Mendelian randomization across 804 patients with combined genomic and proteomic data. In clinical samples, GPNMB protein is measurably elevated in PD plasma and cerebrospinal fluid, with levels correlating with disease severity and genotype at rs199347. Brody et al. 2024⁷⁷ Brody et al. 2024
GPNMB Biomarker Levels in GBA1 Carriers with Lewy Body Disorders. Mov Disord, 2024 confirmed that rs199347 functions as a protein quantitative trait locus (pQTL) — not merely an mRNA eQTL — with GPNMB levels differing significantly by genotype in both plasma (p=0.022) and CSF (p=0.007).

The overall effect size (OR ~1.10 per A allele) is modest, consistent with a polygenic common variant that contributes to population-level PD risk without deterministically causing disease. At the population level, however, the A allele is the major allele in Europeans (~59%), meaning this variant contributes substantially to attributable risk across the population.

Parkinson's disease has no genetic test that predicts disease onset with certainty from a single variant of this effect size. The practical value of knowing your rs199347 genotype lies in motivating specific protective behaviors with established evidence in neuroprotection and lysosomal health.

The strongest behavioral lever is exercise. Multiple prospective cohort studies, including a meta-analysis of over 1 million participants, show that regular vigorous physical activity reduces PD incidence by 25–30%. Exercise upregulates TFEB-driven autophagy and lysosomal biogenesis — directly opposing the lysosomal stress that GPNMB is compensating for — and improves dopaminergic neuron resilience independently of genetics. For AA carriers, this is the most evidence-based neuroprotective intervention available.

Lysosomal health is also supported through dietary patterns rich in polyphenols (especially resveratrol and quercetin) that activate TFEB and autophagy, and through time-restricted eating protocols that induce autophagy through mTOR suppression. Both approaches target the upstream lysosomal biology that rs199347 modulates. Caffeine deserves specific mention: it is the only dietary factor with consistent epidemiological evidence of reduced PD risk (OR ~0.70 across multiple studies), and it operates through adenosine receptor blockade that reduces dopaminergic neuron vulnerability — a distinct mechanism from lysosomal biology.

rs199347 is biologically adjacent to rs356182 in SNCA (alpha-synuclein). GPNMB mediates aSyn internalization, while rs356182 modulates aSyn expression. Carriers with elevated GPNMB expression (AA at rs199347) who also carry elevated aSyn expression risk (rs356182 risk genotype) face a potential double burden: more aSyn protein being produced, and more efficient cellular machinery for importing extracellular aSyn fibrils into neurons — the two-hit scenario that drives propagation of Lewy body pathology. This interaction is biologically compelling but has not been formally tested in combined genotype studies.

rs75932628 in TREM2 is a related neuroinflammatory longevity SNP. GPNMB operates partly through microglial neuroinflammation, as does TREM2. Elevated GPNMB in AA carriers may compound with TREM2-mediated microglial dysfunction to produce a more severe neuroinflammatory phenotype in aging brain tissue.