Elevated 15-PGDH levels were observed in both in vitro and in vivo PD models, as well as in postmortem human PD brain tissue. Both pharmacologic inhibition and genetic partial reduction of 15-PGDH mitigated motor deficits, oxidative damage, neuroinflammation, and dopaminergic neuron loss across MPTP, intranigral LPS, and α-synuclein PD mouse models. Notably, 15-PGDH-mediated protection in the α-synuclein PD mouse model occurred without any change in phosphorylated α-synuclein accumulation, demonstrating that therapeutic benefit can be achieved independently of this aspect of synuclein pathology. This parallels our recent finding that 15-PGDH-mediated neuroprotection in an amyloid-based AD mouse model was similarly independent of alteration in amyloid pathology [12], contributing to a growing body of evidence that potent therapeutic effect can be achieved by targeting the brain's damage and inflammatory response to the classically considered primary drivers of disease [[45], [46], [47]]. Mechanistically, bulk RNA-seq and confirmatory studies revealed that 15-PGDH inhibition downregulates genes associated with oxidative stress and neuroinflammation, including Cybb, Lcn2, and Il1β. Consistent with these findings, 15-PGDH inhibition suppressed microglial and astrocytic activation and protected against oxidative damage in PD models.

We utilized both pharmacologic inhibition (SW033291 treatment) and genetic partial reduction of 15-PGDH. SW033291 has high specificity for inhibiting 15-PGDH, with IC50 < 1 nM and no interaction with other structurally related short-chain dehydrogenases [6]. Nonetheless, to control for any previously unidentified off-target drug effects, we compared 15-PGDH inhibition with SW033291 versus genetic inhibition of 15-PGDH via knockout of one Hpgd allele.

Elevated 15-PGDH levels were observed in both in vitro and in vivo PD models, as well as in postmortem human PD brain tissue. Both pharmacologic inhibition and genetic partial reduction of 15-PGDH mitigated motor deficits, oxidative damage, neuroinflammation, and dopaminergic neuron loss across MPTP, intranigral LPS, and α-synuclein PD mouse models. Notably, 15-PGDH-mediated protection in the α-synuclein PD mouse model occurred without any change in phosphorylated α-synuclein accumulation, demonstrating that therapeutic benefit can be achieved independently of this aspect of synuclein pathology.

In the MPTP study, mice underwent a one-week acclimation period prior to receiving intraperitoneal injections of the 15-PGDH inhibitor SW033291 (0.5 or 5 mg/kg, twice daily) for two days, after which treatment was continued alongside daily MPTP injections (30 mg/kg) for seven more days.

Behavioral assessments were performed at designated time points, and brains were collected for biochemical and histological analysis at 33 days post-injection (Fig. 4A). Consistent with our previous reports [39,40], the combination of AAV-α-synuclein and PFF administration induced pronounced motor dysfunction, reflected by shortened rotarod latency (Fig. 4B), elevated hind-limb clasping scores (Fig. 4C), and prolonged pole test times (Fig. 4D). SW033291 treatment prevented these impairments and maintained motor performance comparable to AAV-eGFP/PBS controls, with no change in body weight (Fig. 4B–D and Fig. S8A).Fig. 415-PGDH inhibition ameliorates motor deficits, dopaminergic neuronal loss, and oxidative stress in the ɑ-synuclein mouse PD model(A) Experimental procedure for evaluating the efficacy of SW033291 in the α-synuclein mouse model.(B) SW033291 protects mice from AAV-ɑ-synuclein/PFF-induced motor deficits in the rotarod test (n = 5-6 per group, ∗∗p < 0.01, one-way ANOVA and Tukey's post hoc analysis).(C) SW033291 protects mice from AAV-ɑ-synuclein/PFF-induced from hindlimb clasping behavior (n = 5-6 per group, ∗p < 0.05, ∗∗p < 0.01, one-way ANOVA and Tukey's post hoc analysis).(D) SW033291 protects mice from AAV-ɑ-synuclein/PFF-induced increased time to orient downward (time to turn) and total time to descend to the floor (time to go down) in the pole test (n = 5-6 per group, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, one-way ANOVA and Tukey's post hoc analysis).(E) Representative images and quantification of TH immunoreactivity in the substantia nigra of mice treated with vehicle or SW033291. SW033291 administration protected mice from AAV-ɑ-synuclein/PFF-induced loss of TH intensity (scale bar = 800 μm, n = 5-6 per group, ∗p < 0.05, one-way ANOVA and Tukey's post hoc analysis).(F) Representative images and quantification of 4-HNE immunoreactivity in the substantia nigra of mice treated with vehicle or SW033291. SW033291 administration protected mice from AAV-ɑ-synuclein/PFF-induced increase in 4-HNE (scale bar = 100 μm, n = 5-6 per group, ∗p < 0.05, one-way ANOVA and Tukey's post hoc analysis).(G) Representative images and quantification of 3-NT immunoreactivity in the substantia nigra of mice treated with vehicle or SW033291. SW033291 administration protected mice from AAV-ɑ-synuclein/PFF-induced increase in 3-NT (scale bar = 100 μm, n = 5-6 per group, ∗p < 0.05, ∗∗∗p < 0.001, one-way ANOVA and Tukey's post hoc analysis).(H) RT-qPCR analysis shows that SW033291 treatment protected mice from administration protected mice from AAV-ɑ-synuclein/PFF-induced increase in Cybb mRNA expression (n = 3 per group, ∗∗p < 0.01, ∗∗∗p < 0.001, one-way ANOVA and Tukey's post hoc analysis).(I) Representative images and quantification of Iba1 immunoreactivity in the substantia nigra of mice treated with vehicle or SW033291. SW033291 administration protected mice from AAV-ɑ-synuclein/PFF-induced increase in Iba1 (scale bar = 200 μm, n = 5-6 per group, ∗∗p < 0.01, ∗∗∗p < 0.001, one-way ANOVA and Tukey's post hoc analysis).(J) Representative images and quantification of GFAP immunoreactivity in the substantia nigra of mice treated with vehicle or SW033291. SW033291 administration protected mice from AAV-ɑ-synuclein/PFF-induced increase in GFAP (scale bar = 300 μm, n = 5-6 per group, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, one-way ANOVA and Tukey's post hoc analysis).(K) RT-qPCR analysis shows that SW033291 administration protected mice from AAV-ɑ-synuclein/PFF-induced increase in Lcn2 mRNA expression (n = 3 per group, ∗p < 0.05, ∗∗∗p < 0.001, one-way ANOVA and Tukey's post hoc analysis).(L) Representative images and quantitative analysis of p-α-synuclein immunoreactivity in the substantia nigra of mice treated with vehicle or SW033291. SW033291 treatment had no effect on p-α-synuclein accumulation in AAV-ɑ-synuclein/PFF mice (scale bar = 20 μm, n = 5-6 per group, ∗∗∗∗p < 0.0001, one-way ANOVA and Tukey's post hoc analysis).

Male C57BL/6J mice were purchased from Daehan BioLink Co., Ltd (Korea) and maintained at the Animal Center for Pharmaceutical Research of Seoul National University under temperature, light, and humidity-controlled conditions with free access to food and water. Hpgd +/- males and females were bred and maintained under the same conditions as C57BL/6J mice. All animal work was approved by the Institutional Animal Care and Use Committee (IACUC, SNU-221026-6-5) of Seoul National University.

The enzyme 15-hydroxyprostaglandin dehydrogenase (15-PGDH) occupies a central regulatory role in bioactive lipid metabolism, catalyzing nicotinamide adenine dinucleotide (NAD+)-dependent degradation of prostaglandins and autocoids, with prostaglandin E2 (PGE2) representing its prototypical substrate [[5], [6], [7], [8]]. Our prior work identified SW033291 as a potent small-molecule 15-PGDH inhibitor that enhances tissue regeneration by activating stem cell populations in peripheral organs [6]. Subsequent studies revealed that 15-PGDH inhibition attenuates oxidative stress and inflammation, conferring protection against renal and hepatic injury, albeit by incompletely understood mechanisms [[9], [10], [11]].

Here, we establish the therapeutic potential of 15-PGDH inhibition in PD, a mechanistically different neurodegenerative disorder. We observed significant 15-PGDH upregulation in the substantia nigra of postmortem human PD brains as well as three different mouse models of PD: systemic 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) [13], intranigral lipopolysaccharide (LPS) [14], intrastriatal AAV-α-synuclein with intra-ventral tegmental area α-synuclein preformed fibrils (PFFs) [15]. We utilized both pharmacologic inhibition (SW033291 treatment) and genetic partial reduction of 15-PGDH. SW033291 has high specificity for inhibiting 15-PGDH, with IC50 < 1 nM and no interaction with other structurally related short-chain dehydrogenases [6]. Nonetheless, to control for any previously unidentified off-target drug effects, we compared 15-PGDH inhibition with SW033291 versus genetic inhibition of 15-PGDH via knockout of one Hpgd allele.

Postmortem human substantia nigra tissue from PD subjects displayed elevated mRNA for Hpgd (the gene encoding 15-PGDH) (Fig. 1A; Table S1). Similarly, exposure of mice to the neurotoxin MPTP, which selectively targets the substantia nigra dopaminergic neurons implicated in PD [[17], [18], [19], [20]], also increased Hpgd mRNA levels, 15-PGDH protein levels, and 15-PGDH enzymatic activity in the substantia nigra (Fig. 1B–D). Hpgd mRNA was similarly elevated in mice receiving intranigral LPS, intrastriatal AAV-α-synuclein with intra-ventral tegmental area α-synuclein PFFs, and intrastriatal PFF (Fig. 1E–G). Lastly, Hpgd mRNA induction was observed in SH-SY5Y cells that were transduced to overexpress wild-type α-synuclein, a cellular model of PD (Fig. 1H).

(A) HPGD mRNA is increased in the substantia nigra of human PD brain, relative to control subjects (n = 5-8 per group, ∗p < 0.05, unpaired t-test).

Here, we establish the therapeutic potential of 15-PGDH inhibition in PD, a mechanistically different neurodegenerative disorder. We observed significant 15-PGDH upregulation in the substantia nigra of postmortem human PD brains as well as three different mouse models of PD: systemic 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) [13], intranigral lipopolysaccharide (LPS) [14], intrastriatal AAV-α-synuclein with intra-ventral tegmental area α-synuclein preformed fibrils (PFFs) [15].

Mice subsequently received intranigral injections of either AAV-GFP (comparator) or AAV–α-synuclein (disease model), followed by intra-VTA injection of PBS (comparator) or α-synuclein preformed fibrils (PFFs) (disease model).

We utilized both pharmacologic inhibition (SW033291 treatment) and genetic partial reduction of 15-PGDH. SW033291 has high specificity for inhibiting 15-PGDH, with IC50 < 1 nM and no interaction with other structurally related short-chain dehydrogenases [6]. Nonetheless, to control for any previously unidentified off-target drug effects, we compared 15-PGDH inhibition with SW033291 versus genetic inhibition of 15-PGDH via knockout of one Hpgd allele. Both drug and genetic inhibition of 15-PGDH showed complete concordance in protecting from motor deficits, dopaminergic cell loss, BBB degradation, and induction of multiple molecular markers of inflammation and ROS-induced tissue damage.

In the MPTP study, mice underwent a one-week acclimation period prior to receiving intraperitoneal injections of the 15-PGDH inhibitor SW033291 (0.5 or 5 mg/kg, twice daily) for two days, after which treatment was continued alongside daily MPTP injections (30 mg/kg) for seven more days. SW033291 significantly reduced 15-PGDH activity and increased PGE2 levels in the substantia nigra without affecting body weight (Fig. S1A–C). Behavioral assessments were conducted 2–3 h after the final MPTP injection (Fig. 2A).

Mice pretreated with SW033291 (5 mg/kg, twice daily) or vehicle received bilateral substantia nigra LPS injections, with continued SW033291 treatment (Fig. 2I).

Seven-week-old C57BL/6J mice were acclimated for one week and then pretreated with either vehicle or the 15-PGDH inhibitor SW033291 (5 mg/kg, intraperitoneally, twice daily) for two days prior to surgery. Mice subsequently received intranigral injections of either AAV-GFP (comparator) or AAV–α-synuclein (disease model), followed by intra-VTA injection of PBS (comparator) or α-synuclein preformed fibrils (PFFs) (disease model). Behavioral assessments were performed at designated time points, and brains were collected for biochemical and histological analysis at 33 days post-injection (Fig. 4A).

MPTP administration induced significant motor deficits, evidenced by reduced rotarod latency (Fig. 2B), increased hind-limb clasping (Fig. 2C and Fig. S1D and E), and prolonged pole test completion times (Fig. 2D). Notably, SW033291 treatment dose-dependently prevented all these impairments, preserving motor performance to near non-MPTP exposed levels (Fig. 2B–D and Fig. S1E).

After MPTP exposure, heterozygous Hpgd ±/- mice exhibited superior motor performance compared to WT Hpgd+/+ littermates, demonstrating longer rotarod retention (Fig. 2F), reduced hindlimb clasping (Fig. 2G and Fig. S1G), and faster pole test completion (Fig. 2H).

Immunohistochemical analysis revealed a significant reduction in tyrosine hydroxylase (TH), a marker of dopaminergic neurons, in the substantia nigra of MPTP-treated mice compared to vehicle-treated controls (Fig. 2M and N).

To investigate the neuroprotective mechanism of 15-PGDH inhibition in PD, we performed bulk RNA sequencing of the substantia nigra from SW033291-treated versus vehicle-treated mice in both the MPTP and intranigral LPS-induced PD models. Joint analysis of the SW033291 effect across both models revealed that 15-PGDH inhibition upregulated 126 genes (p < 0.05 and Log2(FC) > 0.5) and downregulated 367 genes (p < 0.05 and Log2(FC) < -0.5) (Fig. S3A; Table S2).

Inspection of GO biological processes associated with SW033291 downregulated genes (Fig. S3C; Table S4) and of individual genes downregulated by SW033291 (Fig. S3A and D; Table S2) identified SW033291 targeting of several highly pathogenic known disease mediators. First, SW033291 treatment decreased levels of both Cybb and Cyba, which respectively encode gp91phox (commonly called NOX2) and gp22phox, cooperating subunits of the ROS generating NADPH oxidase complex [23]. SW033291 treatment also decreased expression of Lcn2, an inflammation-associated protein that mediates dopaminergic neuronal death and glial activation [[24], [25], [26]]. Furthermore, 15-PGDH inhibition markedly reduced the levels of the pro-inflammatory cytokine gene Il1β. While modulation of each of these gene targets was individually statistically significant (Fig. S3A), they did not reach a threshold for FDR, multiple testing corrected, significance (Table S2). Accordingly, to further investigate these candidate SW033291 effectors, we generated samples from new and independent validation sets of animals, as shown below.

To further delineate the basis of these 15-PGDH pathway driven effects, we evaluated regulation of NOX2 expression and superoxide production in BV2 murine microglial cells. PGE2, the prototypical 15-PGDH degradation substrate, dose-dependently inhibited LPS-induced NOX2 upregulation (Fig. 3K). Furthermore, PGE2 also completely inhibited LPS-induced generation of superoxide (Fig. 3K), that was fully derived from NOX2, as shown by its blockade by the NOX2 inhibitor GSK2795039 (Fig. S4E).

PGE2 signals through four receptor types: EP1, EP2, EP3, and EP4 [35]. To determine which receptor mediates the protective effect of PGE2, we employed receptor-specific antagonists. While EP1, EP2, and EP3 antagonists failed to block PGE2's suppression of LCN2, the EP4-specific antagonist L-161982 dose-dependently inhibited the protective effect of PGE2 on LCN2 expression in BV2 cells exposed to LPS (Fig. S6E–H). These results demonstrate that PGE2 attenuation of LPS-driven LCN2 expression in microglia is predominantly via the PGE2-EP4 signaling pathway.

Further investigation into the regulatory mechanisms governing Hpgd expression will be important for understanding the upstream processes that drive 15-PGDH elevation in PD. While in vitro experiments have demonstrated that PGE2 can reduce LCN2 and NOX2 expression, in vivo studies will be necessary to fully interrogate the interactions among these signaling pathways. Additional experiments, including in vivo EP4 knockout and inhibition, NOX2 silencing, and LCN2 neutralization, will further validate these findings and clarify their mechanistic and therapeutic implications. In addition, further validation in additional human patient samples will be important in future studies.

Immunofluorescence analysis demonstrated a marked increase in ionized calcium-binding adapter molecule 1 (Iba1) intensity, a microglial marker, in the substantia nigra and striatum of MPTP-exposed mice, which was attenuated by SW033291 (Fig. S5A–D). To assess astrocyte reactivity, we analyzed glial fibrillary acidic protein (GFAP) expression, a canonical marker of reactive astrocytes [33], and observed that 15-PGDH inhibition also dose-dependently reduced GFAP-positive astrocytes in these regions in MPTP-exposed mice (Fig. S5E–H). Critically, 15-PGDH inhibition also suppressed these markers of microglial and astrocytic activation in the intranigral LPS model (Fig. S5I–L).

Electron microscopy revealed that pharmacological inhibition of 15-PGDH with SW033291 significantly reduced MPTP-induced structural damage to capillary endothelia in the substantia nigra and striatum (Fig. S7A). To confirm functional BBB compromise, we assessed parenchymal infiltration of endogenous immunoglobulin G (IgG), a marker of barrier leakage. MPTP administration markedly increased IgG deposition in brain tissue, an effect fully abrogated by SW033291 treatment (Fig. S7B).

Notably, SW033291 treatment did not significantly reduce phosphorylated α-synuclein accumulation (Fig. 4L), indicating that 15-PGDH inhibition primarily alleviates oxidative stress and neuroinflammation independent of α-synuclein pathology, and that this is sufficient to eliminate the motor impairment associated with PD.

Additional experimental details (MPTP administration, intranigral LPS administration, stereotaxic injection of α-syn PFF with AAV-human α-syn, stereotaxic injection of α-syn PFF, RT-qPCR, western blotting, 15-PGDH enzyme activity, behavioral analysis, immunohistochemistry, quantification of immunohistochemistry, measurement of MAO-B activity, bulk RNA sequencing, superoxide measurement, electron microscopy, tissue sample preparation for prostaglandin extraction, LC-MS/MS-based prostaglandin analysis, quantification and statistical analysis) are provided in Supplemental Materials.

(N) Quantification of TH fluorescence signal shows that SW033291 treatment dose-dependently protects MPTP-exposed mice from loss of TH intensity (n = 4 per group, ∗p < 0.05, ∗∗∗∗p < 0.0001, one-way ANOVA and Tukey's post hoc analysis).

Further investigation into the regulatory mechanisms governing Hpgd expression will be important for understanding the upstream processes that drive 15-PGDH elevation in PD. While in vitro experiments have demonstrated that PGE2 can reduce LCN2 and NOX2 expression, in vivo studies will be necessary to fully interrogate the interactions among these signaling pathways. Additional experiments, including in vivo EP4 knockout and inhibition, NOX2 silencing, and LCN2 neutralization, will further validate these findings and clarify their mechanistic and therapeutic implications. In addition, further validation in additional human patient samples will be important in future studies.

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:Yeojung Koh has patent issued to n/a. Edwin Vazquez-Rosa has patent issued to n/a. Sanford D. Markowitz has patent issued to n/a. Andrew A. Pieper has patent issued to n/a. Min-Kyoo Shin has patent issued to n/a. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Encouragingly, both pharmaceutical and biotechnology companies have initiated development of 15-PGDH inhibitors for peripheral indications, and inhibitor MF-300 has already completed human phase 1 trials. Our results now provide the rationale to repurpose such agents for the treatment of PD.