Enhancing glycolysis attenuates Parkinson's disease progression in models and clinical databases

Abstract

Parkinson's disease (PD) is a common neurodegenerative disease that lacks therapies to prevent progressive neurodegeneration. Impaired energy metabolism and reduced ATP levels are common features of PD. Previous studies revealed that terazosin (TZ) enhances the activity of phosphoglycerate kinase 1 (PGK1), thereby stimulating glycolysis and increasing cellular ATP levels. Therefore, we asked whether enhancement of PGK1 activity would change the course of PD. In toxin-induced and genetic PD models in mice, rats, flies, and induced pluripotent stem cells, TZ increased brain ATP levels and slowed or prevented neuron loss. The drug increased dopamine levels and partially restored motor function. Because TZ is prescribed clinically, we also interrogated 2 distinct human databases. We found slower disease progression, decreased PD-related complications, and a reduced frequency of PD diagnoses in individuals taking TZ and related drugs. These findings suggest that enhancing PGK1 activity and increasing glycolysis may slow neurodegeneration in PD.

Keywords: Neuroscience; Parkinson’s disease.

Figure 1. TZ enhances glycolysis in the mouse brain.

Data points represent individual mice. Blue indicates controls and red indicates TZ treatment. (A) Schematic of ATP production by glycolysis and oxidative phosphorylation. (B) Schematic time course for experiments in C–G. Eight-week-old C57bl/6 mice were given MPTP (20 mg/kg i.p.) or vehicle 4 times at 2-hour intervals. Then, TZ (10 μg/kg) or vehicle was injected i.p. once a day for 1 week. Assays were performed on day 7. (C–E) Pyruvate levels (C), citrate synthase (CS) activity (D), and ATP levels (E) were measured in mouse striatum. TZ doses are indicated in E. n = 6. Statistical comparison was made versus no TZ treatment. (F and G) Pyruvate (F) and ATP (G) levels in the mouse striatal region. Supplemental Table 3 shows statistical tests and P values for all comparisons. Bars and whiskers indicate the mean ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001, by Mann-Whitney U test (C and D), Kruskal-Wallis with Dunn’s test (E), and Kruskal-Wallis with Dwass-Steele-Critchlow-Fligner test (F and G).

Figure 2. TZ improves dopamine neuron and motor function in MPTP-treated mice.

(A) Schematic for experiments in panels B–K. Eight-week-old C57BL/6 mice received 4 i.p. injections of MPTP (20 mg/kg at 2-hour intervals) or vehicle on day 0. Mice were then injected with TZ (10 μg/kg) or vehicle (0.9% saline) once a day for 1 week, and assays were performed on day 7. Other mice began receiving daily TZ or vehicle injections beginning on day 7, and assays were performed on day 14. n = 6. (B–D) Example of Western blots with TH and β-actin (protein loading control) in striatum and SNc on days 7 and 14 (B). Quantification of TH protein normalized to control (C and D). n = 6. (E) Example of immunostaining of TH in SNc and striatum. Scale bars: 100 μm (SNc) and 1 mm (striatum). Quantification of TH-positive neurons in SNc (F) and TH intensity in the striatum (G). n = 6. (H and I) Dopamine (DA) content in striatum and SnC. n = 6. (J) Percentage of TH-positive neurons that were positive for TUNEL staining. n = 6. (K) Behavioral response of mice in the rotarod test. Data reflect the duration that the mice remained on an accelerated rolling rod, normalized to mice on day 0. n = 8. Data represent examples and indicate the mean ± SEM. Blue indicates control and red indicates TZ treatment. *P < 0.05 and **P < 0.01, by Mann-Whitney U test for days 7 and 14. Supplemental Table 3 shows P values for all comparisons.

Figure 3. TZ slows neurodegeneration, increases dopamine, and improves motor performance in 6-OHDA–treated rats.

(A) Schematic for experiments in B–G. 6-OHDA (20 μg) was injected into the right striatum of rats on day 0. TZ (70 μg/kg) or saline was injected i.p. daily for 2 weeks, beginning 2, 3, 4, or 5 weeks after 6-OHDA injection. Assays were performed at 0 and 2–7 weeks. (B) Percentage of TUNEL-positive SNc cells. n = 6. (C) Quantification of TH protein levels assessed by immunoblotting in the striatum, normalized to control. n = 6. (D and E) Percentage of SNc cells positive for TH immunostaining (D) and intensity of TH immunostaining in striatum (E) 7 weeks after 6-OHDA injection. TZ treatment was administered from week 5 to week 7. n = 6. (F) Dopamine content in the right striatum relative to the left (control) striatum. n = 6. (G) Results of the cylinder test. 6-OHDA was injected into the right striatum, impairing use of the left paw. The assay was performed 7 weeks after 6-OHDA injection. TZ treatment was given from week 5 to week 7. n = 4 for control group and n = 10 for the two 6-OHDA groups. In C, D, E, and G, data points represent individual rats, and bars and whiskers indicate the mean ± SEM. Blue indicates controls and red indicates TZ treatment. Supplemental Table 3 shows statistical tests and P values for all comparisons. *P < 0.05, **P < 0.01, and ***P < 0.001, by Mann-Whitney U test (B and F), Kruskal-Wallis with Dwass-Steele-Critchlow-Fligner test (C, D, and E), and Friedman with Dunn’s test (G).

Figure 4. TZ enhances Pgk activity to attenuate rotenone-impaired motor performance.

(A) Schematic for experiments in panels B–F. Flies received rotenone (125 or 250 μM in food) with TZ (1 μM) or vehicle for 7 or 14 days. (B) Relative ATP content in the brains of w¹¹¹⁸ flies that received 250 μM rotenone with or without TZ for 14 days. n = 6, with 200 fly heads for each treatment in each trial. (C) Climbing behavior of flies after 250 μM rotenone with TZ (1 μM) or vehicle for 7 days. Data show the percentage of flies that climbed up a tube (see Methods). n = 3, with 200 flies tested for each treatment in each trial. (D) Knockdown of Pgk in offspring of actin-Gal4 crossed with UAS-Pgk RNAi flies. Offspring of actin-Gal4 crossed with y1 v1 P [CaryP] attP2 were used as a genetic background matched control. n = 3, with RNA collected from 30 fly heads for each sample. (E) Pgk was knocked down in TH neurons by crossing UAS-Pgk RNAi flies with flies carrying the TH neuron-specific promoter (TH-Gal4) to produce TH>Pgk RNAi flies. Rotenone (250 μM) and TZ were administered as indicated for 7 days. Climbing behavior was measured on day 7. n = 8, with 200 flies tested for each treatment in each trial. (F) Pgk (UAS-Pgk) overexpression was driven by a dopaminergic neuron promoter (TH-Gal4), a pan-neuronal promoter (Appl-Gal4), a pan-cell promoter (Actin-Gal4), and a muscle-specific promoter (Mhc-Gal4). Rotenone (250 μM) was administered for 7 days, and climbing behavior was measured on day 7. n = 3, with 200 flies tested for each treatment in each trial. Data points represent individual groups of flies, and bars and whiskers show the mean ± SEM. Blue indicates controls and red indicates TZ treatment. Supplemental Table 3 shows statistical tests and P values for all comparisons. *P < 0.05, **P < 0.01, and ***P < 0.001, by Kruskal-Wallis with a Dwass-Steele-Critchlow-Fligner test (B), 1-way ANOVA with Tukey’s test (C and E), paired t test (D), and unpaired t test (F).

Figure 5. TZ improves TH levels and motor performance in genetic models of PD.

Data points are from individual mice and groups of flies. (A–E) WT (w¹¹¹⁸) and PINK1⁵ flies received TZ or vehicle for 10 days beginning on the first day after eclosion. Day 10 assays included: (A) Example of wing posture defect and percentage of w¹¹¹⁸ and PINK1⁵ flies with wing posture defects. n = 6, with 80 flies for each treatment in each trial. (B and C) Example of TH Western blot (B) and quantification of TH (C). n = 5, with 40 fly heads for each treatment in each trial. (D) ATP content in brains (relative to w¹¹¹⁸). n = 3, with 200 fly heads for each treatment in each trial. (E) Climbing behavior of flies. n = 3, with 100 flies for each treatment in each trial. (F) Climbing behavior of LRRK^ex1 male flies. n = 6, with 100 flies for each treatment in each trial. (G–K) TZ administration to mThy1-hSNCA–transgenic mice. (G) Schematic for experiments in panels H–K. (H) Example of Western blot of α-synuclein in striatum and SNc. (I and J) Quantification of α-synuclein in striatum and SNc. n = 5. (K) Duration that mice remained on an accelerating rotarod. n = 5. Data are from individual groups of flies (A–F) and individual mice (I–K). Bars and whiskers indicate the mean ± SEM. Blue indicates controls and red indicates TZ treatment. Supplemental Table 3 shows statistical tests and P values for all comparisons. *P < 0.05, **P < 0.01, and ***P < 0.001, by 1-way ANOVA with Tukey’s test (D) and Kruskal-Wallis with a Dwass-Steele-Critchlow-Fligner test (A–C, E, F, and I–K).

Figure 6. TZ increases ATP content and decreases α-synuclein accumulation in iPSC-derived dopamine neurons from patients with PD.

(A) iPSC-derived dopamine neurons from 2 patients with PD (subjects 12 and 13) carrying LRRK2^G2019S mutations and a healthy control (subject 11). Thirty-day-old dopamine neurons were plated and were treated with TZ (10 μM) 1 or 3 days later. The neurons were studied 24 hours after addition of TZ. We observed no difference between the 2 start dates and therefore combined the data. Representative immunofluorescence images of α-synuclein (SNCA, green), TH (red), and DAPI (nuclei, blue). (B) Percentage of TH-positive neurons with cytoplasmic accumulation of α-synuclein. n = 12. (C) ATP content in control and LRRK2^G2019S iPSC–derived dopamine neurons. n = 12. Bars and whiskers indicate the mean ± SEM. Blue indicates controls and red indicates TZ treatment. Supplemental Table 3 shows statistical tests and P values for all comparisons. *P < 0.05, **P < 0.01, and ***P < 0.001, by Mann-Whitney U test.

Figure 7. TZ and related drugs slow the progression of motor defects for patients with PD enrolled in the PPMI database.

Movement Disorder Society–Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) Part 3 (motor) scores for patients with PD in the PPMI database. Patients were taking TZ/DZ/AZ (blue, n = 13), tamsulosin (green, n = 24), or none of these drugs (red, n = 269). Data represent scores upon entry into the PPMI database through approximately 1 year and include all measures between those times. All patients taking these drugs were men prescribed TZ/DZ/AZ or tamsulosin, without breaks for benign prostatic hyperplasia or undefined urological problems. Lines are plotted from linear mixed-effect regression analyses. By maximum likelihood estimation, TZ/DZ/AZ differed from controls (P = 0.012).

Figure 8. TZ and related drugs reduce symptoms as assessed by diagnostic codes for patients with PD in the Truven/IBM Watson clinical database.

Data are from the Truven Health Marketscan Commercial Claims and Encounters and Medicare Supplemental Databases for the years 2011–2016. Patients had a diagnosis of PD and were prescribed TZ/DZ/AZ or tamsulosin for at least 1 year. We assessed RRs for 79 previously identified PD-related diagnostic codes. (A) RR for 79 PD-related diagnostic codes for patients taking TZ/DZ/AZ versus tamsulosin. Yellow indicates a statistically significant difference in risk between TZ/DZ/AZ and tamsulosin (P < 0.05) determined by a generalized linear model with a quasi-Poisson distribution. (B) RR for the categories of PD-related diagnostic codes for patients taking TZ/DZ/AZ versus tamsulosin. Data represent the mean and 95% CIs.