Beneficial Effects of the Direct AMP-Kinase Activator PXL770 in In Vitro and In Vivo Models of X-Linked Adrenoleukodystrophy

Abstract

X-linked adrenoleukodystrophy (ALD) is a severe orphan disease caused by mutations in the peroxisomal ABCD1 transporter gene, leading to toxic accumulation of Very Long-Chain Fatty Acids (VLCFA - in particular C26:0) resulting in inflammation, mitochondrial dysfunction and demyelination. AMP-activated protein kinase (AMPK) is downregulated in ALD, and its activation is implicated as a therapeutic target. PXL770 is the first direct allosteric AMPK activator with established clinical efficacy and tolerability. Methods: We investigated its effects in ALD patient-derived fibroblasts/lymphocytes and Abcd1 KO mouse glial cells. Readouts included VLCFA levels, mitochondrial function and mRNA levels of proinflammatory genes and compensatory transporters (ABCD2-3). After PXL770 treatment in Abcd1 KO mice, we assessed VLCFA levels in tissues, sciatic nerve axonal morphology by electronic microscopy and locomotor function by open-field/balance-beam tests. Results: In patients' cells and Abcd1 KO glial cells, PXL770 substantially decreased C26:0 levels (by ∼90%), improved mitochondrial respiration, reduced expression of multiple inflammatory genes and induced expression of ABCD2-3 In Abcd1 KO mice, PXL770 treatment normalized VLCFA in plasma and significantly reduced elevated levels in brain (-25%) and spinal cord (-32%) versus untreated (P < 0.001). Abnormal sciatic nerve axonal morphology was also improved along with amelioration of locomotor function. Conclusion: Direct AMPK activation exerts beneficial effects on several hallmarks of pathology in multiple ALD models in vitro and in vivo, supporting clinical development of PXL770 for this disease. Further studies would be needed to overcome limitations including small sample size for some parameters, lack of additional in vivo biomarkers and incomplete pharmacokinetic characterization. SIGNIFICANCE STATEMENT: Adrenoleukodystrophy is a rare and debilitating condition with no approved therapies, caused by accumulation of very long-chain fatty acids. AMPK is downregulated in the disease and has been implicated as a potential therapeutic target. PXL770 is a novel clinical stage direct AMPK activator. In these studies, we used PXL770 to achieve preclinical validation of direct AMPK activation for this disease - based on correction of key biochemical and functional readouts in vitro and in vivo, thus supporting clinical development.

Fig. 1.

PXL770 treatment improves VLCFA levels in C-ALD and AMN patient-derived fibroblasts. Fibroblasts were isolated from individual patients with AMN (A) and C-ALD (B, C). (A) AMN fibroblasts were exposed for 7 days to PXL770 at 0.1, 0.5, 1, 2, 3.5, 5, 10, 25, and 50 μM or to Metformin at 100, 200, 300, and 400 μM. VLCFA levels were analyzed by LC-MS after extraction of total lipids from pelleted cells. Fibroblasts from a healthy donor, a C-ALD patient (B) or an AMN patient (C) were exposed to control media or the indicated concentrations of PXL770 followed by LC-MS measurements of specific lipids and the ratio of C26:0/C22:0 VLCFA as described for (A). Data are mean ± S.E.M. n = 5 replicates for (A), 3 for (B), and 5 for (C). Similar results were obtained in two additional independent experiments with these same patient donor cell lines. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 versus C-ALD/AMN untreated (Dunnett’s test).

Fig. 2.

PXL770 mediated improvement of VLCFA levels is associated with restored mitochondrial function and increased compensatory gene expression in C-ALD and AMN patients’ fibroblasts. Fibroblasts were isolated from C-ALD and AMN patients and exposed for 72 hours to PXL770 at 5, 10, 25 or 50μM. (A) OCR was measured using a Seahorse Analyzer and bioenergetic parameters were evaluated by sequential additions of: oligomycin (Oligo - 1 μM), FCCP (0.25 μM) and Rotenone-Antimycin A (Rot-AA - 1 μM). Basal = first three measurements, ATP-linked = OCR drop after oligo addition, MOC = OCR after addition of FCCP. Data are mean ± SEM n = 6 replicates per patient cell line. (B) ABCD2 and 3 mRNA levels evaluated by RT-qPCR, normalized by RLP27 expression (no unit). Data are mean ± SEM n = 4-6 replicates per patient cell line. *P ≤ 0.05, **P > 0.01, ***P < 0.001, ****P < 0.0001 (Dunnett’s test versus untreated).

Fig. 3.

PXL770 treatment improves VLCFA levels, restores mitochondrial function and increases compensatory gene expression in Abcd1 KO mouse glial cells. Mixed glial cells (astrocyte-enriched) were isolated from 1-day-old wt and Abcd1 KO mice. 10 days after culture cells were exposed to PXL770 at 5, 10, 25 or 50 μM. (A) Cells were exposed for 7 days to the treatment and VLCFA levels were analyzed by LC-MS after extraction of total lipids from pelleted cells. Data are mean ± S.E.M., n = 2 replicates for wild-type and n = 3 replicates for Abcd1 KO and PXL770. (B) Cells were exposed for 72 hours to the treatment and the OCR was measured using a Seahorse Analyzer. Bioenergetics parameters were evaluated by sequential additions of: oligomycin (1 μM), FCCP (0.5 μM) and Rotenone-Antimycin A (1 μM). Basal = first three measurements, ATP-linked = OCR drop after oligo addition, Maximal Oxidative capacity = after addition of FCCP. Data are mean ± SEM, n = 6 replicates. (C) Cells were exposed for 72 hours to the treatment and mRNA levels were evaluated by RT-qPCR, normalized by Rlp27 expression (no unit). Data are mean ± SEM, n = 3–5 replicates. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 (Dunnett’s test versus untreated).

Fig. 4.

PXL770 treatment improves neuroinflammation gene expression markers in lymphocytes from a C-ALD patient and in Abcd1 KO mouse glial cells. (A) Lymphocytes were isolated from a C-ALD patient and (B) Mixed glial cells (astrocyte-enriched) were isolated from 1-day-old wt and Abcd1 KO mice and matured for 10 days, then stimulated (TI) with TNFα (10ng/ml) and IL1β (10ng/ml) for 70 hours (during PXL770 treatment). Cells were exposed to PXL770 10 μM for 72 hours and mRNA levels were evaluated by RT-qPCR, normalized by RLP27/Rlp27 expression (no unit). Data are mean ± S.E.M., n = 3 replicates. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 (Dunnett’s or Kruskal Wallis test versus TI untreated – test wt versus TI untreated not shown).

Fig. 5.

PXL770 improves VLCFA levels in Abcd1 KO mice. VLCFA content measured by LC-MS in (A) plasma and (B) brain from 6- to 8-week-old Abcd1 KO mice treated with PXL770 at 75 mg/kg (BID – Oral) for 8 weeks. Data are mean ± SEM, n = 12 animals for wild-type and n = 15 for KO untreated and PXL770. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (Dunnett’s test versus untreated).

Fig. 6.

PXL770 treatment improves VLCFA levels in spinal cord, axonal morphology of sciatic nerve and locomotor function in Abcd1 KO mice. 13-month-old Abcd1 KO mice were treated with PXL770 75 mg/kg (BID oral) for 12 weeks. (A) VLCFA content measured by LC-MS in spinal cord. Data are mean ± SEM, n = 8 animals/condition. (B) Axonal morphology of myelin and neurons determined by morphometric analysis of transversal slices of the sciatic nerve by EM (magnification 800X). Data are mean ± SEM, n = 4 animals per condition. Behavioral and locomotor function assessment by (C) open field test and (D) Balance beam test. Data are median ± 95% interval confidence, n = 8 animals for wild-type and KO untreated, 7 animals for PXL770. *P < 0.05, **P < 0.01, ***P < 0.001 ****P < 0.0001 (Dunnett’s test versus untreated).