LIBX-A401: A Novel Selective Inhibitor of Acyl-CoA Synthetase Long Chain Family Member 4 (ACSL4) and Its Binding Mode

Abstract

Acyl-coenzyme A synthetase long-chain family member 4 (ACSL4), a pivotal enzyme in lipid metabolism, has emerged as a therapeutic target for ferroptosis-related conditions and cancer. However, its reference inhibitor, rosiglitazone, has off-target activity on peroxisome proliferator-activated receptor gamma (PPARγ), a key regulator of lipid homeostasis. Here, the discovery of LIBX-A401, a potent ACSL4 inhibitor derived from rosiglitazone devoid of PPARγ activity, is reported. Its binding to ACSL4 is ATP-dependent, stabilizing the C-terminal domain and altering the fatty acid gate region, as shown by Hydrogen-Deuterium Exchange Mass Spectrometry. Photoaffinity labeling identified A329 within the fatty acid binding site, while molecular dynamics and mutagenesis highlighted Q302 as critical for LIBX-A401 binding. LIBX-A401 exhibits anti-ferroptotic properties in cells, supported by target engagement. These findings establish LIBX-A401 as a valuable tool to study ACSL4 in ferroptosis and cancer, while its elucidated binding mode paves the way for the rational design of improved inhibitors.

Keywords: ACSL4 inhibitors; Ferroptosis; Hydrogen Deuterium Exchange (HDx) mass spectrometry; Parkinson's disease; Photoaffinity labeling.

Figure 1

Evaluation of compound 9 against ACSL4. a) Structure of 9 and summary of its binding and inhibitory activity profile against ACSL4 as well as its selectivity over ACSL3 and PPARγ. b) MST traces, with and without 1 mm ATP (in blue and orange, respectively), showing that the binding of 9 to ACSL4 depends on the prior binding of ATP. A K _D value of 720 nM was obtained for 9 in the presence of 1 mm ATP. c) nDSF traces with and without 1 mm ATP (in blue and orange, respectively), confirming that ATP is required for the binding of compound 9. A ΔT _m value of 5.3 °C was obtained for 9 (5 µm) in the presence of 1 mm ATP. K _D and ΔT _m values (± SD) were determined by three independent experiments for each compound concentration. Curves in panels B and C are shown as the mean ± SD values. IC₅₀ values (± SD) were determined by three independent experiments performed in duplicate for each compound concentration.

Figure 2

HDx‐MS results with compound 9 in the presence of ATP. Changes in deuterium uptake at different exchange times are presented for the identified peptides, which are mapped onto the refined AlphaFold model of ACSL4. C‐terminus and N‐terminus of ACSL4 are depicted in dark gray and white, respectively. Peptides A and B, along with a selected peptide in the C‐terminal region closest in proximity to these two, are shown in cyan. Uptake plots for each illustrated peptide show the impact of compound 9 on deuterium incorporation. Peptide traces shown represent mean ± SD of three independent experiments. Statistically significant using t‐test from HDExaminer, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001.

Figure 3

Photoaffinity labeling identification of compound 9 binding site in ACSL4. a) The structure of photoaffinity labeling probe 10 and representation of its binding peptides (orange) on the refined AlphaFold ACSL4 model for which the number of peptide‐spectrum matches (PSMs) identified are shown. The previously identified peptide A, included in identified peptide C, is represented in cyan. b) Results of the competition assay in targeted MS/MS analysis for peptide C with and without compound 9. c) Results of the labeling assay with and without ATP in targeted MS/MS analysis for peptide C. Data shown are representative of three independent experiments. Graphs in panels B and C are shown as the mean ± SD values. Statistically significant using unpaired t‐test, ^**** p < 0.0001.

Figure 4

Putative binding mode of compound 9 on ACSL4. a) ASCL4‐compound 9 complex after docking and an additional 100 ns MD simulation. b) Representative trace of the contacts that compound 9 forms with the ACSL4 model during the 100 ns MD simulation, showing that compound 9 primarily interacts through hydrophobic contacts and hydrogen bonds with Q302. Each vertical line corresponds to an intermolecular contact, and the color is associated with the interaction type (e.g., green corresponds to a hydrophobic contact and red to a hydrogen bond).

Figure 5

Impact of site‐directed mutations on the inhibitory activity and affinity of compound 9 for ACSL4. a) Inhibition profiles of compound 9 against wild‐type ACSL4 and selected mutants (Q302A in orange, Q302 M in blue, and L325F in purple) showing that the site‐directed mutations affect inhibitory activity of 9. b) Binding curves from MST experiments showing the impact of site‐directed mutations (Q302A in orange, Q302 m in blue, and L325F in purple) on the affinity of compound 9 for ACSL4. c) Table summarizing IC₅₀ and K _D values (± SD) of compound 9 against wild‐type ACSL4 and selected mutants. IC₅₀ values and curves were determined by three independent experiments performed in duplicate for each compound concentration. Similarly, K _D values and curves were determined by three independent experiments for each compound concentration. Curves in panels B and C are shown as the mean ± SD values.

Figure 6

Compound 9 protects cells from ferroptosis. a) Cell viability analysis of HEK293 cells pretreated with 2.5 µm of 9 for 24 h, followed by treatment with RSL3 for 48 h. Data represent mean and SD (n = 4). Statistical analysis was performed using Tukey's multiple comparisons test. ^**** p< 0.001. b) Cell viability analysis of HT‐1080 cells pretreated with 2.5 µm of 9 for 24 h, followed by treatment with RSL3 for 48 h. Data represent mean and SD (n = 4). Statistical analysis was performed using Tukey's multiple comparisons test. ^**** p< 0.001. c) Cell viability analysis of LUHMES cells transfected with ACSL4 siRNA or pretreated with 2.5 µm of 9 for 4 h, followed by treatment with AA + Fe for 48 h. Data represent mean and SD (n = 3). Statistical analysis was performed using Tukey's multiple comparisons test. ^**** p< 0.001. d) Percentage of lipid peroxidation in LUHMES cells transfected with ACSL4 siRNA or pretreated with 2.5 µm of 9 for 4 h, followed by treatment with AA + Fe for 24 h. Lipid peroxidation in cells was measured by flow cytometry using the C11 BODIPY 581/591 probe. The staining data obtained at 530 nm (oxidized C11 BODIPY 581/591) are plotted as a histogram. Data represents mean and SEM (n = 4). Statistical analysis was performed using Tukey's multiple comparisons test. ^**** p< 0.001. e) Representative western blots showing the thermal stability of ACSL4 in LUHMES cells treated with either DMSO or 10 µm compound 9. ACSL4 levels were quantified by immunoblotting. From three independent experiments, T _m values of 48.3 ± 0.5 °C (DMSO) and 50.9 ± 0.1 °C (compound 9) were determined, corresponding to a ΔT _m of 2.6 ± 0.6 °C. f) Evaluation of radical‐trapping antioxidant activity of 9 (50 µm) compared to trolox (50 µm) using the DPPH assay. Data represent mean and SD (n = 3).