Discovery of dual rho-associated protein kinase 1 (ROCK1)/apoptosis signal–regulating kinase 1 (ASK1) inhibitors as a novel approach for non-alcoholic steatohepatitis (NASH) treatment

In the current study we suggest a novel approach to curb non-alcoholic steatohepatitis (NASH) progression, and we suggest privileged scaffolds for the design of novel compounds for this aim. NASH is an advanced form of non-alcoholic fatty liver disease that can further progress into fibrosis, cirrhosis, and hepatocellular carcinoma. It is a widely emerging disease affecting 25% of the global population and has no current approved treatments. Protein kinases are key regulators of cellular pathways, of which, Rho-associated protein kinase 1 (ROCK1) and apoptosis signal–regulating kinase 1 (ASK1) play an important role in the progression of NASH and they stand out as promising targets for NASH therapy. Interestingly, their kinase domains are found to be similar in sequence and topology; therefore, dual inhibition of ROCK1 and ASK1 is expected to be amenable and could achieve a more favourable outcome. To reach this goal, a training set of ROCK1 and ASK1 protein structures co-crystalized with type 1 (ATP-competitive) inhibitors was constructed to manually generate receptor-based pharmacophore models representing ROCK1 and ASK1 inhibitors’ common pharmacophoric features. The models produced were assessed using a test set of both ROCK1 and ASK1 actives and decoys, and their performance was evaluated using different assessment metrics. The best pharmacophore model obtained, showing a Mathew’s correlation coefficient (MCC) of 0.71, was then used to screen the ZINC purchasable database retrieving 6178 hits that were filtered accordingly using several medicinal chemistry and pharmacokinetics filters returning 407 promising compounds. To confirm that these compounds are capable of binding to the target kinases, they were subjected to molecular docking simulations at both protein structures. The results were then assessed individually and filtered, setting the spotlight on various privileged scaffolds that could be exploited as the nucleus for designing novel ROCK1/ASK1 dual inhibitors. Supplementary Information The online version contains supplementary material available at 10.1186/s13065-023-01081-3.


S1. Selection of the X-ray crystallographic structures for ROCK1 and ASK1 and training set generation
Co-crystalized ROCK1 inhibitors interact mainly at three region: the hinge region, the catalytic residue Lys105, and gate area.They interact with the hinge region and the catalytic residue Lys105 via classical hydrogen bonds and hydrophobic interactions.Interactions with Asp216 of the DFG-motif at the gate area were also observed in very few structures.Focusing on the hinge region, interactions frequently occur with Tyr155 and Met156 and to a lesser extent with Gln154.Most inhibitors also interact with the catalytic residue, Lys105.Moreover, the inhibitors are involved in hydrophobic interactions with the hydrophobic residues Leu82 and Val90 as well.
ASK1 inhibitors show a similar binding pattern interacting with the hinge region and the catalytic residue Lys709 via classical hydrogen bonds and hydrophobic interactions.
Interactions with Asp822 of the DFG-motif at the gate area were observed in one inhibitor only.At the hinge region, ASK1 inhibitors usually interact with Gly756 and Val757 and less frequently with Met754 and Gln755.All inhibitors interacted with the catalytic Lys709.
Furthermore, most of the inhibitors interacted with Leu686.
Afterward, the inhibitors were grouped according to their observed interactions after their analysis individually.Overall, ROCK1 inhibitors of the selected group interact through hydrogen bonds with Tyr155 and Met156 (at the hinge region), Lys105 (catalytic residue), and hydrophobic interactions through aromatic rings.On the other hand, ASK1 selected inhibitors interacted with the kinase domain through hydrogen bond interactions at Val757 (hinge region), Lys709 (catalytic residue), and various hydrophobic interactions at Leu686.The co-crystallized inhibitors in the chosen ROCK1 and ASK1 protein structures constituting the training set include the following compounds:

ROCK1 inhibitors:
1) VFS (PDB ID: 7JOU) was discovered during a structure-activity relationship optimization of some phenylpyrazole amides to identify potential dual ROCK1 and ROCK2 inhibitors.It exhibited a high potency with IC50 values of 7.9 nM and 2.3 nM on ROCK1 and ROCK2, respectively.VFS binds in a similar pattern at both ROCK1 and ROCK2 which is seen among other ROCK1 and ROCK2 crystal structures.In ROCK1, VFS binds to the hinge region via hydrogen bonding with Tyr155 and Met156.Moreover, the amide carbonyl group interacts with the conserved Lys105 via a hydrogen bond with the terminal phenyl ring directed under the p-loop.
2) 4KH (PDB ID: 4YVC) was discovered during a study aimed to design and establish the structure-activity relationship of pyridine-based ROCK inhibitors.The whole study was based on this compound which was identified through high-throughput screening.It was chosen due to its potential to be developed further producing more potent and selective inhibitors.It exhibited Ki of 1.1 μM more potent than against the off-target protein kinase A (Ki > 5 μM) which is a member of the AGC family as well.4KH pyridine group interacts within the hinge region via a hydrogen bond with Met156 and its amide carbonyl is within a hydrogen bonding distance with the catalytic Lys105.
3) 3J7 (PDB ID: 4W7P) was discovered through structure-based drug design and was tested through pharmacokinetic/pharmacodynamic experiments to identify potential treatment options for pulmonary arterial hypertension through targeting ROCK1.3J7 has an IC50 of 32 nM and binds to ROCK1 hinge region Met156 located via its indazole ring, moreover, it binds to the catalytic Lys105 via a nitrogen atom at its pyridopyrimidine core.It also achieves an interesting hydrogen bond interaction with Asp160.

ASK1 inhibitors:
1) RFG (PDB ID: 6VRE) was discovered through medicinal chemistry studies that aimed to discover potent and CNS-penetrating ASK1 inhibitors.Due to its involvement in apoptosis and inflammation, ASK1 is believed to have a therapeutic value in neurological diseases.RFG has an IC50 of 138 nM and binds to ASK1 hinge region Val757 via an amide carbonyl oxygen and to its catalytic Lys709 by its triazole nitrogen.
2) 8V7 (PDB ID: 5V24) was discovered in a study that aimed to integrate several lead generation and optimization techniques to design novel and potent ASK1 inhibitors.It has an IC50 of 400 nM and binds to Val757 of the hinge region through its amide carbonyl.Through its triazole ring, 8V7 binds to the hinge region Met754 and the catalytic Lys709 via hydrogen bonding.
3) 9E1 (PDB ID: 5VIL) was discovered during efforts to further optimize potent ASK1 inhibitors to increase potency and selectivity and improve physicochemical properties.This inhibitor has a high potency (IC50 = 1.5 nM) and binds to ASK1 hinge region Val757 through its amide carbonyl oxygen.Moreover, it interacts with Gly756 through a hydrogen bond with its sulphonamide.Through its triazole group, it interacts with the catalytic Lys709.4) 8GY (PDB ID: 5UOX) was discovered by utilizing structure-based drug design, deconstruction, and re-optimization of an already known ASK1 inhibitor.When tested, this compound displayed efficient MAP3K pathway inhibition.Furthermore, when investigated in a heart model of cardiac injury, it reduced the size of the infarcted tissue.8GY has a significantly high potency with an pIC50 of 8.3±0.06.Through its amide carbonyl, it interacts with Gly756 and Val757 at the hinge region and by its triazole group, it interacts with hinge region Met754 and catalytic loop Lys709.Finally, by its hydroxyl group, it interacts with Asn808 and Ser821.
S2. Assessment metrics used to assess the performance of the constructed pharmacophore models.

Assessment
Metric Equation Insight The ability of the model to identify truly active molecules as hits.

Specificity (Sp) 𝑆𝑝 = 𝑇𝑁 𝑁 − 𝐴
The ability of the model to identify decoys and discard them.

Accuracy (Acc) 𝑎𝑐𝑐 = 𝑇𝑃 + 𝑇𝑁 𝑁
The accuracy of the pharmacophore in differentiating between actives and inactive compounds.

Discrimination Ratio (DR) 𝐷𝑅 = 𝑆𝑒 𝑆𝑝
Measures the predictive accuracy of the pharmacophore model when screening active and inactive compounds.
F1 score (f1) Measures the overall quality of the model in differentiating between active and inactive compounds.
Generates a high score only if the pharmacophore model can predict the majority of actives and inactive compounds occurrences.
TP is the number of true positive, TN is the number of true negative, FP is the number of false positive, FN is the number of false negative, A is the number of actives, n is the number of selected compounds as hits and N is the total number of compounds in the dataset.

Figure S1 .Figure S2 .
Figure S1.2D ligand interactions diagram of the co-crystalized ligand (a) and 2D diagram (b) and 3D representation (c) of the overlay of the self-docking pose (green) and co-crystalized ligand (red) in the ROCK1 binding site (PDB ID: 4W7P).

Table S2 .
The quantitative and qualitative description of the generated Pharmacophore models' feature ** Radius of excluded volume in Å.