In silico studiesVirtual screening and hit drug identification

To identify the hits that bind to DBD of FOXM1, we used the pharmacophore features of FDI-6 as a positive control, as depicted in (Fig. 1). The pharmacophore model consists of six distinct pharmacophore features as illustrated in (Fig. 1). Among these features, two hydrogen bond acceptor features are associated with the nitrogen atoms of the thieno[2,3-b]pyridine ring, while another hydrogen bond acceptor feature corresponds to the carbonyl group. Additionally, three aromatic ring features represent the phenyl ring attached to the NH group, the thieno[2,3-b]pyridine ring, and the thiophene ring. Finally, a hydrophobic group is centered around the CF3 group. This comprehensive description of the pharmacophore model provides valuable insights into the structural requirements necessary for ligand binding and interaction within the FOXM1-DBD. We employed a 3D query for virtual screening of the DrugBank (approximately 10,000 compounds) and Selleckchem databases (10,439 compounds). Virtual screening was conducted using Pharmacophore Mapping and Screen Library protocols within the DS software. The screen library protocol was configured to retrieve the top 20 hits, resulting in the identification of 391 compounds (185 from DrugBank and 206 from Selleckchem). Subsequently, 11 compounds were chosen as hits, primarily based on their FDA approval status, along with considerations of fit value, drug’s chemical properties, and structural similarity to FDI-6.

Fig. 1

Pharmacophore features of FDI-6 generated by DS. The hydrogen bond acceptors, aromatic features, and hydrophobic are displayed in mesh spheres of green, orange, and blue, respectively. The arrows indicate the constraint direction (Color figure online)

Based on previously published studies, it is well established that three conserved amino acid residues, including Asn283, Arg286, and His287, play crucial roles in the FOXM1 protein-DNA interaction [28, 31, 36, 44]. His287 establishes direct and indirect hydrogen bonds with the DNA base, mediated by a water molecule [28, 44]. Perturbation of this hydrogen bonding network, possibly due to a His287 mutation, reduces the binding affinity of FOXM1 to DNA. Furthermore, Asn283 exhibits two conformations: vertical, engaging in two hydrogen bonds with DNA, and horizontal, forms minimal DNA interactions but engages in two hydrogen bonds with Arg286. Mutations in Asn283 decrease DNA-binding affinity of FOXM1, while mutations in Arg286 abolish its DNA hydrogen binding, potentially favoring the vertical conformation of Asn283 and enhancing its DNA interactions [28, 31, 36, 44]. Understanding the structural and functional significance of these amino acid residues sheds light on the mechanism of FOXM1-DNA binding and its implications in cellular processes governed by FOXM1.

Molecular docking studies

Based on an assessment of FDA approval status, fit values, drug chemical properties, and structural similarity to FDI-6, we identified eleven potential drugs (hits) among the FDA-approved drugs. Subsequently, molecular docking was performed using AutoDock Vina, followed by molecular docking analysis in the DS, to evaluate the binding affinity of the selected compounds. Each of the eleven potential candidates underwent docking and their respective docking scores and fit values are documented in (Table 1.) These hits exhibited molecular interactions with critical amino acids (His287, Arg286, and Asn283) that play a role in inhibiting FOXM1-DBD. These findings underscore the potential of these selected compounds as promising candidates for further evaluation in the context of binding to and inhibiting FOXM1 function.

Table 1 Docking scores and fit values of the selected 11 drugs targeting the FOXM1-DBD

Based on the molecular docking scores and ligand/drug interaction diagrams (Fig. 2, 3), four potential drugs were identified, including Silibinin, Famotidine, Pantoprazole, and Rabeprazole. Notably, Silibinin and Pantoprazole exhibited higher binding energies of − 6.1 kcal/mol and − 5.9 kcal/mol, respectively, compared to the positive control and reference compound FDI-6 (− 5.6 kcal/mol). On the other hand, Rabeprazole and Famotidine displayed lower binding energies compared to the positive control (− 5.5 kcal/mol, − 4.5 kcal/mol, respectively). The docking results for the selected drugs and standard FDI-6 are summarized in (Table 1).

Fig. 2

The key interactions of the drugs with FOXM1-DBD; the 2D structural representation of FOXM1-DBD residues and 3D representation showing the docked hits interacting with the binding site residues of FOXM1-DBD. a FDI-6, b Silibinin, c Pantoprazole, d Rabeprazole, and e Famotidine

Fig. 3

The comparative docking poses of Pantoprazole (blue), Silibinin (red), Famotidine (yellow), and Rabeprazole (orange) with FDI-6 (green) within the active site of FOXM1-DBD. a The docking poses of all selected drugs alongside FDI-6 in FOXM1-DBD and individual comparisons, depicting the docking poses of Famotidine (b), Silibinin (c), Rabeprazole (d), and Pantoprazole (e), with FDI-6 in the active site of FOXM1-DBD (Color figure online)

Interaction of selected compounds with active site residues on FOXM1

The interactions of selected candidates including Pantoprazole, Rabeprazole, Famotidine, and natural compound Silibinin with the active site residues of FOXM1-DBD were analyzed and compared with the positive control FDI-6. FDI-6 interacted with His287, Asn283, Arg286, and Trp308 through hydrophobic interactions and conventional hydrogen bonds, as shown in (Fig. 2a). Pantoprazole interacted with Trp308, Ser290, Arg286, and His287, while Rabeprazole interacted with Asn288, His287, Asn283, and Arg286. Pantoprazole formed four hydrogen bonds, as shown in (Fig. 2c), and Rabeprazole exhibited various interactions as depicted in (Fig. 2d). Famotidine interacted with Ser284, Giy280, and His287, forming two conventional hydrogen bonds and a pi-sulfur bond, as shown in (Fig. 2e). Silibinin showed the interactions with Arg286, Ser290, Trp308, Leu291, Arg297, and Leu259 forming four conventional hydrogen bonds (Fig. 2b). These findings highlight the specific residues involved in the binding interactions of the inhibitors with FOXM1-DBD providing insights into the molecular interactions and potential mechanisms of inhibition.

Analysis of the interactions between the selected inhibitors and the active site residues of FOXM1-DBD revealed that the standard inhibitor FDI-6 interacts with the critical amino acids His287, Asn283, and Arg286 through two conventional hydrogen bonds. Rabeprazole also displayed the same interaction pattern, forming three conventional hydrogen bonds with these important amino acids. This indicates that Rabeprazole exhibits a stronger interaction with the target compared to the positive control FDI-6. These findings from our in silico studies suggest that Rabeprazole is a more potent inhibitor of FOXM1-DBD than FDI-6.

Pantoprazole exhibited notable interactions with two critical amino acids His287 and Arg286 forming three conventional hydrogen bonds. Notably, it showed superior binding compared to FDI-6 and like rabeprazole, it did not interact with Asn283. This finding highlights the FOXM1-DBD inhibitory potential of Pantoprazole. On the other hand, Silibinin formed two conventional hydrogen bonds both with Arg286, while Famotidine displayed weaker binding by engaging in only one hydrophobic interaction with the amino acid His287. The analysis of the interactions between the selected drugs and FDI-6 showed that Pantoprazole and Rabeprazole interact more strongly with critical amino acids in FOXM1 compared to FDI-6. These findings provide valuable insights into the structural basis of inhibitor–FOXM1–-DBD interactions and suggest Rabeprazole and Pantoprazole as promising candidates for testing in in vitro studies for the inhibition of FOXM1-DBD. Next, we aimed to elucidate the potential binding modes of the selected hits with FOXM1-DBD and to compare their binding modes with that of the standard inhibitor FDI-6 (Fig. 3a).

Silibinin which exhibited interactions with a single critical amino acid for target inhibition demonstrated a binding mode relatively similar to that of FDI-6 (Fig. 3c). Rabeprazole, interacting with all essential amino acids for target inhibition, demonstrated coordination modes comparable to those observed with FDI-6 (Fig. 3d). However, Famotidine displayed a significantly different binding mode compared to FDI-6, suggesting a completely different binding mode (Fig. 3b). Pantoprazole, interacting with two critical amino acids, shared coordination modes like those observed in FDI-6 (Fig. 3e).

The common coordination modes between Rabeprazole and Pantoprazole with regard to FDI-6 suggest their potential as novel FOXM1 inhibitors. Moreover, a relatively close association binding mode exhibited by Silibinin in comparison with FDI-6, underscores its distinct binding characteristics. In contrast, the unique binding model displayed by Famotidine, differing from that of FDI-6, emphasizes its divergent binding properties. These findings elucidate the binding dynamics between the selected drugs and FOXM1, offering valuable insights into the development of novel inhibitors targeting this critical protein.

Based on the studies in the literature, FDI-6 is a potent inhibitor of FOXM1 and acts by disrupting the interaction of FOXM1 with target gene promoters, leading to decreased transcriptional activity [31] By binding to the DNA-binding domain DBD in which it targets critical amino acids such as, His287, Asn283, and Arg286, has been shown be effective in preclinical studies [17, 24, 31]. In vitro preclinical TNBC models, FDI-6 has demonstrated significant antitumor effects, including inhibition of cell proliferation, induction of apoptosis, and suppression of migration and invasion [17, 31]. FDI-6 also sensitizes TNBC cells to chemotherapy, potentially overcoming treatment resistance [17, 31]. However, the safety profile of FDI-6 has not been documented.

In a recent study conducted in 2022 by Zhang et al. [28], the inhibitory potential of XST-20 on the FOXM1-DBD was demonstrated in ovarian cancer cells. Notably, XST-20 demonstrated promising binding interactions with key amino acids, such as His287, Asn283, and Arg286 within the FOXM1-DBD, indicating that this compound also holds significant importance as a FOXM1 inhibitor. However, preclinical in vivo studies demonstrating its in efficacy in tumor models in mice are needed to pave the way for potential clinical applications targeting FOXM1 in various cancers.

MD simulation studies

Throughout the MD simulations of the protein-hit complexes, we conducted a comprehensive analysis of the possible interactions of Rabeprazole and Pantoprazole. The top docking pose and the best conformation interactions with FOXM1 were utilized. The average root-mean-square deviation (RMSD) value was approximately 3.70 Å for the FOXM1-DBD-FDI-6 complex (Fig. S1a) Supplementary Information), suggesting a moderate degree of structural deviation from the initial conformation during the MD simulation. This indicates that while the complex undergoes some structural fluctuations, it maintains a relatively stable overall conformation throughout the simulation. The RMSF plot provides additional insight into the dynamic behavior of individual amino acid residues within the complex, illustrating their varying degrees of mobility over the course of the simulation. Analyzing the RMSF (Fig. S1b) plot enables a detailed examination of regions within the protein that experience significant structural flexibility or rigidity, which can further advance our understanding of the functional dynamics of the FOXM1-DBD-ligand (i.e., FDI-6) complex. In the FOXM1/FDI-6 complex (Fig. 4a), significant hydrogen bond interactions were observed between the ligand and the amino acid residues of Asn283 and His287, as well as Ser290 and Arg286, indicating that these hydrogen bonds contribute to the stability and binding affinity of the complex. Furthermore, hydrophobic interactions involving Trp308 and positively charged atom interactions mediated by the backbone interactions of Arg286. These hydrophobic and electrostatic interactions play a crucial role in the overall structure and function of the complex.

Fig. 4

MD Simulation analysis of protein–ligand (drug) interactions and contact plot with residues on FOXM1 protein for a FDI-6, b Pantoprazole, and c Rabeprazole

In the complex involving Pantoprazole (Fig. 4b), a network of hydrogen bond interactions was identified among the ligand and Asn283, Ser284, Arg286, and His287, which play critical roles in establishing specific contacts and fostering the stability of the complex. Additionally, side chain interactions involving Ser313, Asn315, Asp328, and Leu327 were observed, contributing to the overall binding affinity of the complex. Furthermore, the hydrophobic interactions with Pro279, Leu327, and Tyr317 were noted, thereby augmenting the stability and interactions within the complex. Polar molecular interactions involving His287, Asn283, Ser284, Asn315, and Ser313 were found to bolster the overall structural integrity and specificity of the complex. The average root-mean-square deviation (RMSD) value of approximately 3.30 Å for the FOXM1-DBD-Pantoprazole complex (Fig. S1c) indicates a moderate degree of structural deviation from the initial conformation during the molecular dynamics’ simulation, suggesting a relatively stable overall conformation. The root-mean-square fluctuation (RMSF) (Fig. S1d) plot provided additional insights into the dynamic behavior of individual amino acid residues within the complex, elucidating their varying degrees of mobility throughout the simulation. Analyzing the RMSF plot facilitates a comprehensive examination of regions within the protein experiencing notable structural flexibility or rigidity, thereby enhancing our understanding of the functional dynamics of the FOXM1-DBD-Pantoprazole complex.

In the complex with Rabeprazole and FOXM1 (Fig. 4c), we identified hydrogen bond interactions between Ala277, Phe324, Tyr239, and Asn283. These hydrogen bonds play a significant role in the formation and stabilization of the complex. Moreover, side chain interactions involving Asp321, Gln322, Ile276, Val323, and His275 were observed, contributing to the overall binding affinity and specificity of the complex. The hydrophobic interactions with Ala277, Phe324, Tyr239, Val323, and Ile276 were also detected, further strengthening the stability and interactions within the complex. We also identified negatively charged atom interactions, mediated by the backbone interactions to Asp321 be involved in the complex formation. Polar molecular interactions involving His287, Asn283, Gln322, and His275 were identified, providing additional stability and specificity to the complex. The average root-mean-square deviation (RMSD) value was detected as approximately 2.30 Å for the FOXM1-DBD-Rabeprazole complex (Fig. S1e) indicates a moderate degree of structural deviation from the initial conformation during the MD simulation, suggesting a relatively stable overall conformation. The root-mean-square fluctuation (RMSF) (Fig. S1f) plot provided additional insights into the dynamic behavior of individual amino acid residues within the complex, elucidating their varying degrees of mobility throughout the simulation. Analyzing the RMSF plot facilitates a comprehensive examination of regions within the protein experiencing notable structural flexibility or rigidity, thereby enhancing our understanding of the functional dynamics of the FOXM1-DBD-Rabeprazole complex.

These detailed findings shed light on the molecular interactions within the protein–ligand complexes, uncovering key features that contribute to their potential as inhibitors of FOXM1-DBD. Understanding these interactions is crucial for rational drug design and optimization. The information obtained from the MD simulations can guide future studies aiming to develop novel and potent inhibitors targeting FOXM1-DBD for cancer treatment. Further investigations, including experimental validation and structure–activity relationship studies, are warranted to fully explore the therapeutic potential of these compounds.

In vitro activitySelected drugs reduced cell proliferation and viability of breast cancer cells

To evaluate the effects of selected drugs on breast cancer cell line proliferation and viability, we investigated their impact on (TNBC) BT-20, and estrogen receptor-positive (ER +) MCF-7 cell lines using the MTT cell viability assay. Cells were exposed to varying concentrations of Pantoprazole, Famotidine, Rabeprazole, and Silibinin for 96 h and their viability was evaluated through MTT assays.

Our findings demonstrated that Rabeprazole significantly affected the viability of BT-20 and MCF-7 cells at concentrations ranging between 5 µM and 20 µM (Fig. 5a, b). Pantoprazole exhibited significantly affected cell proliferation and viability at 30 µM and 70 µM in BT-20 and MCF-7 cell lines, respectively (Fig. 5c, d). However, Famotidine and Silibinin did not induce cytotoxic effects within the tested concentration range (5–100 µM). Consequently, Silibinin and Famotidine were excluded from western blot analysis due to their limited effects on cancer cells at high concentrations compared to Rabeprazole and Pantoprazole.

Fig. 5

Rabeprazole and pantoprazole inhibit breast cancer cell proliferation. The effects of Rabeprazole and Pantoprazole were evaluated by MTT cell viability assay after 96 h of treatment of BT-20 and MCF-7 breast cancer cells. Rabeprazole and Pantoprazole reduced the cell viability on BT-20 (a, c) and MCF-7 (b, d) cells. Data were presented as means ± S.D. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001

Rabeprazole and pantoprazole inhibit FOXM1 in breast cancer cells

The inhibitor activity of the two drugs against FOXM1 was determined in BT-20 and MCF-7 cell lines. For this purpose, we evaluated FOXM1 levels after treatments with Rabeprazole and Pantoprazole in the cell lines at various concentrations. Both drugs exhibited strong FOXM1 inhibition at effective concentrations that showed inhibition of cell proliferation. Rabeprazole showed the inhibitor activity at 10 µM in BT-20 and MCF-7 cell lines (Fig. 6a, b). Pantoprazole exhibited FOXM1 inhibition at 30 µM and 70 µM in BT-20 and MCF-7 cells, respectively (Fig. 6a, b). These results indicate that Rabeprazole is the most potent inhibitor against FOXM1 in both cancer cell lines. Previously, we have shown that TNBC cells (MDA-MB-231, BT-20) have higher FOXM1 expression than other breast cancer cell lines, including the MCF-7 cells [43]. Therefore, TNBC cells, BT-20 are expected to be more responsive to Pantoprazole compared to MCF-7 cells response to Pantoprazole. We have previously reported that FOXM1 transcriptionally drives expression of elongation factor 2 kinase (eEF2K), which promotes cell proliferation and growth TNBC tumor xenografts in mice [42]. FOXM1 inhibition leads to downregulation of eEF2K and suppresses cell proliferation and tumor growth in TNBC models [42, 45]. Therefore, we investigated the inhibitor effect of two drugs on eEF2K expression that is the downstream mediator of FOXM1 [43]. The results showed that the treatment of cells with both drugs reduced levels of eEF2K and its direct downstream target p-EF2 (Thr56) in (TNBC) BT-20 and (ER +) MCF-7 cell lines.

Fig. 6

Rabeprazole and pantoprazole inhibit FOXM1 in breast cancer cells. The cells were treated with siRNA FOXM1#1 or FOXM1#2 or control siRNAs (50 nM, 72 h); the cells were treated with Rabeprazole and Pantoprazole or DMSO control vehicle. Rabeprazole and Pantoprazole inhibit FOXM1 expression in TNBC BT-20 (a) and MCF-7 (b) cells. Protein extracts were isolated 72 h after treatments and β-Actin was used as a loading control. RAB rabeprazole, PAN pantoprazole

To confirm whether Rabeprazole and Pantoprazole suppress FOXM1 and its downstream target eEF2K expressions, we also genetically silenced FOXM1 gene using a specific siRNA targeting FOXM1 mRNA in two different breast cancer cells and evaluated expressions of FOXM1 and eEF2K by Western blot analysis. As expected, knockdown of FOXM1 by siRNA decreased the FOXM1 and its downstream targets eEF2K expression as well as phosphorylation of EF2 (p-EF2) at Thr56 in MCF-7 and BT-20 cells compared to control treatments. These studies suggested that genetic and pharmacological inhibition of FOXM1 by siRNA, Rabeprazole, and Pantoprazole, respectively, suppress expression of FOXM1 in both BT-20 and MCF-7 cells (Fig. 6a, b). In this analysis, β-actin was used as a loading control. Overall, these results indicate that the Rabeprazole and Pantoprazole treatments inhibit FOXM1 in breast cancer cells and these drugs may be used as promising FOXM1 anticancer agents for targeting FOXM1 in solid tumors.

Pantoprazole and Rabeprazole are the FDA-approved anti-acid proton pump inhibitors (PPIs) primarily used for the treatment of acid-related gastrointestinal disorders and ulcers to reduce acid secretion from the stomach. While their main indication is not cancer treatment, there is increasing interest in exploring their potential roles in cancer management [46,47,48].

Pantoprazole has been shown to exert antiproliferative effects in gastric cancer, leukemic cancer, colorectal cancer, and pancreatic cancer [46,47,48], inhibit tumor growth, induce apoptosis, and enhance the effectiveness of chemotherapy drugs [46]. Pantoprazole has been shown to inhibit proteasomal degradation [49]. Similarly, Rabeprazole has been shown to demonstrate anticancer properties in gastric cancer, colorectal cancer, and esophageal cancer [50]. Rabeprazole inhibits tumor growth and the expression of cancer-promoting genes and sensitize cancer cells to chemotherapy [50]. However, the mechanism of action of the antiproliferative effects of these drugs was not fully understood.

We have previously demonstrated that FOXM1 is overexpressed in TNBC cells and higher expression of FOXM1 is associated with shorter survival and worse prognosis of patients [6, 8, 42, 43]. Furthermore, we showed for the first time that FOXM1 gene using siRNA suppressed TNBC tumor growth in mice and validated it as a molecular target for TNBC [6, 8, 42, 43]. Our current study provides the first evidence that Pantoprazole and Rabeprazole can bind to FOXM1 and inhibit its activity and downstream signaling including eEF2K and pEF2 in breast cancer cells. As expected, inhibition of FOXM1 with Pantoprazole and Rabeprazole leads to significant reduced of cell proliferation and survival in two different breast cancer cells.

Overall, our studies also shed light on the anticancer effects of two anti-acid drugs Pantoprazole and Rabeprazole. While further research is needed to fully elucidate the potential of Pantoprazole and Rabeprazole in cancer treatment, these findings highlight the potential for repurposing these commonly used PPIs for oncological purposes [46, 49].

Structure–activity relationship (SAR)

According to our combined in silico and in vitro results, we analyzed the chemical structures of reference molecule FDI-6 and approved drugs. A schematic overview of this study is shown in (Fig. 7). Although reference molecule has thieno[2,3-b] pyridine scaffold, the studied drugs contribute to FOXM1 activity with a benzimidazole ring. Furthermore, the sulfinyl group of the approved drugs could take the place of amide substitution in FDI-6. Rabeprazole decreased the FOXM1 activity levels in Western blot analysis due to the pyridine ring bearing long ethylene groups (-CH2-). A difluoromethoxy group at 5-position on benzimidazole scaffold of Pantoprazole may contribute to the activity more than a trifluoromethane substituent at 4-position on FDI-6 reference molecule. In addition, pyridine ring including various substituents can be used in the further design of new analogues of FDI-6 in place of substituted phenyl ring bearing an electron-withdrawing fluorine atom. As a result, we demonstrate that the benzimidazole scaffold and O-substituted pyridine ring can be used in place of the thieno[2,3-b] pyridine scaffold of FDI-6.

Fig. 7

An overview for future perspectives