Computational approaches to targeting protein–protein interactions in cancer: a pathway to drug discovery

Abstract

Future Drug DiscoveryAhead of Print CommentaryOpen AccessComputational approaches to targeting protein–protein interactions in cancer: a pathway to drug discoveryMelody Okereke, Kenneth Bitrus David & Oluwakorede Joshua AdedejiMelody Okereke *Author for correspondence: Tel.: +2348039209527; E-mail Address: melokereke30@gmail.comhttps://orcid.org/0000-0003-2533-6785Faculty of Pharmaceutical Sciences, University of Ilorin, Ilorin, Kwara State, NigeriaSearch for more papers by this author, Kenneth Bitrus David https://orcid.org/0000-0002-4688-5591Faculty of Pharmaceutical Sciences, Kaduna State University, Kaduna, NigeriaSearch for more papers by this author & Oluwakorede Joshua Adedeji https://orcid.org/0000-0002-7859-1889Faculty of Pharmaceutical Sciences, University of Ilorin, Ilorin, Kwara State, NigeriaSearch for more papers by this authorPublished Online:5 May 2023https://doi.org/10.4155/fdd-2023-0003AboutSectionsPDF/EPUB ToolsAdd to favoritesDownload CitationsTrack Citations ShareShare onFacebookTwitterLinkedInRedditEmail Keywords: artificial intelligencecancerdrug discoveryinteractionsPPIproteintargetingProtein–protein interactions (PPIs) are physical links that exist between proteins and their partners [1]. They are essential for many cellular processes, including protein folding, gene expression, signal transduction and cell differentiation. PPIs have been demonstrated to be dysregulated in cancer [1]. Therefore, understanding the structure and dynamics of PPIs is crucial for developing new therapeutic strategies for these diseases.PPIs are essential in regulating various signaling pathways, including the Hedgehog pathway, the MAPK pathway, the Notch pathway, the Wnt signaling pathway and the TGF-beta pathway [2]. In the MAPK pathway, PPIs mediate the sequential phosphorylation of several kinases, leading to apoptosis, differentiation and proliferation of cells [2]. In the Wnt pathway, PPIs mediate the binding of Wnt proteins to Frizzled receptors, activating intracellular signaling cascades [2]. Dysregulation of these pathways due to aberrant PPIs has been implicated in cancer and other diseases [1].The large and often flat interfaces involved in interactions make it challenging to target PPIs [2]. However, advances in computational methods and structural biology have enabled the rational design of small-molecule inhibitors that can disrupt PPIs in disease states [3]. These inhibitors can target specific protein–protein interfaces or allosteric sites, disrupting interactions and restoring normal cellular function.Role of PPIs in cancer development & progressionPPIs are involved in the development of cancer in a number of ways, one of which is the abnormal activation of oncogenic signaling pathways [4]. An instance worth highlighting is colorectal cancer. Here, mutations in the APC gene are a common cause of Wnt signaling pathway activation, resulting in the build-up of β-catenin and aberrant activation of downstream target genes [5]. PPIs between β-catenin and its binding partners, such as TCF/LEF transcription factors, are critical for the activation of the pathway [5]. Similarly, in chronic myeloid leukemia, the constitutive activation of the BCR-ABL fusion protein results in the aberrant activation of the Ras/MAPK and PI3K/Akt signaling pathways, which contribute to cell survival and proliferation [6]. PPIs between Bcr-Abl and its downstream effectors, such as Grb2 and Gab2, are critical for the activation of these pathways [6].In addition to activating oncogenic pathways, dysregulated PPIs can also contribute to tumor suppression by disrupting the function of tumor suppressor proteins. In the case of the regulation of the cell cycle and the response to DNA damage, for instance, the tumor suppressor protein p53 plays an essential role [7]. PPIs with their binding partners, such as Mdm2, regulate the stability and activity of p53, which facilitates its degradation and ubiquitination [7]. Dysregulation of these PPIs can result in the loss of p53 function, leading to the accumulation of genomic instability and the development of cancer [8].Targeting PPIs in cancer is challenging due to the large and often flat interfaces involved in these interactions [2]. However, advances in computational methods and structural biology have enabled the rational design of small-molecule inhibitors that can disrupt PPIs in disease states [3]. For example, small-molecule inhibitors that target PPIs between Bcr-Abl and its downstream effectors have been developed and are currently used in the treatment of chronic myeloid leukemia [9]. Similarly, small-molecule inhibitors that target PPIs between Mdm2 and p53 are being developed as potential cancer therapies [7].Traditional approaches to targeting PPIs: traditional small-molecule inhibitorsSmall-molecule inhibitors have been widely used as therapeutics for a variety of diseases, including cancer. These inhibitors typically target enzymes or receptors, and work by binding to specific sites on these proteins and disrupting their activity [10]. The efficacy of traditional small-molecule inhibitors, however, may be constrained by a number of factors, which can also result in the development of drug resistance. Resistance can arise through various mechanisms, including mutations in the target protein that prevent inhibitor binding, upregulation of alternative signaling pathways that bypass the inhibited pathway and the selection of pre-existing drug-resistant cells within the tumor [11].One limitation of traditional small-molecule inhibitors is their lack of specificity. Many inhibitors target proteins that are involved in multiple signaling pathways, and they can therefore have off-target effects that lead to unintended consequences [12]. For example, the widely used tyrosine kinase inhibitor imatinib was initially developed to target the Bcr-Abl fusion protein in chronic myeloid leukemia, but it also inhibits other kinases such as c-Kit and PDGFR, which can lead to side effects such as gastrointestinal toxicity and fluid retention [12]. Another limitation of traditional small-molecule inhibitors is their inability to target PPIs notwithstanding their potential as a viable therapeutic approach [8]. However, PPIs typically involve large and flat interfaces that are difficult to target with small molecules [2].Alternative approaches to PPI targetingWhile traditional small-molecule inhibitors have limitations in targeting PPIs, there are alternative approaches that are being explored to target these interactions. One approach is the use of biologics, such as monoclonal antibodies, that can specifically bind to PPI interfaces and disrupt the interaction [12]. For example, rituximab (a monoclonal antibody) has been successfully used in the treatment of B-cell lymphomas due to its ability to target the CD20 protein on B-cells and induce their death [13].Another approach is the use of stapled peptides, which are synthetic peptides that are stabilized with a covalent bond between two amino acids to maintain a specific conformation for binding to PPI interfaces. Stapled peptides have shown promise in preclinical studies for targeting PPIs involved in cancer, such as the interaction between MDM2 and p53 [13].In addition to biologics and stapled peptides, other approaches to PPI targeting include the use of protein degradation strategies, such as the use of proteolysis-targeting chimeras (PROTACs), which can induce the degradation of specific target proteins by recruiting them to an E3 ubiquitin ligase for ubiquitination and subsequent degradation by the proteasome. PROTACs have shown promise in targeting PPIs involved in cancer, such as the interaction between the estrogen receptor and the coregulator protein SRC-3 [3].Computational approaches to PPI targeting in cancerComputational approaches involve the use of computer simulations, algorithms and data analysis techniques to study biological systems and their interactions [3]. Computational methods have several advantages over traditional experimental approaches for PPI targeting, including their ability to rapidly screen large numbers of compounds, predict binding affinities and selectivity and identify potential off-target effects [3]. Computational methods can also reduce the need for expensive and time-consuming experimental assays. Various computational methods developed for PPI targeting in cancer include molecular docking, molecular dynamics simulations and virtual screening [14]. The following are examples of successful computational approaches that have been used in PPI targeting in cancer:Inhibitors of MDM2-p53 interactionThe growth and survival of cells are regulated by the MDM2-p53 interaction, which is often dysregulated in cancer [14]. Small-molecule inhibitors of this interaction, such as the US FDA-approved drug nutlin-3, have been identified using computational approaches [15]. To determine the binding location and mechanism of action of Nutlin-3, as well as to design inhibitors that are more potent and selective, molecular docking and dynamics simulations, respectively, were used.Inhibitors of Bcl-2 family of proteinsThe proteins of the Bcl-2 family regulate apoptosis and are often dysregulated in cancer [16]. Small-molecule inhibitors of this family, such as the US FDA-approved drugs venetoclax and navitoclax, have been identified using computational approaches [15]. The identification of the binding site and mechanism of action of these inhibitors and the design of more potent and selective inhibitors were carried out using molecular dynamics simulations.Inhibitors of NF-κB signaling pathwayThe NF-κB signaling pathway regulates inflammatory and immunological responses, and its activity is often dysregulated in cancer [17]. Computational methods have been used to identify small-molecule inhibitors of this pathway, including the US FDA-approved drug bortezomib [15]. The identification of the binding site and mechanism of action of bortezomib and the design of more potent and selective inhibitors were carried out using virtual screening and molecular dynamics simulations.With the aforementioned examples, computational approaches have proven to be successful in identifying potential PPI inhibitors in cancer. Through the combination of molecular docking, molecular dynamics simulations and virtual screening, researchers can rapidly identify potential inhibitors, predict their binding affinity and selectivity and optimize their activity, making computational approaches a powerful tool in PPI drug discovery. However, despite significant progress in this field, several challenges and limitations remain.Challenges & limitations of computational approaches to PPI targeting in cancerComputational approaches offer several advantages over traditional approaches for targeting PPIs in cancer. However, accurately predicting the binding affinity between a small molecule or biologic and its target PPI remains a significant challenge [8]. Different scoring functions used by different docking programs can yield varying results, and accurately predicting binding affinity is particularly difficult for more complex PPI interfaces [8]. Furthermore, experimental validation is necessary to confirm the predicted binding and efficacy. The absence of experimental data risks false positives and wasted resources.The accuracy of computational predictions depends heavily on the availability and quality of structural data for the target PPI [3]. If high-quality structural data is not available, computational methods may not accurately predict the binding interface and affinity [3]. Additionally, computational approaches may not account for important factors such as protein flexibility, water-mediated interactions and post-translational modifications that can affect PPI formation and stability [3,8]. These factors are especially important for PPIs that undergo significant conformational changes upon binding or involve partially disordered regions.Furthermore, computational methods may miss potential inhibitors or biologics that are not included in the library. This limitation highlights the need for rational design approaches that leverage computational methods to guide the synthesis of novel compounds with optimal properties for PPI targeting. However, potential challenges to these approaches exist, such as the difficulty of accurately modeling the effects of protein flexibility, the need for large computational resources and the potential for overfitting models to specific datasets. These challenges must be considered and addressed to ensure that computational approaches continue to advance the field of PPI targeting in cancer.To overcome these challenges and limitations of computational approaches for PPI targeting in cancer, several strategies can be employed. It is important to integrate experimental data with computational predictions to confirm the predicted binding and efficacy. This can help reduce the risk of false positives and improve the accuracy of predictions. Additionally, developing more accurate scoring functions and improving the resolution and quality of structural data for target PPIs can improve the accuracy of computational predictions. Furthermore, incorporating protein flexibility, water-mediated interactions and post-translational modifications in computational models can improve their predictive power. Finally, the use of machine learning algorithms can help overcome the limitations of small libraries and improve the identification of potential inhibitors or biologics. Overall, a combination of experimental and computational methods can overcome the challenges and limitations of computational approaches for PPI targeting in cancer.Future directions in PPI targeting & cancer therapyOver the years, PPI targeting has been of major interest to cancer drug researchers leading to various advancements in the field. The future is thus limitless owing to recent advances and the potential of computational drug discovery in understanding and accelerating the development of treatments that influence PPIs. Since the US FDA approved the kinase inhibitor imatinib in 2001 [15], almost 50% of cancer therapeutic discovery efforts are focused on developing new medications that target kinases. While protein kinase inhibition constitutes only one of the mechanisms in PPI targeting, increasing interest has highlighted its benefits in numerous cancers. Since the capacity of many aberrantly produced proteins to interact with a protein-binding partner directly contributes to their ability to promote tumor growth in the malignant state, addressing PPIs important to cancer therapy discovery is essential. Of such importance is the development of novel inhibitors by Cheng et al. where the computer-aided designed Complex 1 showed potential as the first molecule capable of inhibiting the CDK9-cyclin T1 PPI and improving the outcomes commonly associated with triple-negative breast cancer [18]. Compared with conventional chemotherapy involving cytotoxic drugs, protein-targeted anticancer drugs like kinase inhibitors have been shown to cause fewer side effects, thus increasing their potential for more use in the future.Despite the tolerability and effectiveness of PPI-targeted therapy, drug resistance remains an issue that could be further explored in future research [11]. Primary resistance to direct inhibitors may theoretically come from the presence of particular comutations or mutational heterogeneity in a tumor. Knowledge in this area is quite limited and requires more research in terms of large-scale omics analyses to understand, identify and utilize key factors that occur during pretreatment and are responsible for such resistance. Acquired resistance, referring to the de novo development of early resistance on the administration of tyrosine kinase inhibitors, is another issue in PPI therapy requiring further research. As ERK activity is suppressed in RAS-targeted therapy, MYC target genes, such as those encoding RTKs and their ligands, are often derepressed, which leads to the rapid reactivation of the RAS-MAPK pathway to some degree and is referred to as adaptive resistance [19]. In addition, the development of adaptive resistance in anaplastic lymphoma kinase-tyrosine kinase inhibitors is partially understood and requires further study. Given its predominant role in treatment failure, the novel concept of drug-tolerant persister (DTP) cells has garnered many studies and still requires more. In this regard, it is claimed that DTP cells continue to function after being exposed to anticancer drugs, and their DNA repair processes are changed to promote adaptive mutation, which explains the establishment of drug-resistant mutations [20]. All of these resistance mechanisms require further research to enable progress in PPI targeting for cancer therapy.With less than 20% of protein kinases targeted in cancer treatment, this presents an opportunity for more research to explore useful alternative kinases for inhibition in cancer therapy. More research is also anticipated in the production of next-generation kinase inhibitors with enhanced selectivity and CNS penetration [15]. Poor selectivity of PPI therapies is associated with a wide range of side effects that reduce tolerability and the role of CNS penetration is essential in bypassing the blood–brain barrier (BBB) in the management of brain metastases. A number of PPI therapies are quite limited in their ability to cross the BBB, hence the need for more developments in the use of modified formulations and drug administrations to enhance BBB penetration. To find and create small-molecule inhibitors that can target kinases like the TAM kinases, significant research will also be focused on the involvement of the tumor stromal microenvironment and immune biology in the defense of cancer cells [9]. While PPI treatments are currently effective in early-stage cancers, understanding how they can be improved to be beneficial in late-stage treatments as well is crucial.ConclusionPPIs play important roles in cellular signaling, a number of which play significant roles in cancer pathogenesis and progression. While they were formerly regarded as “undruggable targets,” current progress has debunked that and continuous advances have accelerated the development of novel therapies effective in various cancer types. Compared with the traditional drug development process, rational computational drug design, which uses molecular modeling methods like pharmacophore modeling, molecular dynamics, virtual screening and molecular docking to describe the activity of biomolecules and defines molecular determinants for interaction with the drug target, aids in the development of more effective drug candidates with less time and resources. The future of PPI targeting is predicated on the modern use of computational tools and approaches. Artificial intelligence and machine learning can undoubtedly provide immense benefits in the analysis of protein structures, PPIs and targeting in drug design and discovery for cancer therapy.Author contributionsM Okereke conceptualized the idea for this paper. M Okereke, KB David and OJ Adedeji wrote the first drafts of the paper. M Okereke wrote the final draft and reviewed the paper for intellectual content, accuracy and comprehension. All authors agreed and approved the final manuscriptFinancial & competing interests disclosureThe authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.No writing assistance was utilized in the production of this manuscript.Open accessThis work is licensed under the Attribution-NonCommercial-NoDerivatives 4.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/