We developed PS4DR, an automated workflow that enables the integration of multimodal datasets together with pathway information from different canonical pathway databases to predict drug repositioning candidates in different diseases (Figure 1). We showcase PS4DR using real-world gene expression signatures (i.e., Open Targets (Koscielny et al., 2016) and LINCS) and GWAS data (i.e., GWASdb (Li et al., 2011), GWAS Catalog (Welter et al., 2013), GRASP (Leslie et al., 2014), and PheWAS (Denny et al., 2013)). First, the workflow filters disease and drug transcriptomics (i.e., gene expression signatures) with the help of GWAS data. The next step involves calculating pathway signatures for diseases and drugs via pathway enrichment analysis with the filtered dataset. Finally, PS4DR performs an anti-correlation analysis by calculating correlation scores between the pathway signatures of drugs and diseases to prioritize drugs for each disease. Below, we show the utility of the workflow with three applications on how this tool can serve to i) identify drug repositioning candidates, ii) prioritize drug combinations, and iii) propose drugs that simulate disease dysregulations.
Figure 1: An overview of the PS4DR workflow. The workflow requires three different datasets as inputs, (i) disease perturbed gene expression signatures, (ii) genome-wide association study (GWAS) data, and (iii) drug perturbed gene expression signatures. The first and optional part of the workflow involves different filtering steps based on gene set intersection operations that enable the identification of genes in the gene expression signatures that have also been identified in a GWAS of the studied disease. To retain the maximum flexibility in the workflow, users can decide which of the filtering steps they wish to apply, if any. The next step uses the transcriptomics datasets, filtered or not, to conduct pathway enrichment analysis and evaluate the direction of perturbation for each affected pathway in a particular disease context. While the dotted lines in the figure represent all possible combinations of the filtering steps that can be applied and lead to the pathway enrichment step, solid lines show the option we chose to demonstrate the workflow. Finally, the last step uses the correlation of the pathway scores calculated by the previous step to prioritize drugs that are predicted to invert the pathway signatures observed in a given disease context.
1.1. Identifying Drug Repositioning Candidates
As a first application, we explored the list of 26 diseases for which our workflow predicted drug repositioning candidates. While our workflow predicted plenty of drug candidates, we considered two criteria to prioritize predicted drugs. First, we prioritized all drugs in each disease based on their negative correlation scores. However, a drug could have a negative correlation score by only reverting a minority of the pathways dysregulated in the disease. Therefore, we also consider the relative number of the dysregulated pathways reverted by a drug for the prioritization process. While this prioritization approach facilitated narrowing down the candidate lists, we are aware that each of the drugs exhibiting negative correlation scores might have the potential to revert the disease condition even if they alter very few dysregulated pathways.
The distribution and Q-Q plots for the majority of the diseases that output drug predictions demonstrate that the correlation scores follow a normal distribution (Additional file: Figure S1 and Figure S2). Hence, we applied an arbitrary threshold to the correlation score to prioritize the proposed candidate drugs in each disease. We would like to point out that we used the same threshold for all diseases since we are exploring multiple indications; however, this threshold could be selected individually for each disease based on their underlying correlation score distributions. The applied threshold discarded drugs with a correlation score greater than -0.4 or drugs which did not cover more than 50% of the affected pathways in the disease. This filtering step, intended to reduce the number of hits and facilitate the manual investigation of the results, returned a list of predicted drug candidates for 19 diseases (Additional file 1: Table S1). We further investigated the proposed drugs for five conditions to see whether PS4DR was able to identify FDA-approved drugs for their known indications and predict new indications for existing drugs in the prioritized list.
First, we focused on the predicted drug list for melanoma. We searched DrugBank (Wishart et al., 2017) and scientific literature to collect evidence for the proposed drugs and summarized our findings in Table 1. Seven of nine predicted drugs are either already being used as cancer drugs or currently being studied in different clinical trials. This motivates further investigation of these drugs as repositioning candidates for the treatment of melanoma.
The topmost drug in our predicted shortlist, Crizotinib, a non-small cell lung cancer (NSCLC) drug, has been reported for its positive effect on melanoma by two studies (Surriga et al., 2013; Wiesner et al., 2014). While Surriga et al. (2013) suggested that Crizotinib could be used in adjuvant therapy for uveal melanoma due to its c-Met activity inhibition, recent research reported strong kinase fusion association with different melanoma subtypes (Turner et al., 2017) and encouraged the testing of kinase fusion inhibitor Crizotinib for melanoma treatment (Wiesner et al., 2014). The third drug, Sepantronium, a selective small-molecule survivin suppressant, was reported to reduce the accumulation of survivin in G2/M mitotic arrest and induce apoptosis in human malignant melanoma cells in combination therapy with docetaxel (Yamanaka et al., 2011; Lewis et al., 2011). The following drug in Table 1, Bortezomib, is an approved drug for multiple myeloma that was suggested as a treatment for melanoma in combination therapy with temozolomide due to its ability to induce apoptosis and autophagic formation in human melanoma tumors (Amiri et al., 2004; Selimovic et al., 2013). Another FDA approved drug Olaparib (for breast and pancreatic carcinoma), was also found to be effective against melanoma by inhibiting repair of single-strand DNA breaks in different combination therapies (Czyż et al., 2016; McNeil et al., 2013).
The last two approved drugs in the list (i.e., Tivozanib for renal cell carcinoma and Belinostat for peripheral T-cell lymphoma) have been positively associated with a better response in melanoma (Gimsing et al., 2009; Friedman et al., 2015). Moreover, another mTOR inhibitor drug, Vistusertib (AZD-2014), currently in phase II clinical trial for meningioma, was reported to have a positive impact by mTORC1/2 inhibition of the resistance to MAPK pathway inhibitors in melanomas with high oxidative phosphorylation (Gopal et al., 2014; Schmid et al., 2017). Interestingly, we also have two drugs, Olmesartan, for hypertension, and Fluspirilene, for schizophrenia, from very different therapeutic areas in our shortlist. While no reports of their potential role in melanoma treatment have been found yet, numerous studies have suggested their applicability in different cancer treatments (Masamune et al., 2013; Abd-Alhaseeb et al., 2014; Shi et al., 2015; Patil et al., 2015).
Table 1: Drug repositioning candidates for melanoma. Drugs showing a negative correlation score less than or equal to -0.40 and affecting more than 50% of the dysregulated pathways in melanoma. The last column outlines the current uses of the given drug in other conditions according to DrugBank and scientific literature.
We have found three drugs in breast carcinoma (Additional file 1: Table S1). The first drug, AT-7519, a selective inhibitor of specific Cyclin-Dependent Kinases (CDKs), is under investigation for the treatment of leukemia, lymphoma, myelodysplastic syndrome, and solid tumors (Wishart et al., 2017). This is in concordance with the study by Yu et al. (2006) describing how a subgroup of breast cancer patients benefited from the treatment of CDK4 kinase inhibitors. The next drug, Omacetaxine Mepesuccinate, used for chronic myeloid leukemia, is in a clinical trial (NCT01844869) for treating advanced solid tumors (i.e., breast, lung, colorectal and melanoma). Finally, Rigosertib has shown potent antitumor activity in various preclinical models such as breast cancer and pancreatic cancer xenografts and is currently under clinical trial (Nuthalapati et al., 2012).
Similarly, we found that six out of eight drugs proposed for pancreatic carcinoma are either already being used in different cancers or have been suggested in the literature, as we discuss below (Additional file 1: Table S1). The first drug, Fenofibrate, an antilipemic agent, was reported to inhibit pancreatic cancer cell proliferation via activation of p53 mediated by upregulation of MEG3 (Hu et al., 2016). The next drug, Menadione, was found to induce reactive oxygen species to promote apoptosis via redox cycling in pancreatic cells (Criddle et al., 2006; Osada et al., 2008). Fluoxetine, originally an antidepressant agent, was also reported to work as a chemosensitizer and acts with other cancer drugs to overcome multidrug resistance in cancer cells (Zhou et al., 2012). An investigational cancer drug, Tosedostat, was found to be well-tolerated and clinically active against pancreatic ductal adenocarcinoma patients in phase I/II clinical trial (Wang-Gillam et al., 2017; NCT02352831). Another drug, AZD-6482, a selective PI3Kβ inhibitor, could be useful in pancreatic cancer treatment because of its apoptotic effect in cancer cell lines (Xu et al., 2019). Praziquantel was reported to inhibit cancer cell growth when used synergistically with paclitaxel via downregulating the expression of X-linked inhibitor of apoptosis protein (XIAP) (Wu et al., 2012).
While our workflow showed very promising results in cancer, we wanted to explore the results in complex disorders with no available treatments, such as Alzheimer’s disease (AD) and multiple sclerosis (MS). In the case of AD, the workflow provided fourteen shortlisted candidates (Table 2). The top drug on the list is Sirolimus (rapamycin), an immunosuppressant, already proposed for the treatment of AD by different studies (Spilman et al., 2010; Bové et al., 2011; Cai and Yan 2013). It has been suggested that the therapeutic effect of this drug is due to the reduction of amyloid-beta levels caused by its inhibition of the mTOR signaling pathway (Spilman et al., 2010). Another compound, Pimozide, an antipsychotic agent, was recently suggested as a potential AD therapeutic which was reported to reduce toxic forms of tau protein by enhanced autophagy activity via AMPK-ULK1 axis stimulation (Kim et al., 2017). Interestingly, we have two cancer drugs, Pevonedistat and Nilotinib, which could have potentially positive effects on AD treatment (Andérica-Romero et al., 2016; Scudder and Patrick 2015; Lonskaya et al., 2014; NCT02947893). Pevonedistat, a neddylation inhibitor, could prevent neuronal damage and ameliorates cognitive deficits by preventing NRF2 protein degradation via inhibiting neddylation (Andérica-Romero et al., 2016; Scudder and Patrick 2015). Nilotinib, a tyrosine kinase inhibitor, has also been found to be very promising to delay the progression of AD by enhanced amyloid-beta clearance (Lonskaya et al., 2014; NCT02947893).
Animal studies have demonstrated that the blockade of muscarinic receptors results in increased levels of acetylcholine and improve cognition (Clader and Wang, 2005). Therefore, another proposed drug, Terfenadine which is a muscarinic receptor antagonist and has not yet been linked to AD, could be a potential repositioning candidate. Similarly, several 5-HT6R antagonists have advanced to different phases of clinical trials (Benhamú et al., 2014; NCT02258152; NCT02580305) as treatments for AD. The results also suggest another drug in the list, Ritanserin, that has not been directly indicated for AD. The high score proposed by our workflow to this serotonin receptor antagonist may be explained by its regulation of the neuronal cholinergic and glutamatergic pathways, both dysregulated in AD. Furthermore, there is increasing evidence showing that neuroinflammation significantly contributes to AD pathogenesis (Lee et al., 2010; Rubio-Perez and Morillas-Ruiz, 2012). Hence, it is not surprising to find two anti-inflammatory agents in our list (i.e., Betamethasone and Halcinonide) that could be worth investigating as potential repositioning drugs. Finally, Doxylamine Succinate, a neurotransmitter agent and histamine antagonist, is also a promising candidate since the beneficial effects of histamine antagonists in AD have been reported in multiple studies (Nuutinen and Panula, 2010; Passani and Blandina, 2011; Vohora and Bhowmik, 2012).
Finally, we investigated the top ranked drugs proposed by PS4DR for multiple sclerosis (MS). Ranked at the top of the list, PS4DR successfully recovered methylprednisolone, a corticosteroid with anti-inflammatory action prescribed to treat acute exacerbations in patients with MS (Sloka and Stefanelli, 2005) (Additional File 1: Table S1).
Table 2: Drug repositioning candidates for Alzheimer’s disease (AD). Drugs showing a negative correlation score less than or equal to -0.40 and affecting more than 50% of the dysregulated pathways in AD.
1.2. Prioritizing Drug Combinations
Although we have illustrated that our workflow is able to identify candidate compounds for drug repositioning, combining multiple drugs can provide more benefits since the number of affected pathways can be increased by taking advantage of their synergistic effects. Therefore, we applied our workflow to all drug pair combinations in all diseases in order to identify therapies that could have a greater effect than single-drug treatments. For this application, we exclusively considered combinations of two drugs for two reasons: i) application of multiple drugs is usually counterproductive since it increases the number of side effects and ii) calculation time increases exponentially with an increasing number of drugs.
We investigated the predictions of our workflow in breast cancer to verify if we have more drugs with a good negative correlation score and affected pathways (%). While we had three drugs from our single-drug prediction approach, we were able to retrieve 489 drug pairs from the drug combination approach with the same thresholds. To facilitate manual investigation, we increased our threshold of correlation score to less than or equal to -0.50 and affected pathways greater than or equal to 80% and were still able to retrieve 34 drug pairs (Additional file 1: Table S2). Here, all 19 new drugs in these 34 pairs are partnered with one of the top two drugs, AT-7519 or Omacetaxine Mepesuccinate, from the single-drug approach. 14 of the new drugs have partnered with both AT-7519 or Omacetaxine Mepesuccinate. While we have found literature evidence for the beneficial role of seven of these new drugs in the treatment of breast cancer, another six drugs are reported to have positive effects in other solid tumor based cancer treatment as described below. The third drug from the single-drug approach, Rigosertib, which was reported to have antitumor activity in breast cancer cell lines (Nuthalapati et al., 2012), has partnered with both AT-7519 or Omacetaxine Mepesuccinate. BGJ-398, a fibroblast growth factor receptor inhibitor in the list, significantly prevented the outgrowth of tumor organoids in metastatic breast cancers (Wendt et al., 2014). An approved cancer drug, Erlotinib Hydrochloride, epidermal growth factor receptor inhibitor, has shown a very positive response rate when treated combinedly with Capecitabine and Docetaxel in advanced breast cancer patients (Twelves et al., 2008). Another drug Selumetinib, a tyrosine kinase inhibitor, is currently being tested in several clinical trials (i.e., NCT03162627; NCT03742102; NCT02503358) for different cancer types, including breast cancer. TAK-715 is a p38 MAP kinase inhibitor in the list that cross-reacts with casein kinase ɛ (CKIɛ). Since CKIɛ mutations have been linked with the proliferation of different breast cancer cell lines, this drug could be explored to repurpose it for breast cancer treatment (Verkaar et al., 2011). Another investigated drug, Tivantinib, has also shown positive effect on breast cancer model by reducing the metastasis via c-MET inhibition (Previdi et al., 2012). Megestrol Acetate, a progesterone receptor agonist, is under various clinical trials either alone or in combination with other cancer drugs for breast cancer treatment (i.e., NCT03306472 and NCT03024580).
AZD-1775, a drug that inhibits the G2–M cell-cycle checkpoint gatekeeper WEE1 kinase, has been used in multiple trials studying the treatment of lymphoma, ovarian cancer, and adult glioblastoma (Matheson et al., 2016; Wishart et al., 2017). Another drug, Axitinib, a selective vascular endothelial growth factor receptor (VEGFR) inhibitor, is under investigation in different clinical trials for various cancer types (i.e., NCT02129647; NCT03494816; NCT03472560). Moreover, four other drugs i,e., BMS-777607, PF-04217903, R-406, and Isotretinoin are reported to have positive effects in different solid tumor cancer types in different studies (Wishart et al., 2017; Zou et al., 2012; Ghotra et al., 2015; Hong et al., 1990).
1.3. Proposing Drugs that Simulate Disease Pathway Signatures
While we have initially focused on the drugs with the most negative correlation scores, we also anticipated a potential utility for drugs showing positive correlations. Well-characterized drugs with high positive correlation scores can provide information about how pathways or targets could be implicated in the molecular basis of the disease. Hence, as an extended application, the workflow may be used additionally as a prioritization tool to identify drugs that could be potentially employed to generate in-vitro or in-vivo models. By investigating the correlation scores (Figure 2), researchers can readily identify drugs that could be used for this purpose. Our workflow predicted induction of disease pathway signatures for Pevonedistat in diabetes mellitus, Alvocidib in Crohn's disease, and Entinostat and panobinostat in systemic lupus erythematosus (SLE) through very high positive correlation scores in addition to their broad coverage of affecting disease pathways. We see the need for further investigations of all the drugs with both high positive correlation scores and a high percentage of affected pathways for their use in potential disease model development.
Figure 2: Combined scatter plots of the drug’s correlation scores against affected pathways (%) in each disease. The relative number of target pathways affected by the drug in the disease context is plotted along the x-axis and correlation scores on the y-axis. Drugs in the top-right corner of the plot might be interesting for developing in vitro disease models since this group of drugs shows positive correlation scores, covering a broad range of the affected pathways. The circles represent drugs and the color coding indicates their respective disease indication, as shown at the bottom.