Drug repurposing represents a pivotal strategy for the expeditious administration of efficacious treatments for specific diseases in a more time- and cost-efficient manner [1]. The majority of successful repurposing instances adhere to a conventional paradigm derived from novel drug development. This paradigm is rooted in the concept of “one-drug-one-target-one-disease”, wherein drugs are tailored for individual diseases and their corresponding gene targets. This research examines the potential of biological pathways as a complementary means of enhancing the success of drug repurposing initiatives [2]. The current study is based on data from the DISNET platform [3], which aims to enhance our understanding of diseases and facilitate the repurposing of drugs by integrating biomedical data on a large scale.
A comprehensive analysis was conducted on successful cases of drug repurposing, focusing on instances where biological pathways serve as the foundational framework for success. These cases are defined as DREBIOP (Drug REpurposing based on BIOlogical Pathways). By identifying these pathway-based repurposing cases, we delved into their underlying patterns, analyzing the diverse biological elements and pathways involved. Our analysis put forward a distinct collection of drug repurposing cases, named DREBIOP as previously defined, which are characterized by the dominant role of biological pathways. Furthermore, we employed the Mann-Whitney U Test to compare gene-disease association values and illustrate the reduced significance of drug target genes in these cases. This was done by contrasting the values of gene-disease association scores in the DREBIOP set with overall association values found in DISNET, as well as with the drug target gene-based repurposing cases (which we name as DREGE (Drug REpurposing based on GEnes)). A significant difference was observed between these gene disease association scores of DREBIOP and DREGE cases (p-value < 0.05).
After identifying cases in the DREBIOP set in which the repurposing process was a result of connections through biological pathways, and discovering the most relevant characteristics of these cases, a new drug repurposing approach was developed. This novel approach is based on the direct relationship that exists between different diseases through a biological pathway. Thus, if we have "Disease 1" - Pathway - "Disease 2", drugs used to treat Disease 2 could be considered as candidates to treat Disease 1, inferred by its association to the same pathway.
To perform this method, direct disease-pathway associations were obtained from WikiPathways1. “As of 02/24/2023, a total of 725 different associations were obtained comprised of 325 different pathways identified by WikiPathways identifiers, and 355 different diseases identified by the Concept Unique Identifier from the Unified Medical Language System (from now on, UMLS CUIs).”. These were used to create "Disease 1" - Pathway - "Disease 2" triplets, where disease sharing pathways were related. Through this association, 2,000 different triplets were obtained. Drug candidates for Disease 2 were then determined to be either previously associated with Disease 1 or newly discovered using this method. By associating Disease 1 with all possible candidate treatments for Disease 2, a total of 21,968 records were obtained using the DISNET database to check whether the association was new or had already been previously described (known), 76.7% of the associations between Disease 1 and potential drugs were found to be new. Of the original 2,000 triplets, new cases were obtained for 821 of them. Thus, a very large number of new potential candidates for the treatment of these diseases have been identified through the current analysis.
A review of the scientific literature was then conducted to examine specific examples of these associations to determine if these findings had been previously investigated in clinical trials or if there were studies supporting this relationship. This process paved the way to understand the potential of repurposing each of these drugs for the identified disease. One of the examples studied confirming the potential of this method was the association between Huntington's disease (with UMLS CUI C0020179) and Schizophrenia (with UMLS CUI C0036341) through the phosphodiesterase in neuronal function pathway (with WikiPathways Identifier WP4222). Of the Huntington's disease-related drugs, 7 are potential candidates for the treatment of schizophrenia due to their common pathway. One of these drugs is Tetrabenazine, a vesicular monoamine transporter 2 (VMAT) inhibitor used to treat chorea associated with Huntington's disease [4]. Several studies have investigated the efficacy of this drug in schizophrenia [5], [6]. A recent meta-analysis concluded that tetrabenazine was moderately or highly effective in about three quarters of patients with schizophrenia or psychosis who were treated with it [7].
In conclusion, biological pathways emerge as one potential element to be considered for some specific drug repurposing efforts. Our research reveals discernible patterns that distinguish cases of pathway-based repurposing from the classical paradigm. These findings provide not just insights into the importance of pathway-based repurposing but also provide a new approach to drug repurposing based on biological pathways.
Smart Citations