In a completely different approach to drug discovery, a team led by Dr. Samir K. Brahmachari at Delhi’s CSIR-Institute of Genomics and Integrative Biology (CSIR-IGIB) has used a combination of approaches to predict potential drug targets in Mycobacterium tuberculosis, the TB causing bacteria. The novel method not only helps in speeding up drug discovery by finding potential, non-toxic drug targets but will also cost far less by reducing the chances of failure. The results were published in the journal Scientific Reports.
“Conventionally, drug discovery was never looked at from a Systems Biology point of view. The approach we used was rather unconventional. We looked at finding targets first based on evolutionary conservation principle in an organism,” says Dr. Divneet Kaur from IGIB and the first author of the paper. Evolutionary conservation is based on the premise that genes that are very critical for the bacteria do not undergo any mutation.
Based on a previous study that used the Systems Biology Spindle Map (SBSM) approach, the team was able to identify 890 non-toxic genes responsible for metabolism as drug targets. Using computational approaches, the potential drug targets were reduced to 116; these 116 genes are so vital that any inhibition would kill the bacteria.
In order to identify drug targets with the least likelihood of side effects, the 116 essential genes were compared with the human genome and human microbiome at the sequence level to identify genes that did not have any similarity (homology) with human genome sequences.
Of the 116 genes, 104 genes were found to have no similarity with the human genome sequences, meaning any drug developed targeting these 104 genes will only target the TB bacteria and not cause any harm to human cells.
“There is a need for drug discovery to move from Wright brothers’ era of trial and error method. The trial and error approach is slow and too expensive.”The potential drug targets were further shortlisted to 33 genes. The 33 genes play an essential role in bacteria metabolism and have not undergone any mutation in any of the 1,623 TB strains, including the 1,084 multidrug-resistant TB strains, isolated from people with TB. The presence or absence of mutations in any of the 33 genes was evaluated using the Genome-wide Mycobacterium tuberculosis variation (GMTV) database. The genes which are essential for bacteria never undergo any mutations as that would be lethal for their survival.
The crystal structure, which is essential for carrying out drug discovery process, was available for 15 of the 33 targets. “For the 15 genes which have a crystal structure, work has already begun in finding out novel lead compounds for the targets,” says Dr. Mukta Sharma, who is presently involved in the study but is not an author of the paper.
“Once we have the targets and have structures of these targets then can tailor-make molecules to inhibit even MDR-TB and XDR-TB,” says Dr. Kaur. Dr. Debasis Dash from IGIB and one of the authors of the paper helped identify the amino acid changes and mapped them onto the proteins.
The team has already carried out druggability assessment (to find out if the targets have certain properties for a drug to bind to receptors) for all the 33 gene targets. “Most of the 33 genes were found to be highly druggable and none was found to be non-druggable,” says Dr. Sharma. “This validates the approach adopted by this study.”
“There is a need for drug discovery to move from Wright brothers’ era of trial and error method. The trial and error approach is slow and too expensive,” says Dr. Brahmachari. “Through computational approach we first make sure that the target is the Achilles’ heel of M. Tuberculosis and is absent in human cells. Using our new approach, we can work on all organisms.”
According to Dr. Brahmachari, all the genes, targets and even ligands will be in open source so anyone can develop new drug molecules.
In order to understand the drug resistance seen in many TB drugs, the researchers evaluated the gene targets of known drugs — isoniazid, pyrazinamide, ethambutol for any mutations. “The current drug targets show a high degree of mutations,” says the paper. While isoniazid showed relatively lower variation compared with other drugs, the target of bedaquiline (for MDR-TB) drug showed no mutation. According to the paper, this “supports the hypothesis of the importance of completely invariant genes as potential targets for the successful development of novel antibiotics”.
Metformin, a drug used for diabetes care, was found to have minimum number of mutations, so can be repurposed as an adjunct therapy for MDR-TB.