New path for drug development

The 1960s was a decade that witnessed humans setting foot on the moon and a bloody war that changed the history. In between these events, a drug that would be the mainstay for doctors for decades to fight a smart bug, was born.

Rifampicin was the last novel class of antibiotics against Mycobacterium tuberculosis till the arrival of bedaquiline at the fag end of 2012. Discovered in 1965, Rifampicin was marketed in Italy in 1968 and was approved by the US regulatory body in 1971. Half-a-century later, it still remains one of the best bets in the hands of doctors to treat the dreaded bacteria that kill millions around the world.

Notwithstanding its efficacy, Rifampicin and other medicines used for the first line treatment of tuberculosis (TB) come with a problem. They form a treatment regimen that lasts for six to nine months for standard cases and close to two years for drug resistant ones. The medicine’s effectiveness is compromised because of the long duration of the treatment, which leads to poor adherence and subsequent development of drug resistance.

The limitations are more evident in resource-poor countries like India, which is driving many researchers to look for a short and simple drug regimen that is effective, safe and robust. However, this is easier said than done for various reasons. For one, pharmaceutical companies are not interested because TB is a poor man’s disease.

TB was left to publicly-funded scientists, who simply didn’t have the wherewithal to convert the laboratory leads into a commercial drug. Drug development is an expensive business with a failure-success ratio of nearly 5000:1. That’s why even though the first genome sequences of the TB bacteria came in 1998 there was little follow-up action for almost a decade.


With the help of computers

This was the time when a group of Indian scientists decided to approach the TB drug development differently. “When the Wright brothers made the first airplane, they followed a trial-and-error method. But Boeing-777 is created in a computer. Similarly, we wanted to build an in-silico model for TB drug, moving away from the trial-and-error method that the pharmaceutical industry follows,” said Samir K Brahmachari, a former professor of Indian Institute of Science, Bengaluru.

Though sounds easy, in reality the task was an extremely challenging one. It involved deciphering of close to 50,000 scientific literature and extensive studying of the 1,623 sequenced MTb (Mycobacterium tuberculosis) strains to come out with the genetic and biochemical building blocks needed to recreate the microbe in the computer.

In an unprecedented move, the scientists decided to rope in hundreds of students in an open platform to analyse the data for annotating the genome. By then, Brahmachari became the director general of the Council of Scientific and Industrial Research, which finally rolled out the Open Source Drug Discovery programme in 2008 to make sense of the knowledge generated over the years.

Understanding the complex biological responses or system biology of an organism is important in helping improve and  quicken the process of TB drug discovery. A system level analysis was undertaken with the integration of proteomics and genomics data, incorporating all the experimental evidence from the literature to identify novel, non-toxic drug targets. After almost six years, around 2014, the building blocks were in place. It was time to simulate the bug in the computer.

The team took help from biologist Rohit Vashisht and two researchers from Indian Institute of Technology, Madras and National Institute of Technology, Bhubaneswar for the coding process. After several months, a novel system biology spindle map of the microbe was created depicting the complex biochemical interplay within the microbe triggered by 890 metabolic genes. It also shows role of 961 metabolites in 1,152 reactions inside the bacteria. This map serves as a unique tool in the hands of researchers looking for new drugs against TB.

But out of all these genes, which ones are the best drug targets? The scientists propose that genes showing no variation in the entire 1,623 strains were considered as invariant and could prove to be critical. They found 33 genes (and the proteins they generate) with no variation, which could be the best drug targets. The findings were published in the April issue of Nature Scientific Reports.

The researchers followed on by analysing the crystal structure of 15 of these proteins and discovered that 12 of them are highly drugable. “The discovery of bedaquiline took 20 years as it was done in a trial-and-error method. The computational approach reduces the chances of failure, thereby bringing down the cost, and saves time. It would eliminate those structures that are not permitted and streamline the pipeline,” said Brahmachari.


High degree of mutation

Several leads are already out. The analysis showed that the current set of targets show a high degree of mutation in the bacterial genome. It suggests why doctors found some of the existing medicines like kanamycin and isoniazid becoming increasingly ineffective to tackle TB. The bacteria also has a special set of genes which help it survive when facing an assault. On the other hand, medicines like lansoprazole (used for heart burn) and metformin (used for type-II diabetes) could be an effective anti-TB medicine, because they target those genes that didn’t mutate at all. Similarly, the reason behind bedaquiline’s efficacy is that it targets one of those highly conserved regions in the genome.

Buoyed by the scientific results, the Indian Council of Medical Research (ICMR) plans to undertake a clinical trial to see the efficacy of Metfomin as an anti-TB drug. “While it is a good and logical approach, it needs to be checked if those genetic mutations are in the Indian strain as well. The idea they developed, needs to be taken forward,” commented Soumya Swaminathan, the director-general of ICMR, who is not associated with the study.

Brahmachari, with his students and associates including Divneet Kaur, Rintu Kutum and Debasis Dash, now plans to test the efficacy of most of the 4,000 plus medicines approved by the US Food and Drug Administration for other indications, as an anti-TB drug. “We already found four medicines (used for other indications) to inhibit the bacteria,” Brahmachari said.