If you are interested in understanding how they hide, one thing to do is to map their genetic expression—which genes are turned
off and which genes stay active during the latent phase—and try to derive targets. Now, for the first time ever, we are applying
the same tools that we use to find cancer or Alzheimer's or diabetes targets to these bugs.
Are these expression maps used often within the industry?
In the infective area, of course—the genomes of all of these pathogenic organisms are being published constantly. But there
are very few people working in TB—four or five companies.
Now another very interesting point is that most of the original antibiotics were natural compounds because—well, maybe I can
ask you a little test question. Who invented antibiotics?
Was it Alexander Flemming?
Well, a lot of people think it's Flemming, but actually that's wrong. Penicillin was invented by bugs against bugs through
millions of years of evolution. Antibiotics originally were secondary metabolites of bacteria, which they were using to kill
each other. And Flemming learned relatively late that you can find within natural organisms compounds that can kill others.
That was the origin of the human use of antibiotics.
Now we try to synthesize them. Today we have our big libraries with a few million compounds. Whenever we select a new target
out of the genome, we first test our library against it and see if we get hits, compounds that interact with that target.
If we do, we take those as the starting point for a chemical-optimization program to transform the hits into what could be
In the case of both TB and dengue, our compound libraries very often do not contain any drugs interacting with these targets.
It seems that because nobody worked on these bugs for at least two generations, nobody made any compounds against them. And
given that most antibiotics used to be natural compounds, made by bugs for bugs, the synthetic libraries do not contain any
compounds that interact with these bacterial targets.
What this means is now we have to apply the most sophisticated rational-chemistry mechanism, proteomics. We must determine
the three-dimensional structure of those enzymes or signaling molecule pathways that we want to attack receptors, and then
design a new compound or family of compounds that will interact with those. We always have to create new libraries.
This is another way that the absolutely newest form of structural chemistry, rational chemistry, is being rolled out against
neglected diseases, which was never the case before. At the Novartis Institute for Tropical Diseases, with our partners, we
have solved the three-dimensional structure of at least three dengue enzymes, which was essential because that was one of
those areas where we didn't find any lead compounds.
The philosophy behind the NITD is to take sophisticated new drug-discovery mechanisms that are developed for our mainstream
business and systematically apply them to the pathogens of neglected diseases.
Given that there are not a lot of therapies available and the unmet need for treatment is so urgent, are you implementing
any new tactics to speed up the very long discovery and development time?
Well, if I did, I would apply it to the commercial part first, because it has the same problem. Anything to shorten the process
for drug discovery and development and to reduce the attrition rate would have a monumental impact on the whole industry.
Originally, we thought that developing treatments for TB and dengue might be faster, because it's simpler to kill a bug straight
off than to tweak a complex biological mechanism in the human body. But we realized that was a naïve illusion when we looked
at how intimately the pathogen and the host are interconnected.
I've spoken to other researchers about their drug-development innovations, and they've talked about doing Phase I/II or Phase
That's the one thing that Novartis was pioneering—with the proof of concept in man—for all our drugs going into the clinic.
It is in contrast to what we were doing before—Phase I in normal volunteers and looking at the side effect profiles, Phase
II dose-escalation studies, and Phase III for efficacy.