BioWorld International Correspondent

LONDON - Detailed pictures of the structure of an enzyme that is crucial to the survival of the bacterium Mycobacterium tuberculosis inside the host cell may help scientists to design a new generation of drugs to treat tuberculosis.

Work by a team of researchers based in Switzerland has shown that the vital enzyme has a unique "pocket" that allows it to accommodate a highly specific inhibitor.

Jean Pieters, group leader at Biozentrum at the University of Basel, Switzerland, told BioWorld International, "Now that we have this structure, this suggests that we will be able to design compounds that can inhibit this bacterial enzyme even more specifically." He currently is looking for partners to help develop the work further.

Pieters, together with collaborators led by Michel Steinmetz from the Paul Scherrer Institute in Villigen, near Zurich, Switzerland, reported the findings in a paper in the July 3 Proceedings of the National Academy of Sciences titled: "Structural basis for the specific inhibition of protein kinase G, a virulence factor of Mycobacterium tuberculosis".

New drugs to treat tuberculosis are desperately needed. No new drugs have come onto the market for more than 40 years, and resistance to those in use is widespread. Health authorities such as the TB Alliance predict that there will be 1 billion new tuberculosis infections and 36 million deaths within the next 20 years.

M. tuberculosis, the main cause of tuberculosis in humans, can survive in host cells because it prevents the cell from delivering it to the lysosome, where pathogenic bacteria normally are broken down.

In 2004, a team led by Pieters discovered that M. tuberculosis produces an enzyme called protein kinase G (PknG), which allows the bacteria to survive in cells and avoid the lysosome. That work also confirmed the importance of PknG by showing that mycobacteria in which PknG has been inactivated are transferred rapidly to lysosomes and killed.

Pieters and his colleagues then identified a low-molecular-weight inhibitor of PknG, a compound called AX20017. When AX20017 is added to a culture of host cells infected with M. tuberculosis, the cells rapidly transfer the mycobacteria to their lysosomes, and intracellular bacteria are killed, Pieters' group showed. What worried the researchers, however, was the possibility that AX20017 also would inhibit kinases present in human cells. "Critics said, 'If you are going to inhibit PknG, which is homologous to the host kinases, you are going to perturb a lot of host cell activities and you will never make this inhibitor into an antibacterial drug,'" Pieters recalled.

Fortunately, the sequencing of the human genome meant that the DNA code for all 491 human kinases was by then available. Analysis of the "kinome," as that information became known, and comparison with PknG, showed that PknG contained a unique domain. The mycobacterial enzyme had a unique set of amino acid side chains that are not found in any human kinase.

Pieters and his colleagues also carried out experiments that showed that if they mutated any of the amino acids in PknG's unique domain, AX20017 would no longer inhibit the enzyme. "We established that this domain is not found in any of the almost 500 human kinases, nor in any of the other bacterial kinases," Pieters said. "Somehow, it seems that this PknG has evolved to form a specific pocket in its active site, which we just happened to hit by chance using the most rudimentary compounds."

The next step was to crystallize the structure of AX20017 in combination with PknG, in collaboration with Steinmetz's group. That was not an easy task. "This was a nightmare," Pieters said. "We spent several years purifying proteins here in Basel, but upon arrival in the Villigen laboratory, we could only grow miserable crystals."

Eventually, however, all the problems were solved, and the team obtained a high-quality X-ray diffraction pattern that allowed them to produce an atomic model of PknG in the presence of the inhibitor.

Pieters' group is embarking on a project to screen low-molecular-weight compounds for those that will fit the active site of PknG. Pieters also plans to use modeling to determine ways of optimizing the activity of the inhibitors identified in that way.

He said: "It was difficult to optimize the activity of the early inhibitors we tried, at the same time as optimizing their stability. Now that we know the important sites on these compounds, we can rationally design compounds that are highly effective against PknG."