This research project encompasses a number of different approaches to both understand how current anti-tubercular chemotherapy works using the most modern technologies and to use this information to develop new and improved therapies and therapeutic approaches. Individual projects within this framework are;(1) understanding the activity of various drugs in animal models of tuberculosis therapy, (2) correlating responses seen in animal models of disease with the pathology and response to therapy observed in human TB patients, (3) developing structural and functional imaging techniques using CT/PET for use in live, infected animals, and (4) developing techniques for assessing drug penetration and pharmacokinetics in vivo during laparoscopy and bronchoscopy. One important aspect of the project relies on the development of advanced animal models for predicting drug efficacy under conditions that exactly mimic those experienced by TB patients. In partnership with scientists at the University of Pittsburg, section scientists have been exploring the microenvironment of tuberculosis in both rabbits and non-human primates infected with Mtb. Understanding the physical characteristics of the local microenvironment in which Mycobacterium tuberculosis resides is an important goal that may allow targeting of metabolic processes to shorten drug regimens. In 2010, we completed analysis of 18F fluoro-deoxyglucose (FDG) PET/CT images documenting disease development in rabbits and treatment progress with two well known anti-TB drugs INH and RIF. Lesions were observed to be dynamic in both size and metabolic activity even when untreated. As lesions formed the proportion of dense tissue in the lung increased, and once established, lesion density increased until about 10 weeks. Treatment was observed to decrease the volume, density, and metabolic activity of lesions, but each lesion responded independently. Interestingly, changes in PET reported activity occurred must rapidly and could be robustly detected with as little as 1 week of therapy with INH or RIF. Most of our PET-CT studies used 18F fluorodeoxyglucose to image the metabolism of the eukaryotic cells in TB lesions;we are also making attempts to identify the location, abundance and metabolic state of the bacteria in lesions. In an effort to identify molecules that could be used to specifically label MTb in vivo, we capitalized on the unusually broad substrate specificity of the MTb antigen 85 enzymes, which we found could transfer mycolates to a variety of sugars including trehalose modified with bulky substituents. As a result, trehalose labeled with FITC was specifically taken up by MTb in vitro as well as MTb growing in macrophages allowing imaging of live MTb in situ and we could show that the FITC-trehalose was incorporated into mycobacterial cell wall lipids such as trehalose monomycolate and trehalose dimycolate. This offered prospects of using trehalose analogs for labeling and subsequent imaging of MTb in infected animals. We found that this approach worked in vitro and in cultured macrophages, so now we are using these enzymes to incorporate 18F trehalose into bacteria in the lesions of infected rabbits. A series of different positions and methods for attaching the 18F to the sugar are being explored to see which is most efficiently incorporated. In addition, we have begun developing methods for 18F labeling of TB drugs to develop the tools to assess their localization to lesions and the bacteria themselves real-time. In addition, the penetration of TB drugs into TB rabbit lesions has been investigated using a number of different drugs. Using non-compartmental and population pharmacokinetic approaches, we modeled the rate and extent of distribution of isoniazid, rifampicin, pyrazinamide and moxifloxacin in rabbit lung and lesions. Moxifloxacin reproducibly showed favorable partitioning into lung and granulomas, while the exposure of isoniazid, rifampicin and pyrazinamide in lesions was markedly lower than in plasma. Spatial 2-D distribution of rifampicin, isoniazid within lesions is also being studied in collaboration with researchers at the University of Vanderbilt and using matrix-assisted laser desorption/ionization mass spectrometry (MALDI/MS). The 2-D MALDI/MS data for both INH and RIF also indicate a lack of concentration of these drugs into lesion tissue. A similar study is underway in collaboration with Novartis IBR for pyrazinamide and 5 fluoroquinolones. The 2-D MALDI/MS moxifloxacin images confirm the its favorable distribution in to lesion tissue and more specifically into the macrophage-rich region just distal to the necrotic core of the TB lesions. More recently we have begun exploring the lesion penetration of the 7-alkyl substituted PA 824 analogues and tail modifies analogues of PA 824, where the MIC of the compounds is less than 0.3M and MAC less than 3M. The objective of these studies is to compare the lesion penetration of the PA-824 analogues and use lesion penetration as a factor in the selection of a better candidate for future preclinical studies. Our results suggest 3 of standard TB drugs, INH, RIF, PZA do not penetrate into lesion tissue as well as they do into lung, this suggests that these drugs are not reaching as high a concentration in the lesion as might be possible if delivery to the site was increased. We have been performing a series of experiments to determine if treatment with an agent that promotes normalization of blood vessel structure such that hypoxia is decreased and drug penetration is increased could improve drug access to the lesion. In collaboration with Mass General Hospital we have determined that TB lesion-associated vasculature is structurally and functionally abnormal, which might be the cause of poor oxygenation and reduced drug delivery in the rabbit model and human TB disease. Attempts to normalize the vasculature of lesions will continue with and without anti-TB drugs, with results monitored by FDG-PET/CT imaging, lesion histology, drug quantification and bacterial load. Recently, the section began developing a new, non-human primate model for tuberculosis - the common marmoset. The motivation for this was four fold: i) the animal is very small and experiments with new agents can be accomplished with as little as 4 grams of the compound, ii) the animal is used for toxicology studies and so both efficacy and toxicology data could be gained from the same animals, iii) the animals are known to tolerate anesthesia well and were expected to be amenable to prolonged PET-CT studies without morbidity or mortality, and iv) over 80% of live births result in identical twins, allowing us to significantly reduce variability in outcome using genetically identical controls and treated animals. We have shown that the disease displayed by the marmoset is as similar to human disease as that of the larger primate models although no subclinical disease has been documented as is sometimes seen in macaques experimentally infected with a very low dose of Mtb. In addition, we have documented differential virulence with different Mtb strains measured by bacterial load in the lungs and dissemination to other organs, lesion burden, and time to systemic disease. In 2010, we began exploring if the marmoset model accurately reflects the response to treatment by providing standard TB treatment (RIF, INH, PZA, and EMB) to infected symptomatic marmosets. In two separate experiments we have succeeded in documenting response to treatment in both reduction in bacterial burden, reduction in lesion volume by PET/CT, and improved clinical score including weight gain. The analysis of these experiments is ongoing, but the models performance has encouraged us enough to begin planning head to head studies of new potential anti-tubercular agents.
Showing the most recent 10 out of 42 publications
