Rifabutin central nervous system concentrations in a rabbit model of tuberculous meningitis.

Tuberculous meningitis (TBM) has a high mortality, possibly due to suboptimal therapy. Drug exposure data of antituberculosis agents in the central nervous system (CNS) are required to develop more effective regimens. Rifabutin is a rifamycin equivalently potent to rifampin in human pulmonary tuberculosis. Here, we show that human-equivalent doses of rifabutin achieved potentially therapeutic exposure in relevant CNS tissues in a rabbit model of TBM, supporting further evaluation in clinical trials.

A preclinical model of TB meningitis to determine drug penetration and activity at the sites of disease.

Tuberculosis meningitis (TBM) is essentially treated with the first-line regimen used against pulmonary tuberculosis, with a prolonged continuation phase. However, clinical outcomes are poor in comparison, for reasons that are only partially understood, highlighting the need for improved preclinical tools to measure drug distribution and activity at the site of disease. A predictive animal model of TBM would also be of great value to prioritize promising drug regimens to be tested in clinical trials, given the healthy state of the development pipeline for the first time in decades. Here, we report the optimization of a rabbit model of TBM disease induced via inoculation of Mycobacterium tuberculosis into the cisterna magna, recapitulating features typical of clinical TBM: neurological deterioration within months post-infection, acid-fast bacilli in necrotic lesions in the brain and spinal cord, and elevated lactate levels in cerebrospinal fluid (CSF). None of the infected rabbits recovered or controlled the disease. We used young adult rabbits, the size of which allows for spatial drug quantitation in critical compartments of the central nervous system that cannot be collected in clinical studies. To illustrate the translational value of the model, we report the penetration of linezolid from plasma into the CSF, meninges, anatomically distinct brain areas, cervical spine, and lumbar spine. Across animals, we measured the bacterial burden concomitant with neurological deterioration, offering a useful readout for drug efficacy studies. The model thus forms the basis for building a preclinical platform to identify improved regimens and inform clinical trial design.

Spatial quantitation of antibiotics in bone tissue compartments by laser-capture microdissection coupled with UHPLC-tandem mass spectrometry.

Bones are the site of multiple diseases requiring chemotherapy, including cancer, arthritis, osteoporosis and infections. Yet limited methodologies are available to investigate the spatial distribution and quantitation of small molecule drugs in bone compartments, due to the difficulty of sectioning undecalcified bones and the interference of decalcification methods with spatially resolved drug quantitation. To measure drug concentrations in distinct anatomical bone regions, we have developed a workflow that enables spatial quantitation of thin undecalcified bone sections by laser-capture microdissection coupled to HPLC/tandem mass spectrometry, and spatial mapping on adjacent sections by mass spectrometry imaging. The adhesive film and staining methods were optimized to facilitate histology staining on the same sections used for mass spectrometry image acquisition, revealing drug accumulation in the underlying bone tissue architecture, for the first time. Absolute spatial concentrations of rifampicin, bedaquiline, doxycycline, vancomycin and several of their active metabolites are shown for both small rodent bones and larger rabbit bones that more closely resemble human bone density. Overlaid MALDI mass spectrometry images of drugs and histology staining enabled the generation of semi-quantitative data from regions of interest within anatomical bone compartments. These data correlated with absolute drug concentrations determined by HPLC-MS/MS in laser-capture microdissection samples. Collectively, these techniques enable semi- and fully quantitative drug distribution investigations within bone tissue compartments for the first time. Our workflow can be translated to image and quantify not only drugs but also biomarkers of disease to investigate drug penetration as well as mechanisms underlying bone disorders.

Wild mouse gut microbiota limits initial tuberculosis infection in BALB/c mice.

Mouse models are critical tools in tuberculosis (TB) research. Recent studies have demonstrated that the wild mouse gut microbiota promotes host fitness and improves disease resistance. Here we examine whether the wild mouse gut microbiota alters the immunopathology of TB in BALB/c mice. Conventional BALB/c mice (LabC) and mice born to germ-free BALB/c mothers reconstituted with the wild mouse gut microbiota (WildR) were used in our studies. WildR mice controlled initial TB infection better than LabC mice. The microbial gut communities of LabC mice and WildR mice had similar richness but significantly different composition prior to infection. TB reduced the gut community richness in both cohorts while differences in community composition remained indicating a general TB-induced dysbiosis. The wild mouse gut microbiota did not alter the typical lung histopathology of TB in the BALB/c model that includes unstructured immune cell infiltrates with infected foamy macrophages invading alveolar spaces. Animals of both cohorts mounted robust T cell responses in lungs and spleen with lower absolute counts of CD4 and CD8 T cells in lungs of WildR mice during acute infection, corresponding with observed differences in pathogen load. In summary, LabC mice and WildR mice showed largely overlapping TB immunopathology and pathogen kinetics, with WildR mice controlling early acute infection better than LabC mice.