Abstract
Antibiotic therapy cures infection predominantly by killing the infecting pathogen, but for infections such as tuberculosis (TB), which are accompanied by chronic inflammation, the salutary effects of antibiotic therapy may reflect a combination of pathogen killing and microbiome alteration. This question has not been examined in humans due to the difficulty in dissociating the immunologic effects of antibiotic induced pathogen clearance and microbiome alteration. We analyzed sputum TB bacterial load, microbiome composition, and peripheral blood transcriptomics from a clinical trial (NCT02684240) comparing two antimicrobial therapies for tuberculosis, only one of which was clinically effective. We confirm that standard TB therapy (HRZE) rapidly depletes Clostridia from the intestinal microbiota. The antiparasitic drug nitazoxanide (NTZ), although ineffective in reducing Mycobacterium tuberculosis (Mtb) bacterial load in the sputum, caused profound alterations to host microbiome composition overlapping with alterations generated by HRZE. We then evaluated the effect of these two treatments on the TB driven inflammatory state and found that whereas HRZE normalized proinflammatory TB-associated gene sets, NTZ exacerbated these pathways. Using Random Forest Regression, we identify both pathogen sterilization and microbiome disruption as the top predictors of changes in TB-associated inflammatory transcriptomic markers. We then validate the observed microbiome-peripheral gene expression associations in an independent human cohort of healthy subjects in which the abundance of Clostridia was positively associated with homeostatic, and negatively associated with pro-inflammatory pathways, while the abundance of Bacilli and Proteobacteria species displayed the opposite trend. Our findings indicate that antibiotic-induced reduction in pathogen burden and changes in the microbiome are independently associated with treatment-induced changes of the inflammatory response of active TB, and more broadly indicate that response to antibiotic therapy may be a combined effect of pathogen killing and microbiome driven immunomodulation. Additionally, to our knowledge, this is the first analysis to directly test the hypothesis that the microbiome composition is associated with peripheral gene expression inflammatory profile in humans.
Competing Interest Statement
MFW is currently an employee and shareholder of Regeneron Pharmaceuticals, Inc. MSG reports consulting fees 699 and equity in Vedanta Biosciences, Inc., and consulting fees from Takeda Pharmaceutical Co., Ltd. VB reports consulting fees 699 and equity in Vedanta Biosciences, Inc.
Clinical Trial
NCT02684240
Funding Statement
MFW, CKV, YT, KW, CN, DWF, MSG acknowledge funding from the Tri-I TBRU (grant: U19AI111143). CV acknowledges support from K08AI132739. MFW acknowledges support from the National Center for Advancing Translational Sciences (grant: TL1TR002386-02). Support for NCT02684240 came primarily from the Abby and Howard P. Milstein Program in Chemical Biology and Translational Medicine. VB acknowledges support from the National Science Foundation (grant: 1458347). This work was supported by P30 CA008748.
Author Declarations
All relevant ethical guidelines have been followed; any necessary IRB and/or ethics committee approvals have been obtained and details of the IRB/oversight body are included in the manuscript.
Yes
All necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived.
Yes
I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).
Yes
I have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable.
Yes
Data Availability
Deposition of data. 16S rDNA sequencing data is deposited with the SRA under accession no. PRJNA445968 (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA445968) Code used for 16S analysis is available at https://wipperman.github.io/TBRU/. Code to analyze the data and to reproduce all the figures and results is available on Github at https://wipperman.github.io/TBRU/.