Computational and Empirical Studies Predict Mycobacterium tuberculosis-Specific T Cells as a Biomarker for Infection Outcome

Distinguishing multiple roles of T cell and macrophage involvement in determining lymph node fates during Mycobacterium tuberculosis infection

Tuberculosis (TB) is a disease of major public health concern with an estimated one-fourth of the world currently infected with M. tuberculosis (Mtb) bacilli. Mtb infection occurs after inhalation of Mtb, following which, highly structured immune structures called granulomas form within lungs to immunologically restrain and physically constrain spread of infection. Most lung granulomas are successful at controlling or even eliminating their bacterial loads, but others fail to control infection and promote disease. Granulomas also form within lung-draining lymph nodes (LNs), variably affecting immune function. Both lung and LN granulomas vary widely in ability to control infection, even within a single host, with outcomes ranging from bacterial clearance to uncontrolled bacterial growth. While lung granulomas are well-studied, data on LN granulomas are scarce; it is unknown what mechanisms drive LN Mtb infection progression and variability in severity. Recent data suggest that LN granulomas are niches for bacterial replication and can reduce control over lung infection. To identify mechanisms driving LN Mtb infection, we developed a multi-scale compartmental model that includes multiple lung-draining LNs, blood. We calibrated to data from a nonhuman primate TB model (one of the only models that parallels human TB infection). Our model predicts temporal trajectories for LN macrophage, T-cell, and Mtb populations during simulated Mtb infection. We also predict a clinically measurable infection feature from PET/CT imaging, FDG avidity. Using uncertainty and sensitivity analysis methods, we identify key mechanisms driving LN granuloma fate, T-cell efflux rates from LNs, and a role for LNs in pulmonary infection control.

A brief overview of mathematical modeling of the within-host dynamics of Mycobacterium tuberculosis

Tuberculosis (TB), a disease caused by bacteria Mycobacterium tuberculosis (Mtb), remains one of the major infectious diseases of humans with 10 million TB cases and 1.5 million deaths due to TB worldwide yearly. Upon exposure of a new host to Mtb, bacteria typically infect one local site in the lung, but over time, Mtb disseminates in the lung and in some cases to extrapulmonary sites. The contribution of various host components such as immune cells to Mtb dynamics in the lung, its dissemination in the lung and outside of the lung, remains incompletely understood. Here we overview different types of mathematical models used to gain insights in within-host dynamics of Mtb; these include models based on ordinary or partial differential equations (ODEs and PDEs), stochastic simulation models based on ODEs, agent-based models (ABMs), and hybrid models (ODE-based models linked to ABMs). We illustrate results from several of such models and identify areas for future resesarch.

A framework for multi-scale intervention modeling: virtual cohorts, virtual clinical trials, and model-to-model comparisons

Computational models of disease progression have been constructed for a myriad of pathologies. Typically, the conceptual implementation for pathology-related in-silico intervention studies has been ad-hoc and similar in design to experimental studies. We introduce a multi-scale interventional design (MID) framework toward two key goals: tracking of disease dynamics from within-body to patient to population scale; and tracking impact(s) of interventions across these same spatial scales. Our MID framework prioritizes investigation of impact on individual patients within virtual pre-clinical trials, instead of replicating the design of experimental studies. We apply a MID framework to develop, organize, and analyze a cohort of virtual patients for the study of tuberculosis (TB) as an example disease. For this study, we use HostSim: our next-generation whole patient-scale computational model of individuals infected with Mycobacterium tuberculosis. HostSim captures infection within lungs by tracking multiple granulomas, together with dynamics occurring with blood and lymph node compartments, the compartments involved during pulmonary TB. We extend HostSim to include a simple drug intervention as an example of our approach and use our MID framework to quantify the impact of treatment at cellular and tissue (granuloma), patient (lungs, lymph nodes and blood), and population scales. Sensitivity analyses allow us to determine which features of virtual patients are the strongest predictors of intervention efficacy across scales. These insights allow us to identify patient-heterogeneous mechanisms that drive outcomes across scales.

A virtual host model of Mycobacterium tuberculosis infection identifies early immune events as predictive of infection outcomes

Tuberculosis (TB), caused by infection with Mycobacterium tuberculosis (Mtb), is one of the world's deadliest infectious diseases and remains a significant global health burden. TB disease and pathology can present clinically across a spectrum of outcomes, ranging from total sterilization of infection to active disease. Much remains unknown about the biology that drives an individual towards various clinical outcomes as it is challenging to experimentally address specific mechanisms driving clinical outcomes. Furthermore, it is unknown whether numbers of immune cells in the blood accurately reflect ongoing events during infection within human lungs. Herein, we utilize a systems biology approach by developing a whole-host model of the immune response to Mtb across multiple physiologic and time scales. This model, called HostSim, tracks events at the cellular, granuloma, organ, and host scale and represents the first whole-host, multi-scale model of the immune response following Mtb infection. We show that this model can capture various aspects of human and non-human primate TB disease and predict that biomarkers in the blood may only faithfully represent events in the lung at early time points after infection. We posit that HostSim, as a first step toward personalized digital twins in TB research, offers a powerful computational tool that can be used in concert with experimental approaches to understand and predict events about various aspects of TB disease and therapeutics.

Spatial Organization and Recruitment of Non-Specific T Cells May Limit T Cell-Macrophage Interactions Within Mycobacterium tuberculosis Granulomas

Tuberculosis (TB) is a worldwide health problem; successful interventions such as vaccines and treatment require a 2better understanding of the immune response to infection with Mycobacterium tuberculosis (Mtb). In many infectious diseases, pathogen-specific T cells that are recruited to infection sites are highly responsive and clear infection. Yet in the case of infection with Mtb, most individuals are unable to clear infection leading to either an asymptomatically controlled latent infection (the majority) or active disease (roughly 5%-10% of infections). The hallmark of Mtb infection is the recruitment of immune cells to lungs leading to development of multiple lung granulomas. Non-human primate models of TB indicate that on average <10% of T cells within granulomas are Mtb-responsive in terms of cytokine production. The reason for this reduced responsiveness is unknown and it may be at the core of why humans typically are unable to clear Mtb infection. There are a number of hypotheses as to why this reduced responsiveness may occur, including T cell exhaustion, direct downregulation of antigen presentation by Mtb within infected macrophages, the spatial organization of the granuloma itself, and/or recruitment of non-Mtb-specific T cells to lungs. We use a systems biology approach pairing data and modeling to dissect three of these hypotheses. We find that the structural organization of granulomas as well as recruitment of non-specific T cells likely contribute to reduced responsiveness.

Development of New Tuberculosis Drugs: Translation to Regimen Composition for Drug-Sensitive and Multidrug-Resistant Tuberculosis

Tuberculosis (TB) kills more people than any other infectious disease. Challenges for developing better treatments include the complex pathology due to within-host immune dynamics, interpatient variability in disease severity and drug pharmacokinetics-pharmacodynamics (PK-PD), and the growing emergence of resistance. Model-informed drug development using quantitative and translational pharmacology has become increasingly recognized as a method capable of drug prioritization and regimen optimization to efficiently progress compounds through TB drug development phases. In this review, we examine translational models and tools, including plasma PK scaling, site-of-disease lesion PK, host-immune and bacteria interplay, combination PK-PD models of multidrug regimens, resistance formation, and integration of data across nonclinical and clinical phases.We propose a workflow that integrates these tools with computational platforms to identify drug combinations that have the potential to accelerate sterilization, reduce relapse rates, and limit the emergence of resistance.

CaliPro: A Calibration Protocol That Utilizes Parameter Density Estimation to Explore Parameter Space and Calibrate Complex Biological Models

Introduction

Mathematical and computational modeling have a long history of uncovering mechanisms and making predictions for biological systems. However, to create a model that can provide relevant quantitative predictions, models must first be calibrated by recapitulating existing biological datasets from that system. Current calibration approaches may not be appropriate for complex biological models because: 1) many attempt to recapitulate only a single aspect of the experimental data (such as a median trend) or 2) Bayesian techniques require specification of parameter priors and likelihoods to experimental data that cannot always be confidently assigned. A new calibration protocol is needed to calibrate complex models when current approaches fall short.

Methods

Herein, we develop CaliPro, an iterative, model-agnostic calibration protocol that utilizes parameter density estimation to refine parameter space and calibrate to temporal biological datasets. An important aspect of CaliPro is the user-defined pass set definition, which specifies how the model might successfully recapitulate experimental data. We define the appropriate settings to use CaliPro.

Results

We illustrate the usefulness of CaliPro through four examples including predator-prey, infectious disease transmission, and immune response models. We show that CaliPro works well for both deterministic, continuous model structures as well as stochastic, discrete models and illustrate that CaliPro can work across diverse calibration goals.

Conclusions

We present CaliPro, a new method for calibrating complex biological models to a range of experimental outcomes. In addition to expediting calibration, CaliPro may be useful in already calibrated parameter spaces to target and isolate specific model behavior for further analysis.

A computational model tracks whole-lung Mycobacterium tuberculosis infection and predicts factors that inhibit dissemination

Mycobacterium tuberculosis (Mtb), the causative infectious agent of tuberculosis (TB), kills more individuals per year than any other infectious agent. Granulomas, the hallmark of Mtb infection, are complex structures that form in lungs, composed of immune cells surrounding bacteria, infected cells, and a caseous necrotic core. While granulomas serve to physically contain and immunologically restrain bacteria growth, some granulomas are unable to control Mtb growth, leading to bacteria and infected cells leaving the granuloma and disseminating, either resulting in additional granuloma formation (local or non-local) or spread to airways or lymph nodes. Dissemination is associated with development of active TB. It is challenging to experimentally address specific mechanisms driving dissemination from TB lung granulomas. Herein, we develop a novel hybrid multi-scale computational model, MultiGran, that tracks Mtb infection within multiple granulomas in an entire lung. MultiGran follows cells, cytokines, and bacterial populations within each lung granuloma throughout the course of infection and is calibrated to multiple non-human primate (NHP) cellular, granuloma, and whole-lung datasets. We show that MultiGran can recapitulate patterns of in vivo local and non-local dissemination, predict likelihood of dissemination, and predict a crucial role for multifunctional CD8+ T cells and macrophage dynamics for preventing dissemination.

Tuberculosis (TB) is caused by infection with Mycobacterium tuberculosis (Mtb) and kills 3 people per minute worldwide. Granulomas, spherical structures composed of immune cells surrounding bacteria, are the hallmark of Mtb infection and sometimes fail to contain the bacteria and disseminate, leading to further granuloma growth within the lung environment. To date, the mechanisms that determine granuloma dissemination events have not been characterized. We present a computational multi-scale model of granuloma formation and dissemination within primate lungs. Our computational model is calibrated to multiple experimental datasets across the cellular, granuloma, and whole-lung scales of non-human primates. We match to both individual granuloma and granuloma-population datasets, predict likelihood of dissemination events, and predict a critical role for multifunctional CD8+ T cells and macrophage-bacteria interactions to prevent infection dissemination.

CD4CD8 Double Positive T cell responses during Mycobacterium tuberculosis infection in cynomolgus macaques

BACKGROUND

Tuberculosis (TB) kills millions of people every year. CD4 and CD8 T cells are critical in the immune response against TB. T cells expressing both CD4 and CD8 (CD4CD8 T cells) are functionally active and have not been examined in the context of TB.

METHODS

We examine peripheral blood mononuclear cells (PBMC) and bronchoalveolar lavage cells (BAL) and lung granulomas from 28 cynomolgus macaques during Mycobacterium tuberculosis (Mtb) infection.

RESULTS

CD4CD8 T cells increase in frequency during early Mtb infection in PBMC and BAL from pre-infection. Peripheral, airway, and lung granuloma CD4CD8 T cells have distinct patterns and greater cytokine production than CD4 or CD8 T cells.

CONCLUSION

Our data suggest that CD4CD8 T cells transient the blood and airways early during infection to reach the granulomas where they are involved directly in the host response to Mtb.

Multiscale Agent-Based and Hybrid Modeling of the Tumor Immune Microenvironment

Multiscale systems biology and systems pharmacology are powerful methodologies that are playing increasingly important roles in understanding the fundamental mechanisms of biological phenomena and in clinical applications. In this review, we summarize the state of the art in the applications of agent-based models (ABM) and hybrid modeling to the tumor immune microenvironment and cancer immune response, including immunotherapy. Heterogeneity is a hallmark of cancer; tumor heterogeneity at the molecular, cellular, and tissue scales is a major determinant of metastasis, drug resistance, and low response rate to molecular targeted therapies and immunotherapies. Agent-based modeling is an effective methodology to obtain and understand quantitative characteristics of these processes and to propose clinical solutions aimed at overcoming the current obstacles in cancer treatment. We review models focusing on intra-tumor heterogeneity, particularly on interactions between cancer cells and stromal cells, including immune cells, the role of tumor-associated vasculature in the immune response, immune-related tumor mechanobiology, and cancer immunotherapy. We discuss the role of digital pathology in parameterizing and validating spatial computational models and potential applications to therapeutics.

The Role of Dimensionality in Understanding Granuloma Formation

Within the first 2-3 months of a Mycobacterium tuberculosis (Mtb) infection, 2-4 mm spherical structures called granulomas develop in the lungs of the infected hosts. These are the hallmark of tuberculosis (TB) infection in humans and non-human primates. A cascade of immunological events occurs in the first 3 months of granuloma formation that likely shapes the outcome of the infection. Understanding the main mechanisms driving granuloma development and function is key to generating treatments and vaccines. In vitro, in vivo, and in silico studies have been performed in the past decades to address the complexity of granuloma dynamics. This study builds on our previous 2D spatio-temporal hybrid computational model of granuloma formation in TB (GranSim) and presents for the first time a more realistic 3D implementation. We use uncertainty and sensitivity analysis techniques to calibrate the new 3D resolution to non-human primate (NHP) experimental data on bacterial levels per granuloma during the first 100 days post infection. Due to the large computational cost associated with running a 3D agent-based model, our major goal is to assess to what extent 2D and 3D simulations differ in predictions for TB granulomas and what can be learned in the context of 3D that is missed in 2D. Our findings suggest that in terms of major mechanisms driving bacterial burden, 2D and 3D models return very similar results. For example, Mtb growth rates and molecular regulation mechanisms are very important both in 2D and 3D, as are cellular movement and modulation of cell recruitment. The main difference we found was that the 3D model is less affected by crowding when cellular recruitment and movement of cells are increased. Overall, we conclude that the use of a 2D resolution in GranSim is warranted when large scale pilot runs are to be performed and if the goal is to determine major mechanisms driving infection outcome (e.g., bacterial load). To comprehensively compare the roles of model dimensionality, further tests and experimental data will be needed to expand our conclusions to molecular scale dynamics and multi-scale resolutions.

Dynamic balance of pro- and anti-inflammatory signals controls disease and limits pathology

Immune responses to pathogens are complex and not well understood in many diseases, and this is especially true for infections by persistent pathogens. One mechanism that allows for long-term control of infection while also preventing an over-zealous inflammatory response from causing extensive tissue damage is for the immune system to balance pro- and anti-inflammatory cells and signals. This balance is dynamic and the immune system responds to cues from both host and pathogen, maintaining a steady state across multiple scales through continuous feedback. Identifying the signals, cells, cytokines, and other immune response factors that mediate this balance over time has been difficult using traditional research strategies. Computational modeling studies based on data from traditional systems can identify how this balance contributes to immunity. Here we provide evidence from both experimental and mathematical/computational studies to support the concept of a dynamic balance operating during persistent and other infection scenarios. We focus mainly on tuberculosis, currently the leading cause of death due to infectious disease in the world, and also provide evidence for other infections. A better understanding of the dynamically balanced immune response can help shape treatment strategies that utilize both drugs and host-directed therapies.

Integrating Non-human Primate, Human, and Mathematical Studies to Determine the Influence of BCG Timing on H56 Vaccine Outcomes

Tuberculosis (TB) is the leading cause of death by an infectious agent, and developing an effective vaccine is an important component of the WHO's EndTB Strategy. Non-human primate (NHP) models of vaccination are crucial to TB vaccine development and have informed design of subsequent human trials. However, challenges emerge when translating results from animal models to human applications, and connecting post-vaccination immunological measurements to infection outcomes. The H56:IC31 vaccine is a candidate currently in phase I/IIa trials. H56 is a subunit vaccine that is comprised of 3 mycobacterial antigens: ESAT6, Ag85B, and Rv2660, formulated in IC31 adjuvant. H56, as a boost to Bacillus Calmette-Guérin (BCG, the TB vaccine that is currently used in most countries world-wide) demonstrates improved protection (compared to BCG alone) in mouse and NHP models of TB, and the first human study of H56 reported strong antigen-specific T cell responses to the vaccine. We integrated NHP and human data with mathematical modeling approaches to improve our understanding of NHP and human response to vaccine. We use a mathematical model to describe T-cell priming, proliferation, and differentiation in lymph nodes and blood, and calibrate the model to NHP and human blood data. Using the model, we demonstrate the impact of BCG timing on H56 vaccination response and reveal a general immunogenic response to H56 following BCG prime. Further, we use uncertainty and sensitivity analyses to isolate mechanisms driving differences in vaccination response observed between NHP and human datasets. This study highlights the power of a systems biology approach: integration of multiple modalities to better understand a complex biological system.

Integrating Non-human Primate, Human, and Mathematical Studies to Determine the Influence of BCG Timing on H56 Vaccine Outcomes

Background

Acute lung injury (ALI) is characterized by suppressed fibrinolytic activity in bronchoalveolar lavage fluid (BALF) attributed to elevated plasminogen activator inhibitor-1 (PAI-1). Restoring pulmonary fibrinolysis by delivering tissue-type plasminogen activator (tPA), urokinase plasminogen activator (uPA), and plasmin could be a promising approach.

Objectives

To systematically analyze the overall benefit of fibrinolytic therapy for ALI reported in preclinical studies.

Methods

We searched PubMed, Embase, Web of Science, and CNKI Chinese databases, and analyzed data retrieved from 22 studies for the beneficial effects of fibrinolytics on animal models of ALI.

Results

Both large and small animals were used with five routes for delivering tPA, uPA, and plasmin. Fibrinolytics significantly increased the fibrinolytic activity both in the plasma and BALF. Fibrin degradation products in BALF had a net increase of 408.41 ng/ml vs controls (P < 0.00001). In addition, plasma thrombin–antithrombin complexes increased 1.59 ng/ml over controls (P = 0.0001). In sharp contrast, PAI-1 level in BALF decreased 21.44 ng/ml compared with controls (P < 0.00001). Arterial oxygen tension was improved by a net increase of 15.16 mmHg, while carbon dioxide pressure was significantly reduced (11.66 mmHg, P = 0.0001 vs controls). Additionally, fibrinolytics improved lung function and alleviated inflammation response: the lung wet/dry ratio was decreased 1.49 (P < 0.0001 vs controls), lung injury score was reduced 1.83 (P < 0.00001 vs controls), and BALF neutrophils were lesser (3 × 10⁴/ml, P < 0.00001 vs controls). The mortality decreased significantly within defined study periods (6 h to 30 days for mortality), as the risk ratio of death was 0.2-fold of controls (P = 0.0008).

Conclusion

We conclude that fibrinolytic therapy may be effective pharmaceutic strategy for ALI in animal models.

Addressing diversity in tuberculosis using multidimensional approaches

Tuberculosis is a complex disease, which can affect many organs other than the lungs. Initial infection may be cleared without inducing immunological memory, or progress directly to primary disease. Alternatively, the infection may be controlled as latent TB infection, that may progress to active tuberculosis at a later stage. There is now a greater understanding that these infection states are part of a continuum, and studies using PET/CT imaging have shown that individual lung granulomas may respond to infection independently, in an un-synchronized manner. In addition, the Mycobacterium tuberculosis organisms themselves can exist in different states: as nonculturable forms, as 'persisters', as rapidly growing bacteria and a biofilm-forming cording phenotype. The 'omics' approaches of transcriptomics, metabolomics and proteomics can help reveal the mechanisms underlying these different infection states in the host, and identify biosignatures with diagnostic potential, that can predict the development of disease, in 'progressors' as early as 12-18 months before it can be detected clinically, or that can monitor the success of anti-TB therapy. Further insights can be obtained from studies of BCG vaccination and new TB vaccines. For example, epigenetic changes associated with trained immunity and a stronger immune responses following BCG vaccination can be identified. These omics approaches may be particularly valuable when linked to studies of mycobacterial growth inhibition, as a direct read-out of the ability to control mycobacterial growth. The second generation of omics studies is identifying much smaller signatures based on as few as 3 or 4 genes. Thus, narrowing down omics-derived biosignatures to a manageable set of markers now opens the way to field-friendly point of care assays.

Deletion of TGF-β1 Increases Bacterial Clearance by Cytotoxic T Cells in a Tuberculosis Granuloma Model

Mycobacterium tuberculosis is the pathogenic bacterium that causes tuberculosis (TB), one of the most lethal infectious diseases in the world. The only vaccine against TB is minimally protective, and multi-drug resistant TB necessitates new therapeutics to treat infection. Developing new therapies requires a better understanding of the complex host immune response to infection, including dissecting the processes leading to formation of granulomas, the dense cellular lesions associated with TB. In this work, we pair experimental and computational modeling studies to explore cytokine regulation in the context of TB. We use our next-generation hybrid multi-scale model of granuloma formation (GranSim) to capture molecular, cellular, and tissue scale dynamics of granuloma formation. We identify TGF-β1 as a major inhibitor of cytotoxic T-cell effector function in granulomas. Deletion of TGF-β1 from the system results in improved bacterial clearance and lesion sterilization. We also identify a novel dichotomous regulation of cytotoxic T cells and macrophages by TGF-β1 and IL-10, respectively. These findings suggest that increasing cytotoxic T-cell effector functions may increase bacterial clearance in granulomas and highlight potential new therapeutic targets for treating TB.

Protective immunity against tuberculosis: what does it look like and how do we find it?

•

An absence of immune correlates of protection is a barrier to vaccine development.

•

The immune mechanisms behind tuberculosis progression are not understood.

•

Fluorescent Mtb reporter strains identify permissive and controller host cells.

•

Bacterial burden can be impacted by the magnitude of host cell population.

•

Bacterial reporter strains offer new insights into host immune mechanisms.

Progress towards the development of an effective vaccine against tuberculosis is hampered by the lack of correlative readouts of immune protection, coupled with our limited understanding of the immune mechanisms that determine disease progression versus containment. In this article we discuss the value of microbial readouts of bacterial fitness to probe the host immune environments and determine those host cell subsets that promote or control bacterial growth. Ultimately, we feel that these bacterial reporters will prove to be key in understanding the immune mechanisms underpinning disease outcome, and that this knowledge is critical to any program developing vaccines or immune-modulatory therapeutics as a means of controlling tuberculosis.

Comparing efficacies of moxifloxacin, levofloxacin and gatifloxacin in tuberculosis granulomas using a multi-scale systems pharmacology approach

Granulomas are complex lung lesions that are the hallmark of tuberculosis (TB). Understanding antibiotic dynamics within lung granulomas will be vital to improving and shortening the long course of TB treatment. Three fluoroquinolones (FQs) are commonly prescribed as part of multi-drug resistant TB therapy: moxifloxacin (MXF), levofloxacin (LVX) or gatifloxacin (GFX). To date, insufficient data are available to support selection of one FQ over another, or to show that these drugs are clinically equivalent. To predict the efficacy of MXF, LVX and GFX at a single granuloma level, we integrate computational modeling with experimental datasets into a single mechanistic framework, GranSim. GranSim is a hybrid agent-based computational model that simulates granuloma formation and function, FQ plasma and tissue pharmacokinetics and pharmacodynamics and is based on extensive in vitro and in vivo data. We treat in silico granulomas with recommended daily doses of each FQ and compare efficacy by multiple metrics: bacterial load, sterilization rates, early bactericidal activity and efficacy under non-compliance and treatment interruption. GranSim reproduces in vivo plasma pharmacokinetics, spatial and temporal tissue pharmacokinetics and in vitro pharmacodynamics of these FQs. We predict that MXF kills intracellular bacteria more quickly than LVX and GFX due in part to a higher cellular accumulation ratio. We also show that all three FQs struggle to sterilize non-replicating bacteria residing in caseum. This is due to modest drug concentrations inside caseum and high inhibitory concentrations for this bacterial subpopulation. MXF and LVX have higher granuloma sterilization rates compared to GFX; and MXF performs better in a simulated non-compliance or treatment interruption scenario. We conclude that MXF has a small but potentially clinically significant advantage over LVX, as well as LVX over GFX. We illustrate how a systems pharmacology approach combining experimental and computational methods can guide antibiotic selection for TB.

Tuberculosis (TB) is caused by infection with the bacterium Mycobacterium tuberculosis (Mtb) and kills 1.5 million people each year. TB requires at least 6 months of treatment with up to four drugs, and is characterized by formation of granulomas in patient lungs. Granulomas are spherical collections of host cells and bacteria. Fluoroquinolones (FQs) are a class of drug that could help shorten TB treatment. Three FQs that are used to treat TB are: moxifloxacin (MXF), levofloxacin (LVX) or gatifloxacin (GFX). To date, it is unclear if one FQ is better than the others at treating TB, in part because little is known about how these drugs distribute and work inside the lung granulomas. We use computer simulations of Mtb infection and FQ treatment within granulomas to predict which FQ is better and why. Our computer model is calibrated to multiple experimental data sets. We compare the three FQs by multiple metrics, and predict that MXF is better than LVX and GFX because it kills bacteria more quickly, and it works better when patients miss doses. However, all three FQs are unable to kill a part of the bacterial population living in the center of granulomas. Our results can now inform future experimental studies.

Phenotypic Heterogeneity in Mycobacterium tuberculosis

The interaction between the host and the pathogen is extremely complex and is affected by anatomical, physiological, and immunological diversity in the microenvironments, leading to phenotypic diversity of the pathogen. Phenotypic heterogeneity, defined as nongenetic variation observed in individual members of a clonal population, can have beneficial consequences especially in fluctuating stressful environmental conditions. This is all the more relevant in infections caused by Mycobacterium tuberculosis wherein the pathogen is able to survive and often establish a lifelong persistent infection in the host. Recent studies in tuberculosis patients and in animal models have documented the heterogeneous and diverging trajectories of individual lesions within a single host. Since the fate of the individual lesions appears to be determined by the local tissue environment rather than systemic response of the host, studying this heterogeneity is very relevant to ensure better control and complete eradication of the pathogen from individual lesions. The heterogeneous microenvironments greatly enhance M. tuberculosis heterogeneity influencing the growth rates, metabolic potential, stress responses, drug susceptibility, and eventual lesion resolution. Single-cell approaches such as time-lapse microscopy using microfluidic devices allow us to address cell-to-cell variations that are often lost in population-average measurements. In this review, we focus on some of the factors that could be considered as drivers of phenotypic heterogeneity in M. tuberculosis as well as highlight some of the techniques that are useful in addressing this issue.

Heterogeneity in tuberculosis

Infection with Mycobacterium tuberculosis, the causative agent of tuberculosis (TB), results in a range of clinical presentations in humans. Most infections manifest as a clinically asymptomatic, contained state that is termed latent TB infection (LTBI); a smaller subset of infected individuals present with symptomatic, active TB. Within these two seemingly binary states, there is a spectrum of host outcomes that have varying symptoms, microbiologies, immune responses and pathologies. Recently, it has become apparent that there is diversity of infection even within a single individual. A good understanding of the heterogeneity that is intrinsic to TB - at both the population level and the individual level - is crucial to inform the development of intervention strategies that account for and target the unique, complex and independent nature of the local host-pathogen interactions that occur in this infection. In this Review, we draw on model systems and human data to discuss multiple facets of TB biology and their relationship to the overall heterogeneity observed in the human disease.

A review of computational and mathematical modeling contributions to our understanding of Mycobacterium tuberculosis within-host infection and treatment

Tuberculosis (TB) is an ancient and deadly disease characterized by complex host-pathogen dynamics playing out over multiple time and length scales and physiological compartments. Computational modeling can be used to integrate various types of experimental data and suggest new hypotheses, mechanisms, and therapeutic approaches to TB. Here, we offer a first-time comprehensive review of work on within-host TB models that describe the immune response of the host to infection, including the formation of lung granulomas. The models include systems of ordinary and partial differential equations and agent-based models as well as hybrid and multi-scale models that are combinations of these. Many aspects of M. tuberculosis infection, including host dynamics in the lung (typical site of infection for TB), granuloma formation, roles of cytokine and chemokine dynamics, and bacterial nutrient availability have been explored. Finally, we survey applications of these within-host models to TB therapy and prevention and suggest future directions to impact this global disease.

Model-Driven Experimentation: A New Approach to Understand Mechanisms of Tertiary Lymphoid Tissue Formation, Function, and Therapeutic Resolution

The molecular and cellular processes driving the formation of secondary lymphoid tissues have been extensively studied using a combination of mouse knockouts, lineage-specific reporter mice, gene expression analysis, immunohistochemistry, and flow cytometry. However, the mechanisms driving the formation and function of tertiary lymphoid tissue (TLT) experimental techniques have proven to be more enigmatic and controversial due to differences between experimental models and human disease pathology. Systems-based approaches including data-driven biological network analysis (gene interaction network, metabolic pathway network, cell-cell signaling, and cascade networks) and mechanistic modeling afford a novel perspective from which to understand TLT formation and identify mechanisms that may lead to the resolution of tissue pathology. In this perspective, we make the case for applying model-driven experimentation using two case studies, which combined simulations with experiments to identify mechanisms driving lymphoid tissue formation and function, and then discuss potential applications of this experimental paradigm to identify novel therapeutic targets for TLT pathology.

A Multi-Compartment Hybrid Computational Model Predicts Key Roles for Dendritic Cells in Tuberculosis Infection

Tuberculosis (TB) is a world-wide health problem with approximately 2 billion people infected with Mycobacterium tuberculosis (Mtb, the causative bacterium of TB). The pathologic hallmark of Mtb infection in humans and Non-Human Primates (NHPs) is the formation of spherical structures, primarily in lungs, called granulomas. Infection occurs after inhalation of bacteria into lungs, where resident antigen-presenting cells (APCs), take up bacteria and initiate the immune response to Mtb infection. APCs traffic from the site of infection (lung) to lung-draining lymph nodes (LNs) where they prime T cells to recognize Mtb. These T cells, circulating back through blood, migrate back to lungs to perform their immune effector functions. We have previously developed a hybrid agent-based model (ABM, labeled GranSim) describing in silico immune cell, bacterial (Mtb) and molecular behaviors during tuberculosis infection and recently linked that model to operate across three physiological compartments: lung (infection site where granulomas form), lung draining lymph node (LN, site of generation of adaptive immunity) and blood (a measurable compartment). Granuloma formation and function is captured by a spatio-temporal model (i.e., ABM), while LN and blood compartments represent temporal dynamics of the whole body in response to infection and are captured with ordinary differential equations (ODEs). In order to have a more mechanistic representation of APC trafficking from the lung to the lymph node, and to better capture antigen presentation in a draining LN, this current study incorporates the role of dendritic cells (DCs) in a computational fashion into GranSim.

Do Xpert MTB/RIF Cycle Threshold Values Provide Information about Patient Delays for Tuberculosis Diagnosis?

Introduction

Early diagnosis and initiation to appropriate treatment is vital for tuberculosis (TB) control. The XpertMTB/RIF (Xpert) assay offers rapid TB diagnosis and quantitative estimation of bacterial burden through Cycle threshold (Ct) values. We assessed whether the Xpert Ct value is associated with delayed TB diagnosis as a potential monitoring tool for TB control programme performance.

Materials and Methods

This analysis was nested in a prospective study under the routine TB surveillance procedures of the National TB Control Program in Manhiça district, Maputo province, Mozambique. Presumptive TB patients were tested using smear microscopy and Xpert. We explored the association between Xpert Ct values and self-reported delay of Xpert-positive TB patients as recorded at the time of diagnosis enrolment. Patients with >60 days of TB symptoms were considered to have long delays.

Results

Of 1,483 presumptive TB cases, 580 were diagnosed as TB of whom 505 (87.0%) were due to pulmonary TB and 302 (94.1%) were Xpert positive. Ct values (range, 9.7–46.4) showed a multimodal distribution. The median (IQR) delay was 30 (30–45) days. Ct values showed no correlation with delay (R² = 0.001, p = 0.621), nor any association with long delays: adjusted odds ratios (AOR) (95% confidence interval [CI]) comparing to >28 cycles 0.99 (0.50–1.96; p = 0.987) for 23–28 cycles, 0.93 (0.50–1.74; p = 0.828) for 16–22 cycles; and 1.05 (0.47–2.36; p = 0.897) for <16 cycles. Being HIV-negative (AOR [95% CI]), 2.05 (1.19–3.51, p = 0.009) and rural residence 1.74 (1.08–2.81, p = 0.023), were independent predictors of long delays.

Conclusion

Xpert Ct values were not associated with patient delay for TB diagnosis and cannot be used as an indicator of TB control program performance.

PET CT Identifies Reactivation Risk in Cynomolgus Macaques with Latent M. tuberculosis

Mycobacterium tuberculosis infection presents across a spectrum in humans, from latent infection to active tuberculosis. Among those with latent tuberculosis, it is now recognized that there is also a spectrum of infection and this likely contributes to the variable risk of reactivation tuberculosis. Here, functional imaging with 18F-fluorodeoxygluose positron emission tomography and computed tomography (PET CT) of cynomolgus macaques with latent M. tuberculosis infection was used to characterize the features of reactivation after tumor necrosis factor (TNF) neutralization and determine which imaging characteristics before TNF neutralization distinguish reactivation risk. PET CT was performed on latently infected macaques (n = 26) before and during the course of TNF neutralization and a separate set of latently infected controls (n = 25). Reactivation occurred in 50% of the latently infected animals receiving TNF neutralizing antibody defined as development of at least one new granuloma in adjacent or distant locations including extrapulmonary sites. Increased lung inflammation measured by PET and the presence of extrapulmonary involvement before TNF neutralization predicted reactivation with 92% sensitivity and specificity. To define the biologic features associated with risk of reactivation, we used these PET CT parameters to identify latently infected animals at high risk for reactivation. High risk animals had higher cumulative lung bacterial burden and higher maximum lesional bacterial burdens, and more T cells producing IL-2, IL-10 and IL-17 in lung granulomas as compared to low risk macaques. In total, these data support that risk of reactivation is associated with lung inflammation and higher bacterial burden in macaques with latent Mtb infection.