A systematic efficacy analysis of tuberculosis treatment with BPaL-containing regimens using a multiscale modeling approach

Tuberculosis (TB) is a life-threatening infectious disease. The standard treatment is up to 90% effective; however, it requires the administration of four antibiotics (isoniazid, rifampicin, pyrazinamide, and ethambutol [HRZE]) over long time periods. This harsh treatment process causes adherence issues for patients because of the long treatment times and a myriad of adverse effects. Therefore, the World Health Organization has focused goals of shortening standard treatment regimens for TB in their End TB Strategy efforts, which aim to reduce TB-related deaths by 95% by 2035. For this purpose, many novel and promising combination antibiotics are being explored that have recently been discovered, such as the bedaquiline, pretomanid, and linezolid (BPaL) regimen. As a result, testing the number of possible combinations with all possible novel regimens is beyond the limit of experimental resources. In this study, we present a unique framework that uses a primate granuloma modeling approach to screen many combination regimens that are currently under clinical and experimental exploration and assesses their efficacies to inform future studies. We tested well-studied regimens such as HRZE and BPaL to evaluate the validity and accuracy of our framework. We also simulated additional promising combination regimens that have not been sufficiently studied clinically or experimentally, and we provide a pipeline for regimen ranking based on their efficacies in granulomas. Furthermore, we showed a correlation between simulation rankings and new marmoset data rankings, providing evidence for the credibility of our framework. This framework can be adapted to any TB regimen and can rank any number of single or combination regimens.

Toward mechanistic medical digital twins: some use cases in immunology

A fundamental challenge for personalized medicine is to capture enough of the complexity of an individual patient to determine an optimal way to keep them healthy or restore their health. This will require personalized computational models of sufficient resolution and with enough mechanistic information to provide actionable information to the clinician. Such personalized models are increasingly referred to as medical digital twins. Digital twin technology for health applications is still in its infancy, and extensive research and development is required. This article focuses on several projects in different stages of development that can lead to specific-and practical-medical digital twins or digital twin modeling platforms. It emerged from a two-day forum on problems related to medical digital twins, particularly those involving an immune system component. Open access video recordings of the forum discussions are available.

Forum on immune digital twins: a meeting report

Medical digital twins are computational models of human biology relevant to a given medical condition, which are tailored to an individual patient, thereby predicting the course of disease and individualized treatments, an important goal of personalized medicine. The immune system, which has a central role in many diseases, is highly heterogeneous between individuals, and thus poses a major challenge for this technology. In February 2023, an international group of experts convened for two days to discuss these challenges related to immune digital twins. The group consisted of clinicians, immunologists, biologists, and mathematical modelers, representative of the interdisciplinary nature of medical digital twin development. A video recording of the entire event is available. This paper presents a synopsis of the discussions, brief descriptions of ongoing digital twin projects at different stages of progress. It also proposes a 5-year action plan for further developing this technology. The main recommendations are to identify and pursue a small number of promising use cases, to develop stimulation-specific assays of immune function in a clinical setting, and to develop a database of existing computational immune models, as well as advanced modeling technology and infrastructure.

Structural identifiability analysis of age-structured PDE epidemic models

Computational and mathematical models rely heavily on estimated parameter values for model development. Identifiability analysis determines how well the parameters of a model can be estimated from experimental data. Identifiability analysis is crucial for interpreting and determining confidence in model parameter values and to provide biologically relevant predictions. Structural identifiability analysis, in which one assumes data to be noiseless and arbitrarily fine-grained, has been extensively studied in the context of ordinary differential equation (ODE) models, but has not yet been widely explored for age-structured partial differential equation (PDE) models. These models present additional difficulties due to increased number of variables and partial derivatives as well as the presence of boundary conditions. In this work, we establish a pipeline for structural identifiability analysis of age-structured PDE models using a differential algebra framework and derive identifiability results for specific age-structured models. We use epidemic models to demonstrate this framework because of their wide-spread use in many different diseases and for the corresponding parallel work previously done for ODEs. In our application of the identifiability analysis pipeline, we focus on a Susceptible-Exposed-Infected model for which we compare identifiability results for a PDE and corresponding ODE system and explore effects of age-dependent parameters on identifiability. We also show how practical identifiability analysis can be applied in this example.

A multi-scale pipeline linking drug transcriptomics with pharmacokinetics predicts in vivo interactions of tuberculosis drugs

Tuberculosis (TB) is the deadliest infectious disease worldwide. The design of new treatments for TB is hindered by the large number of candidate drugs, drug combinations, dosing choices, and complex pharmaco-kinetics/dynamics (PK/PD). Here we study the interplay of these factors in designing combination therapies by linking a machine-learning model, INDIGO-MTB, which predicts in vitro drug interactions using drug transcriptomics, with a multi-scale model of drug PK/PD and pathogen-immune interactions called GranSim. We calculate an in vivo drug interaction score (iDIS) from dynamics of drug diffusion, spatial distribution, and activity within lesions against various pathogen sub-populations. The iDIS of drug regimens evaluated against non-replicating bacteria significantly correlates with efficacy metrics from clinical trials. Our approach identifies mechanisms that can amplify synergistic or mitigate antagonistic drug interactions in vivo by modulating the relative distribution of drugs. Our mechanistic framework enables efficient evaluation of in vivo drug interactions and optimization of combination therapies.

Predicting the second wave of COVID-19 in Washtenaw County, MI

The COVID-19 pandemic has highlighted the patchwork nature of disease epidemics, with infection spread dynamics varying wildly across countries and across states within the US. To explore this issue, we study and predict the spread of COVID-19 in Washtenaw County, MI, which is home to University of Michigan and Eastern Michigan University, and in close proximity to Detroit, MI, a major epicenter of the epidemic in Michigan. We apply a discrete and stochastic network-based modeling framework allowing us to track every individual in the county. In this framework, we construct contact networks based on synthetic population datasets specific for Washtenaw County that are derived from US Census datasets. We assign individuals to households, workplaces, schools, and group quarters (such as prisons or long term care facilities). In addition, we assign casual contacts to each individual at random. Using this framework, we explicitly simulate Michigan-specific government-mandated workplace and school closures as well as social distancing measures. We perform sensitivity analyses to identify key model parameters and mechanisms contributing to the observed disease burden in the three months following the first observed cases of COVID-19 in Michigan. We then consider several scenarios for relaxing restrictions and reopening workplaces to predict what actions would be most prudent. In particular, we consider the effects of 1) different timings for reopening, and 2) different levels of workplace vs. casual contact re-engagement. We find that delaying reopening does not reduce the magnitude of the second peak of cases, but only delays it. Reducing levels of casual contact, on the other hand, both delays and lowers the second peak. Through simulations and sensitivity analyses, we explore mechanisms driving the magnitude and timing of a second wave of infections upon re-opening. We find that the most significant factors are workplace and casual contacts and protective measures taken by infected individuals who have sought care. This model can be adapted to other US counties using synthetic population databases and data specific to those regions.

Predicting the second wave of COVID-19 in Washtenaw County, MI

Marissa Renardy and Denise Kirschner University of Michigan Medical School The COVID-19 pandemic has highlighted the patchwork nature of disease epidemics, with infection spread dynamics varying wildly across countries and across states within the US. These heteroge- neous patterns are also observed within individual states, with patches of concentrated outbreaks. Data is being generated daily at all of these spatial scales, and answers to questions regarded re- opening strategies are desperately needed. Mathematical modeling is useful in exactly these cases, and using modeling at a county scale may be valuable to further predict disease dynamics for the purposes of public health interventions. To explore this issue, we study and predict the spread of COVID-19 in Washtenaw County, MI, the home to University of Michigan, Eastern Michigan University, and Google, as well as serving as a sister city to Detroit, MI where there has been a serious outbreak. Here, we apply a discrete and stochastic network-based modeling framework allowing us to track every individual in the county. In this framework, we construct contact net- works based on synthetic population datasets specific for Washtenaw County that are derived from US Census datasets. We assign individuals to households, workplaces, schools, and group quarters (such as prisons). In addition, we assign casual contacts to each individual at random. Using this framework, we explicitly simulate Michigan-specific government-mandated workplace and school closures as well as social distancing measures. We also perform sensitivity analyses to identify key model parameters and mechanisms contributing to the observed disease burden in the three months following the first observed cases on COVID-19 in Michigan. We then consider several scenarios for relaxing restrictions and reopening workplaces to predict what actions would be most prudent. In particular, we consider the effects of 1) different timings for reopening, and 2) different levels of workplace vs. casual contact re-engagement. Through simulations and sensitivity analyses, we explore mechanisms driving magnitude and timing of a second wave of infections upon re-opening. This model can be adapted to other US counties using synthetic population databases and data specific to those regions.

Data-Driven Model Validation Across Dimensions

Data-driven model validation across dimensions in mathematical and computational biology assumptions are often made (e.g., symmetry) to reduce the problem from three spatial dimensions (3D) to two (2D). However, some experimental datasets, such as cell counts obtained via flow cytometry, represent the entire 3D biological object. For purpose of model calibration and validation, it is sometimes necessary to compare these biological datasets with model outputs. We propose a methodology for scaling 2D model outputs to compare with 3D experimental datasets, and we discuss the application of this methodology to two examples: agent-based models of granuloma formation and skeletal muscle tissue. The accuracy of the method is evaluated in artificially generated scenarios.

Unraveling the role of fibrosis in the TB Granuloma

Tuberculosis (TB), a deadly infectious disease caused by the bacterium Mycobacterium tuberculosis (Mtb). The disease is characterized by the development of granulomas consisting of immune cells that form a cluster around the bacteria to limit bacterial growth and disease outcomes. Control of the TB epidemic is limited by a complicated drug regimen, development of antibiotic resistance, and the lack of an effective vaccine against infection and disease. Fibrosis is common in older granulomas, and has been associated with both positive and negative disease outcomes. Little is known about fibrosis in TB, partly due to the fact that fibroblasts are difficult to identify using traditional antibody-based techniques.

To provide insight into the role that fibrosis plays at a single granuloma scale, we have developed a computational, agent-based model of granuloma formation in the lung following infection with Mtb. In previously published work we have identified the mechanisms driving fibrosis within a granuloma i.e. how the granuloma environment effects fibrosis. Using immunohistochemistry, we have characterized fibroblasts and early collagen deposition in TB granulomas. Here we have extended this work to look at how fibrosis affects the ability of a granuloma to control bacteria, focusing on the role of both fibroblasts. Predictions show that early fibrosis alters the structure of the granuloma with few long-term effects on bacterial control, however late fibrosis decreases the promotability of bacterial dissemination. This work has implications on treatment options for TB that typically cause early fibrosis to occur.

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.

A population model capturing dynamics of tuberculosis granulomas predicts host infection outcomes

Granulomas play a centric role in tuberculosis (TB) infection progression. Multiple granulomas usually develop within a single host. These granulomas are not synchronized in size or bacteria load, and will follow different trajectories over time. How the fate of individual granulomas influence overall infection outcome at host scale is not understood, although computational models have been developed to predict single granuloma behavior. Here we present a within-host population model that tracks granulomas in two key organs during Mycobacteria tuberculosis (Mtb) infection: lung and lymph nodes (LN). We capture various time courses of TB progression, including latency and reactivation. The model predicts that there is no steady state; rather it is a continuous process of progressing to active disease over differing time periods. This is consistent with recently posed ideas suggesting that latent TB exists as a spectrum of states and not a single state. The model also predicts a dual role for granuloma development in LNs during Mtb infection: in early phases of infection granulomas suppress infection by providing additional antigens to the site of immune priming; however, this induces a more rapid reactivation at later stages by disrupting immune responses. We identify mechanisms that strongly correlate with better host-level outcomes, including elimination of uncontained lung granulomas by inducing low levels of lung tissue damage and inhibition of bacteria dissemination within the lung.

Macrophage polarization drives granuloma outcome during Mycobacterium tuberculosis infection (IRC5P.475)

Lung granulomas are the pathologic hallmark of tuberculosis (TB), where different immune cell types and bacteria co-localize. Macrophages are the most abundant cells in a granuloma. They are both critical effector and target cells for Mycobacterium tuberculosis (Mtb, the causative agent of TB) infection, and serve as regulators of inflammation and innate immune responses by producing a variety of pro- and anti-inflammatory cytokines. Overall, macrophage function, diversity and location within a granuloma environment are still poorly understood. In this study we used a systems biology approach to better characterize how macrophages are polarized by their microenvironment to mount either a pro- or an anti-inflammatory phenotype. We built on recently published experimental studies on non-human primates that suggest how a continuous dynamic balance between pro and anti-inflammatory signals is achieved to control pathology and restrain bacterial growth within a granuloma. Here we investigated how this balance translates into macrophage polarization and plasticity, and how the relative contribution of these macrophages, broadly classified as ‘M1’ and ‘M2’, evolves over time and leads to an effective immune response, at the granuloma scale. Our in silico granulomas suggest that temporal dynamics of granuloma polarization ratios can be predictive of granuloma outcome. We also find that a signature common to properly formed and functioning granulomas is increased NFkB degradation.

Identifying key mechanisms in fibroblast disregulation using a systems biology approach. (IRM9P.727)

Pulmonary fibrosis, stiffening and scarring of the lung tissue, is a symptom of interstitial lung disease with no available treatments and poor prognosis for patients. It occurs when disregulation of the proliferative and remodeling phases of the wound healing process results in uncontrolled proliferation and activation of fibroblasts, excessive extracellular matrix deposition, and detrimental tissue remodeling. The molecular mechanisms underlying this disregulation are poorly understood providing an untapped pool of potential therapeutic targets. TGFβ is known to be an important factor in fibroblast activation while Pge2 has been shown to counteract TGFβ; however the mechanistic details of this relationship have not been characterized. To this end, we combined computation and experimental techniques to study these cellular and molecular scale events. In vitro experiments are done in parallel with computation experiments and used to build and validate an in silico model. We created a hybrid multi-scale model focusing on the roles of TGFβ receptor ligand signaling dynamics with downstream signaling events leading to fibroblast activation, and fibroblast and epithelial cell co-regulation. Analysis of our model allows evaluation of real-time molecular events in the context of a simultaneous fibroblast/epithelial cell co-culture system. With these tools we have identified key factors driving fibroblast regulation and predict potential therapeutic targets for pulmonary fibrosis.

Predicting T cell responses to multiple antigens in lymph nodes (VAC7P.990)

T cells carrying receptors specific to an epitope are primed and differentiate in lymph nodes when they encounter their peptide-MHC counterpart displayed on dendritic cells. These epitopes are subunits of proteins derived from pathogens, but very few of the epitopes from any given pathogen elicit a measureable cell-mediated immune response. For infectious diseases lacking a protective vaccine, concurrent cellular immunity to several peptides may be necessary to target multiple stages of disease. In this work, we expand upon our previously developed computational model of T cell priming and differentiation in LNs to explore simultaneous induction of memory for multiple antigens. The model incorporates key parameters describing cell interactions, including pMHC-TCR binding parameters, cognate T cell frequency, and fraction of pMHCs specific to a TCR. Our results point to the importance of binding thresholds and pMHC fractions in determining immunodominance hierarchies. These findings are relevant to design of multi-subunit vaccines, where the aim is to simultaneously induce memory to several peptides.

Harnessing the heterogeneity of T cell differentiation fate to fine-tune generation of effector and memory T cells (LYM4P.759)

Recent studies show that naïve T cells bearing identical T cell receptors experience heterogeneous differentiation and clonal expansion processes. The factors controlling this outcome are not well characterized, and their contributions to immune cell dynamics are similarly poorly understood. In this study, we develop a computational model to elaborate mechanisms occurring within and between two important physiological compartments, lymph nodes and blood, to determine how immune cell dynamics are controlled. Our multi-organ (multi-compartment) model integrates cellular and tissue level events and allows us to examine the heterogeneous differentiation of individual precursor cognate naïve T cells to generate both effector and memory T lymphocytes. Using this model, we simulate a hypothetical immune response and reproduce both primary and recall responses to infection. Increased numbers of antigen-bearing dendritic cells are predicted to raise production of both effector and memory T cells, and distinct “sweet spots” of peptide-MHC levels on those dendritic cells exist that favor CD4+ or CD8+ T cell differentiation towards either effector or memory cell phenotypes. This has important implications for vaccine development and immunotherapy.

Harnessing the heterogeneity of T cell differentiation fate to fine-tune generation of effector and memory T cells

Recent studies show that naïve T cells bearing identical T cell receptors experience heterogeneous differentiation and clonal expansion processes. The factors controlling this outcome are not well characterized, and their contributions to immune cell dynamics are similarly poorly understood. In this study, we develop a computational model to elaborate mechanisms occurring within and between two important physiological compartments, lymph nodes and blood, to determine how immune cell dynamics are controlled. Our multi-organ (multi-compartment) model integrates cellular and tissue level events and allows us to examine the heterogeneous differentiation of individual precursor cognate naïve T cells to generate both effector and memory T lymphocytes. Using this model, we simulate a hypothetical immune response and reproduce both primary and recall responses to infection. Increased numbers of antigen-bearing dendritic cells (DCs) are predicted to raise production of both effector and memory T cells, and distinct “sweet spots” of peptide-MHC levels on those DCs exist that favor CD4+ or CD8+ T cell differentiation toward either effector or memory cell phenotypes. This has important implications for vaccine development and immunotherapy.

The roles of immune memory and aging in protective immunity and endogenous reactivation of tuberculosis

Finding more effective vaccines against tuberculosis (TB) and improved preventive treatments against endogenous reactivation of latent TB is strategic to block transmission and reach the WHO goal of eliminating TB by 2050. Key related open questions in TB research include: i) what are the determinants of a strong memory response upon primary infection? ii) what is the role of cytokines towards protective memory response against a secondary infection? iii) what are the mechanisms responsible for the increased risk of reactivation in elderly individuals? To address these questions, we explored a computational model of the immune response to Mycobacterium tuberculosis including a mathematical description of immunosenescence and the generation and maintenance of immune memory. Sensitivity analysis techniques, together with extensive model characterization and in silico experiments, were applied to identify key mechanisms controlling TB reactivation and immunological memory. Key findings of this study are summarized by the following model predictions: i) increased strength and duration of memory protection is associated with higher levels of Tumor Necrosis Factor- (TNF) during primary infection; ii) production of TNF, but not of interferon-, by memory T cells during secondary infection is a major determinant of effective protection; iii) impaired recruitment of CD4+ T cells may promote reactivation of latent TB infections in aging hosts. This study is a first attempt to consider the immune dynamics of a persistent infection throughout the lifetime of the host, taking into account immunosenescence and memory. While the model is TB specific, the results are applicable to other persistent bacterial infections and can aid in the development, evaluation and refinement of TB treatment and/or vaccine protocols.

Modeling socio-demography to capture tuberculosis transmission dynamics in a low burden setting

Evidence of preferential mixing through selected social routes has been suggested for the transmission of tuberculosis (TB) infection in low burden settings. A realistic modelization of these contact routes is needed to appropriately assess the impact of individually targeted control strategies, such as contact network investigation of index cases and treatment of latent TB infection (LTBI). We propose an age-structured, socio-demographic individual based model (IBM) with a realistic, time-evolving structure of preferential contacts in a population. In particular, transmission within households, schools and workplaces, together with a component of casual, distance-dependent contacts are considered. We also compared the model against two other formulations having no social structure of contacts (homogeneous mixing transmission): a baseline deterministic model without age structure and an age-structured IBM. The socio-demographic IBM better fitted recent longitudinal data on TB epidemiology in Arkansas, USA, which serves as an example of a low burden setting. Inclusion of age structure in the model proved fundamental to capturing actual proportions of reactivated TB cases (as opposed to recently transmitted) as well as profiling age-group specific incidence. The socio-demographic structure additionally provides a prediction of TB transmission rates (the rate of infection in household contacts and the rate of secondary cases in household and workplace contacts). These results suggest that the socio-demographic IBM is an optimal choice for evaluating current control strategies, including contact network investigation of index cases, and the simulation of alternative scenarios, particularly for TB eradication targets.

A hybrid multi-compartment model of granuloma formation and T cell priming in tuberculosis

Tuberculosis is a worldwide health problem with 2 billion people infected with Mycobacterium tuberculosis (Mtb, the bacteria causing TB). The hallmark of infection is the emergence of organized structures of immune cells forming primarily in the lung in response to infection. Granulomas physically contain and immunologically restrain bacteria that cannot be cleared. We have developed several models that spatially characterize the dynamics of the host-mycobacterial interaction, and identified mechanisms that control granuloma formation and development. In particular, we published several agent-based models (ABMs) of granuloma formation in TB that include many subtypes of T cell populations, macrophages as well as key cytokine and chemokine effector molecules. These ABM studies emphasize the important role of T-cell related mechanisms in infection progression, such as magnitude and timing of T cell recruitment, and macrophage activation. In these models, the priming and recruitment of T cells from the lung draining lymph node (LN) was captured phenomenologically. In addition to these ABM studies, we have also developed several multi-organ models using ODEs to examine trafficking of cells between, for example, the lung and LN. While we can predict temporal dynamic behaviors, those models are not coupled to the spatial aspects of granuloma. To this end, we have developed a multi-organ model that is hybrid: an ABM for the lung compartment and a non-linear system of ODE representing the lymph node compartment. This hybrid multi-organ approach to study TB granuloma formation in the lung and immune priming in the LN allows us to dissect protective mechanisms that cannot be achieved using the single compartment or multi-compartment ODE system. The main finding of this work is that trafficking of important cells known as antigen presenting cells from the lung to the lymph node is a key control mechanism for protective immunity: the entire spectrum of infection outcomes can be regulated by key immune cell migration rates. Our hybrid multi-organ implementation suggests that effector CD4+ T cells can rescue the system from a persistent infection and lead to clearance once a granuloma is fully formed. This could be effective as an immunotherapy strategy for latently infected individuals.

Deep subthreshold Xi;{-} production in Ar + KCl reactions at 1.76A GeV

We report first results on a deep subthreshold production of the doubly strange hyperon Xi;{-} in a heavy-ion reaction. At a beam energy of 1.76A GeV the reaction Ar + KCl was studied with the High Acceptance Di-Electron Spectrometer at SIS18/GSI. A high-statistics and high-purity Lambda sample was collected, allowing for the investigation of the decay channel Xi;{-} --> Lambdapi;{-}. The deduced Xi;{-}/(Lambda + Sigma;{0}) production ratio of (5.6 +/- 1.2_{-1.7};{+1.8}) x 10;{-3} is significantly larger than available model predictions.

On the global dynamics of a model for tumor immunotherapy

Understanding the dynamics of human hosts and tumors is of critical importance. A mathematical model was developed that explored the immune response to tumors that was used to study a special type of treatment. This treatment approach uses elements of the host to boost its immune response in the hopes that the host can clear the tumor. This model was extensively studied using numerical simulation, however no global analytical results were originally presented. In this work we explore the global dynamics to show under what conditions tumor clearance can be achieved.

Neutralization of Tumor Necrosis Factor (TNF) by Antibody but not TNF Receptor Fusion Molecule Exacerbates Chronic Murine Tuberculosis

Tumor necrosis factor (TNF) plays an essential role in the immunologic maintenance of Mycobacterium tuberculosis infection. Although an increased rate of tuberculosis has been reported in humans treated with anti-TNF biological agents, disparate rates of disease have been observed between those treated with infliximab, an anti-TNF antibody, and etanercept, a TNF-neutralizing TNF receptor (TNFR) fusion molecule. We compared the effects of anti-TNF antibody and soluble TNFR fusion molecule in the murine model of tuberculosis. Systemic TNF neutralization was equivalent between these molecules, and both resulted in rapid morbidity at the initiation of infection. During chronic infection, administration of the receptor fusion molecule allowed the control of infection, whereas antibody treatment caused mice to die within a month. We provide evidence of decreased penetration into the granulomas by the receptor fusion molecule, compared with antibody. These findings begin to clarify the mechanistic difference between anti-TNF agents and their role in the exacerbation of tuberculosis.

Identifying DNA splice sites using hypernetworks with artificial molecular evolution

Identifying DNA splice sites is a main task of gene hunting. We introduce the hyper-network architecture as a novel method for finding DNA splice sites. The hypernetwork architecture is a biologically inspired information processing system composed of networks of molecules forming cells, and a number of cells forming a tissue or organism. Its learning is based on molecular evolution. DNA examples taken from GenBank were translated into binary strings and fed into a hypernetwork for training. We performed experiments to explore the generalization performance of hypernetwork learning in this data set by two-fold cross validation. The hypernetwork generalization performance was comparable to well known classification algorithms. With the best hypernetwork obtained, including local information and heuristic rules, we built a system (HyperExon) to obtain splice site candidates. The HyperExon system outperformed leading splice recognition systems in the list of sequences tested.

Mycobacterium tuberculosis as viewed through a computer

Mathematical models are emerging as important tools in the study of microbiology. As an illustrative example, we present results from several models each generated to study the interaction of Mycobacterium tuberculosis and the immune system. Different mathematical models were formulated on the basis of assumptions regarding system-component interactions, enabling us to explore specific aspects at diverse biological scales (e.g. intracellular, cell-cell interactions, and cell population dynamics). In addition, we were able to examine both temporal and spatial aspects. At each scale, there were consistent themes that emerged as determinative in infection outcome. Factors we identified include both host and microbial characteristics. The use of the models lies in generating hypotheses that can then be tested experimentally. Here, we outline the primary host and bacterial factors that we have identified as key mechanisms that contribute to the success of M. tuberculosis as a human pathogen. Our goal is to stimulate experimentation and foster collaborations between theoretical and experimental scientists.

Identifying control mechanisms of granuloma formation during M. tuberculosis infection using an agent-based model

Infection with Mycobacterium tuberculosis is a major world health problem. An estimated 2 billion people are presently infected and the disease causes approximately 3 million deaths per year. After bacteria are inhaled into the lung, a complex immune response is triggered leading to the formation of multicellular structures termed granulomas. It is believed that the collection of host granulomas either contain bacteria resulting in a latent infection or are unable to do so, leading to active disease. Thus, understanding granuloma formation and function is essential for improving both diagnosis and treatment of tuberculosis. Granuloma formation is a complex spatio-temporal system involving interactions of bacteria, specific immune cells, including macrophages, CD4+ and CD8+ T cells, as well as immune effectors such as chemokine and cytokines. To study this complex dynamical system we have developed an agent-based model of granuloma formation in the lung. This model combines continuous representations of chemokines with discrete agent representations of macrophages and T cells in a cellular automata-like environment. Our results indicate that key host elements involved in granuloma formation are chemokine diffusion, prevention of macrophage overcrowding within the granuloma, arrival time, location and number of T cells within the granuloma, and an overall host ability to activate macrophages. Interestingly, a key bacterial factor is its intracellular growth rate, whereby slow growth actually facilitates survival.

A model for the study of Helicobacter pylori interaction with human gastric acid secretion

We present a comprehensive mathematical model describing Helicobacter pylori interaction with the human gastric acid secretion system. We use the model to explore host and bacterial conditions that allow persistent infection to develop and be maintained. Our results show that upon colonization, there is a transient period (day 1-20 post-infection) prior to the establishment of persistence. During this period, changes to host gastric physiology occur including elevations in positive effectors of acid secretion (such as gastrin and histamine). This is promoted by reduced somatostatin levels, an inhibitor of acid release. We suggest that these changes comprise compensatory mechanisms aimed at restoring acid to pre-infection levels. We also show that ammonia produced by bacteria sufficiently buffers acid promoting bacteria survival and growth.

Analysis of CO2 leakage through ‘low-permeability’ faults from natural reservoirs in the Colorado Plateau, east-central Utah

The numerous CO2 reservoirs in the Colorado Plateau region of the United States are natural analogues for potential geological CO2 sequestration repositories. To understand better the risk of leakage from reservoirs used for long-term underground CO2 storage, we examine evidence for CO2 migration along two normal faults that cut a reservoir in east-central Utah. CO2-charged springs, geysers, and a hydrocarbon seep are localized along these faults. These include natural springs that have been active for long periods of time, and springs that were induced by recent drilling. The CO2-charged spring waters have deposited travertine mounds and carbonate veins. The faults cut siltstones, shales, and sandstones and the fault rocks are fine-grained, clay-rich gouge, generally thought to be barriers to fluid flow. The geological and geochemical data are consistent with these faults being conduits for CO2 moving to the surface. Consequently, the injection of CO2 into faulted geological reservoirs, including faults with clay gouge, must be carefully designed and monitored to avoid slow seepage or fast rupture to the biosphere.

On treatment of tuberculosis in heterogeneous populations

Global eradication of tuberculosis (TB) is an international agenda. Thus understanding effects of treatment of TB in different settings is crucial. In previous work, we introduced the framework for a mathematical model of epidemic TB in demographically distinct, heterogeneous populations. Simulations showed the importance of genetic susceptibility in determining endemic prevalence levels. In the work presented here, we include treatment and investigate different strategies for treatment of latent and active TB disease in heterogeneous populations. We illustrate how the presence of a genetically susceptible subpopulation dramatically alters effects of treatment in the same way a core population does in the setting of sexually transmitted diseases. In addition, we evaluate treatment strategies that focus specifically on this subpopulation, and our results indicate that genetically susceptible subpopulations should be accounted for when designing treatment strategies to achieve the greatest reduction in disease prevalence.

A model for integrative study of human gastric acid secretion

We have developed a unique virtual human model of gastric acid secretion and its regulation in which food provides a driving force. Food stimulus triggers neural activity in central and enteric nervous systems and G cells to release gastrin, a critical stimulatory hormone. Gastrin stimulates enterochromaffin-like cells to release histamine, which, together with acetylcholine, stimulates acid secretion from parietal cells. Secretion of somatostatin from antral and corpus D cells comprises a negative-feedback loop. We demonstrate that although acid levels are most sensitive to food and nervous system inputs, somatostatin-associated interactions are also important in governing acidity. The importance of gastrin in acid secretion is greatest at the level of transport between the antral and corpus regions. Our model can be applied to study conditions that are not yet experimentally reproducible. For example, we are able to preferentially deplete antral or corpus somatostatin. Depletion of antral somatostatin exhibits a more significant elevation of acid release than depletion of corpus somatostatin. This increase in acid release is likely due to elevated gastrin levels. Prolonged hypergastrinemia has significant effects in the long term (5 days) by promoting enterochromaffin-like cell overgrowth. Our results may be useful in the design of therapeutic strategies for acid secretory dysfunctions such as hyper- and hypochlorhydria.

Comparing epidemic tuberculosis in demographically distinct heterogeneous populations

There is wide variation in endemic tuberculosis (TB) levels between countries and we seek to identify possible causes of these differences. In this study we present an epidemiological model of Mycobacterium tuberculosis infection to investigate the effects of host genetics and demographic factors on epidemic TB. We discuss the general framework for this approach and present analytical results to identify important parameters affecting steady-state prevalence and incidence rates of TB disease. We then use numerical simulations of our model to observe the effects of a genetically susceptible subpopulation on TB disease dynamics at the population level. Finally, we simulate infection within a genetically heterogeneous population in two demographic settings: India (a typical population with high TB prevalence) and the USA (a typical population with low TB prevalence). Results show that changes in transmission parameters, the fraction of the population genetically susceptible to infection, and demographic factors strongly affect TB prevalence and incidence rates.

Mathematical analysis of the global dynamics of a model for HTLV-I infection and ATL progression

Mathematical analysis is carried out that completely determines the global dynamics of a mathematical model for the transmission of human T-cell lymphotropic virus I (HTLV-I) infection and the development of adult T-cell leukemia (ATL). HTLV-I infection of healthy CD4(+) T cells takes place through cell-to-cell contact with infected T cells. The infected T cells can remain latent and harbor virus for several years before virus production occurs. Actively infected T cells can infect other T cells and can convert to ATL cells, whose growth is assumed to follow a classical logistic growth function. Our analysis establishes that the global dynamics of T cells are completely determined by a basic reproduction number R(0). If R(0)< or =1, infected T cells always die out. If R(0)>1, HTLV-I infection becomes chronic, and a unique endemic equilibrium is globally stable in the interior of the feasible region. We also show that the equilibrium level of ATL-cell proliferation is higher when the HTLV-I infection of T cells is chronic than when it is acute.

Dynamics of naive and memory CD4+ T lymphocytes in HIV-1 disease progression

Understanding the dynamics of naive and memory CD4+ T cells in the immune response to HIV-1 infection can help elucidate typical disease progression patterns observed in HIV-1 patients. Although infection markers such as CD4+ T-cell count and viral load are monitored in patient blood, the lymphatic tissues (LT) have been shown to be an important viral reservoir. Here, we introduce the first comprehensive theoretical model of disease progression based on T-cell subsets and virus circulating between the two compartments of LT and blood. We use this model to predict several trademarks observed in adult HIV-1 disease progression such as the establishment of a setpoint in the asymptomatic stage. Our model predicts that both host and viral elements play a role in determining different disease progression patterns. Viral factors include viral infectivity and production rates, whereas host factors include elements of specific immunity. We also predict the effect of highly active antiretroviral therapy and treatment cessation on cellular and viral dynamics in both blood and LT.

Reduction-oxidation control of beta-sheet assembly in genetically engineered silk

Genetically engineered spider dragline silk protein was modified to incorporate methionines flanking the beta-sheet forming polyalanine regions. The methionines could be selectively chemically oxidized and reduced. This chemical change altered the bulkiness and charge of the sulfhydryl groups, and in turn, the beta-sheet forming tendencies of the polyalanine domains and solubility of the protein. The genes encoding these redesigned proteins were constructed, cloned and expressed in Escherichia coli. In the reduced state (beta-mercaptoethanol) the approximately 25 kDa protein behaved similarly to native spider dragline silk, crystallizing into beta-sheets based on diffraction analysis and appearing fibrous by TEM. The addition of the methionines into the consensus dragline silk sequence did not disrupt the normal macromolecular assembly behavior of the protein. In the oxidized state (phenacyl bromide) the protein did not form beta-sheet crystals and appeared morphologically featureless based on TEM. A reduction in beta-strand content was also observed upon oxidation based on FTIR and TEM analysis and confirmed by X-ray diffraction analysis. To further confirm changes in assembly behavior observed for the recombinant protein containing the methionines, a model peptide with the same repeat amino acid sequence was synthesized and characterized. Shifts in molecular weight, observed by MALDI, along with corresponding changes in crystallinity, by electron diffraction, agreed with the changes expected on activation and deactivation of the redox trigger. These results support the use of a redox trigger as a useful feature with which to control the assembly of beta-sheet forming proteins.

The coreceptor mutation CCR5Delta32 influences the dynamics of HIV epidemics and is selected for by HIV

We explore the impact of a host genetic factor on heterosexual HIV epidemics by using a deterministic mathematical model. A protective allele unequally distributed across populations is exemplified in our models by the 32-bp deletion in the host-cell chemokine receptor CCR5, CCR5Delta32. Individuals homozygous for CCR5Delta32 are protected against HIV infection whereas those heterozygous for CCR5Delta32 have lower pre-AIDS viral loads and delayed progression to AIDS. CCR5Delta32 may limit HIV spread by decreasing the probability of both risk of infection and infectiousness. In this work, we characterize epidemic HIV within three dynamic subpopulations: CCR5/CCR5 (homozygous, wild type), CCR5/CCR5Delta32 (heterozygous), and CCR5Delta32/CCR5Delta32 (homozygous, mutant). Our results indicate that prevalence of HIV/AIDS is greater in populations lacking the CCR5Delta32 alleles (homozygous wild types only) as compared with populations that include people heterozygous or homozygous for CCR5Delta32. Also, we show that HIV can provide selective pressure for CCR5Delta32, increasing the frequency of this allele.

Histone folds mediate selective heterodimerization of yeast TAF(II)25 with TFIID components yTAF(II)47 and yTAF(II)65 and with SAGA component ySPT7

We show that the yeast TFIID (yTFIID) component yTAF(II)47 contains a histone fold domain (HFD) with homology to that previously described for hTAF(II)135. Complementation in vivo indicates that the yTAF(II)47 HFD is necessary and sufficient for vegetative growth. Mutation of highly conserved residues in the alpha1 helix of the yTAF(II)47 HFD results in a temperature-sensitive phenotype which can be suppressed by overexpression of yTAF(II)25, as well as by yTAF(II)40, yTAF(II)19, and yTAF(II)60. In yeast two-hybrid and bacterial coexpression assays, the yTAF(II)47 HFD selectively heterodimerizes with yTAF(II)25, which we show contains an HFD with homology to the hTAF(II)28 family We additionally demonstrate that yTAF(II)65 contains a functional HFD which also selectively heterodimerizes with yTAF(II)25. These results reveal the existence of two novel histone-like pairs in yTFIID. The physical and genetic interactions described here show that the histone-like yTAF(II)s are organized in at least two substructures within TFIID rather than in a single octamer-like structure as previously suggested. Furthermore, our results indicate that ySPT7 has an HFD homologous to that of yTAF(II)47 which selectively heterodimerizes with yTAF(II)25, defining a novel histone-like pair in the SAGA complex.

A model to predict cell-mediated immune regulatory mechanisms during human infection with Mycobacterium tuberculosis

A key issue for the study of tuberculosis infection (TB) is to understand why individuals infected with Mycobacterium tuberculosis experience different clinical outcomes. Elaborating the immune mechanisms that determine whether an infected individual will suffer active TB or latent infection can aid in developing treatment and prevention strategies. To better understand the dynamics of M. tuberculosis infection and immunity, we have developed a virtual human model that qualitatively and quantitatively characterizes the cellular and cytokine control network operational during TB infection. Using this model, we identify key regulatory elements in the host response. In particular, factors affecting cell functions, such as macrophage activation and bactericidal capabilities, and effector T cell functions such as cytotoxicity and cytokine production can each be determinative. The model indicates, however, that even if latency is achieved, it may come at the expense of tissue damage if the response is not properly regulated. A balance in Th1 and Th2 immune responses governed by IFN-gamma, IL-10, and IL-4 facilitate this down-regulation. These results are further explored through virtual deletion and depletion experiments.

Model of HIV-1 disease progression based on virus-induced lymph node homing and homing-induced apoptosis of CD4+ lymphocytes

Several proposed theories have described the progression of HIV infection. Even so, no concrete evidence supports any as comprehensive, including, for example, why the CD4+ T-cell counts fall from 1000/mm3 of blood to roughly 100/mm3 over an average 10-year period, whereas concomitant viral loads are relatively constant, increasing by several orders of magnitude in late-stage disease. Here, we develop and validate a theoretical model that altered lymphocyte circulation patterns between the lymph system and blood due to HIV-induced enhanced lymph-node homing and subsequent apoptosis of resting CD4+ T cells can explain many aspects of HIV-1 disease progression. These results lead to a recalculation of the CD4+ lymphocyte dynamics during highly active antiretroviral therapy, and also suggest new targets for therapy.

Theoretical and experimental approaches for studying factors defining the Helicobacter pylori-host relationship

Mathematical modeling has helped develop hypotheses about the role of microbial and host parameters in the initial and subsequent phases of Helicobacter pylori colonization. Transgenic mice have been used to test the hypothesis that the outcome of colonization is influenced by whether bacteria can adhere to available epithelial cell receptors. Complementary use of modeling and experimental approaches should facilitate studies of H. pylori pathogenesis.

Dynamics of Helicobacter pylori colonization in relation to the host response

The dynamics of Helicobacter pylori colonization from its acquisition through the development of steady-state are examined through a mathematical model that includes the host response. The model encompasses both host and microbiological variation. The individual capacity of the host response is shown to be a key model parameter, leading to either transient or persistent colonization, whereas the growth rate of that response has little effect. Analyses of competing strains indicate that each must occupy a specific niche, otherwise exclusion occurs. The model implies that there exists a lower bound on the host response to the indigenous microflora that is consistent with current biological views of H. pylori. Parallel models may be useful in understanding other persistent host-microbial interactions.

Dynamics of co-infection with M. Tuberculosis and HIV-1

Since 1985, there has been a renewed epidemic of tuberculosis (TB) that was previously thought to be in check. There is evidence to believe the main factor for this resurgence has been the human immunodeficiency virus (HIV). Co-infection with HIV and M. Tuberculosis has profound implications for the course of both diseases. This study represents a first attempt to understand how the introduction of an opportunistic infection, namely Mycobacterium tuberculosis, the bacteria that causes TB, affects the dynamic interaction of HIV-1 and the immune system. We create a mathematical model using ordinary differential equations to describe the interaction of HIV and TB with the immune system. It is known that infection with TB can decrease the CD4(+) T cell counts-a key marker of AIDS progression; thus, it shortens survival in HIV infected individuals. Another main marker for HIV progression is the viral load. If this load is increased due to the presence of opportunistic infections, the disease progression is much more rapid. We also explore the effects of drug treatment on the TB infection in the doubly-infected patient.

Modeling immunotherapy of the tumor-immune interaction

A number of lines of evidence suggest that immunotherapy with the cytokine interleukin-2 (IL-2) may boost the immune system to fight tumors. CD4+ T cells, the cells that orchestrate the immune response, use these cytokines as signaling mechanisms for immune-response stimulation as well as lymphocyte stimulation, growth, and differentiation. Because tumor cells begin as 'self', the immune system may not respond in an effective way to eradicate them. Adoptive cellular immunotherapy can potentially restore or enhance these effects. We illustrate through mathematical modeling the dynamics between tumor cells, immune-effector cells, and IL-2. These efforts are able to explain both short tumor oscillations in tumor sizes as well as long-term tumor relapse. We then explore the effects of adoptive cellular immunotherapy on the model and describe under what circumstances the tumor can be eliminated.