Publications

IFN-γ-producing CD4 T cells are required for protection against Mycobacterium tuberculosis (Mtb) infection, but the extent to which IFN-γ contributes to overall CD4 T cell-mediated protection remains unclear. Furthermore, it is not known if increasing IFN-γ production by CD4 T cells is desirable in Mtb infection. Here we show that IFN-γ accounts for only  30% of CD4 T cell-dependent cumulative bacterial control in the lungs over the first six weeks of infection, but >80% of control in the spleen. Moreover, increasing the IFN-γ-producing capacity of CD4 T cells by  2 fold exacerbates lung infection and leads to the early death of the host, despite enhancing control in the spleen. In addition, we show that the inhibitory receptor PD-1 facilitates host resistance to Mtb by preventing the detrimental over-production of IFN-γ by CD4 T cells. Specifically, PD-1 suppressed the parenchymal accumulation of and pathogenic IFN-γ production by the CXCR3+KLRG1-CX3CR1- subset of lung-homing CD4 T cells that otherwise mediates control of Mtb infection. Therefore, the primary role for T cell-derived IFN-γ in Mtb infection is at extra-pulmonary sites, and the host-protective subset of CD4 T cells requires negative regulation of IFN-γ production by PD-1 to prevent lethal immune-mediated pathology.

Upon infection of a new host, human immunodeficiency virus (HIV) replicates in the mucosal tissues and is generally undetectable in circulation for 1-2 weeks post-infection. Several interventions against HIV including vaccines and antiretroviral prophylaxis target virus replication at this earliest stage of infection. Mathematical models have been used to understand how HIV spreads from mucosal tissues systemically and what impact vaccination and/or antiretroviral prophylaxis has on viral eradication. Because predictions of such models have been rarely compared to experimental data, it remains unclear which processes included in these models are critical for predicting early HIV dynamics. Here we modified the "standard" mathematical model of HIV infection to include two populations of infected cells: cells that are actively producing the virus and cells that are transitioning into virus production mode. We evaluated the effects of several poorly known parameters on infection outcomes in this model and compared model predictions to experimental data on infection of non-human primates with variable doses of simian immunodifficiency virus (SIV). First, we found that the mode of virus production by infected cells (budding vs. bursting) has a minimal impact on the early virus dynamics for a wide range of model parameters, as long as the parameters are constrained to provide the observed rate of SIV load increase in the blood of infected animals. Interestingly and in contrast with previous results, we found that the bursting mode of virus production generally results in a higher probability of viral extinction than the budding mode of virus production. Second, this mathematical model was not able to accurately describe the change in experimentally determined probability of host infection with increasing viral doses. Third and finally, the model was also unable to accurately explain the decline in the time to virus detection with increasing viral dose. These results suggest that, in order to appropriately model early HIV/SIV dynamics, additional factors must be considered in the model development. These may include variability in monkey susceptibility to infection, within-host competition between different viruses for target cells at the initial site of virus replication in the mucosa, innate immune response, and possibly the inclusion of several different tissue compartments. The sobering news is that while an increase in model complexity is needed to explain the available experimental data, testing and rejection of more complex models may require more quantitative data than is currently available.

Plasmodium remains a major pathogen causing malaria and impairing defense against other infections. Defining how Plasmodium increases susceptibility to heterologous pathogens may lead to interventions that mitigate the severity of coinfections. Previous studies proposed that reduced T cell responses during coinfections are due to diminished recruitment of naive T cells through infection-induced decreases in chemokine CCL21. We found that, although Listeria infections reduced expression of CCL21 in murine spleens, lymphocytic choriomeningitis virus (LCMV)-specific T cell responses were not impaired during Listeria + LCMV coinfection, arguing against a major role for this chemokine in coinfection-induced T cell suppression. In our experiments, Plasmodium yoelii infection led to a reduced CD8(+) T cell response to a subsequent Listeria infection. We propose an alternative mechanism whereby P. yoelii suppresses Listeria-specific T cell responses. We found that Listeria-specific T cells expanded more slowly and resulted in lower numbers in response to coinfection with P. yoelii. Mathematical modeling and experimentation revealed greater apoptosis of Listeria-specific effector T cells as the main mechanism, because P. yoelii infections did not suppress the recruitment or proliferation rates of Listeria-specific T cells. Our results suggest that P. yoelii infections suppress immunity to Listeria by causing increased apoptosis in Listeria-specific T cells, resulting in a slower expansion rate of T cell responses.

Recent studies have highlighted the ability of HIV to escape from cytotoxic T lymphocyte (CTL) responses that concurrently target multiple viral epitopes. Yet, the viral dynamics involved in such escape are incompletely understood. Previous analyses have made several strong assumptions regarding HIV escape from CTL responses such as independent or non-concurrent escape from individual CTL responses. Using experimental data from evolution of HIV half genomes in four patients we observe concurrent viral escape from multiple CTL responses during early infection (first 100 days of infection), providing confirmation of a recent result found in a study of one HIV-infected patient. We show that current methods of estimating CTL escape rates, based on the assumption of independent escapes, are biased and perform poorly when CTL escape proceeds concurrently at multiple epitopes. We propose a new method for analyzing longitudinal sequence data to estimate the rate of CTL escape across multiple epitopes; this method involves few parameters and performs well in simulation studies. By applying our novel method to experimental data, we find that concurrent multiple escapes occur at rates between 0.03 and 0.4 day(-1), a relatively broad range that reflects uncertainty due to sparse sampling and wide ranges of parameter values. However, we show that concurrent escape at rates 0.1-0.2 day(-1) across multiple epitopes is consistent with our patient datasets.

Mycobacterium avium spp. paratuberculosis (MAP) causes a persistent infection and chronic inflammation of the gut in ruminants leading to bacterial shedding in feces in many infected animals. Although there are often strong MAP-specific immune responses in infected animals, immunological correlates of protection against progression to disease remain poorly defined. Analysis of cross-sectional data has suggested that the cellular immune response observed early in infection is effective at containing bacterial growth and shedding, in contrast to humoral immune responses. In this study, 20 MAP-infected calves were followed for nearly 5 years during which MAP shedding, antigen-specific cellular (LPT) and humoral (ELISA) immune responses were measured. We found that MAP-specific cellular immune response developed slowly, with the peak of the immune response occurring one year post infection. MAP-specific humoral immunity expanded only in a subset of animals. Only in a subset of animals there was a statistically significant negative correlation between the amount of MAP shedding and magnitude of the MAP-specific cellular immune response. Direct fitting of simple mechanistic mathematical models to the shedding data suggested that MAP-specific immune responses contributed significantly to the kinetics of MAP shedding in most animals. However, whereas the MAP-specific cellular immune response was predicted to suppress shedding in some animals, in other animals it was predicted to increase shedding. In contrast, MAP-specific humoral response was always predicted to increase shedding. Our results illustrate the use of mathematical methods to understand relationships between mycobacteria and immunity in vivo but also highlight problems with establishing cause-effect links from observational data.

With major advances in experimental techniques to track antigen-specific immune responses many basic questions on the kinetics of virus-specific immunity in humans remain unanswered. To gain insights into kinetics of T and B cell responses in human volunteers we combined mathematical models and experimental data from recent studies employing vaccines against yellow fever and smallpox. Yellow fever virus-specific CD8 T cell population expanded slowly with the average doubling time of 2 days peaking 2.5 weeks post immunization. Interestingly, we found that the peak of the yellow fever-specific CD8 T cell response was determined by the rate of T cell proliferation and not by the precursor frequency of antigen-specific cells as has been suggested in several studies in mice. We also found that while the frequency of virus-specific T cells increased slowly, the slow increase could still accurately explain clearance of yellow fever virus in the blood. Our additional mathematical model described well the kinetics of virus-specific antibody-secreting cell and antibody response to vaccinia virus in vaccinated individuals suggesting that most of antibodies in 3 months post immunization were derived from the population of circulating antibody-secreting cells. Taken together, our analysis provided novel insights into mechanisms by which live vaccines induce immunity to viral infections and highlighted challenges of applying methods of mathematical modeling to the current, state-of-the-art yet limited immunological data.

The kinetics of recirculation of naive lymphocytes in the body has important implications for the speed at which local infections are detected and controlled by immune responses. With a help of a novel mathematical model, we analyze experimental data on migration of 51Cr-labeled thoracic duct lymphocytes (TDLs) via major lymphoid and nonlymphoid tissues of rats in the absence of systemic antigenic stimulation. We show that at any point of time, 95% of lymphocytes in the blood travel via capillaries in the lung or sinusoids of the liver and only 5% migrate to secondary lymphoid tissues such as lymph nodes, Peyer's patches, or the spleen. Interestingly, our analysis suggests that lymphocytes travel via lung capillaries and liver sinusoids at an extremely rapid rate with the average residence time in these tissues being less than 1 minute. The model also predicts a relatively short average residence time of TDLs in the spleen (2.5 hours) and a longer average residence time of TDLs in major lymph nodes and Peyer's patches (10 hours). Surprisingly, we find that the average residence time of lymphocytes is similar in lymph nodes draining the skin (subcutaneous LNs) or the gut (mesenteric LNs) or in Peyer's patches. Applying our model to an additional dataset on lymphocyte migration via resting and antigen-stimulated lymph nodes we find that enlargement of antigen-stimulated lymph nodes occurs mainly due to increased entrance rate of TDLs into the nodes and not due to decreased exit rate as has been suggested in some studies. Taken together, our analysis for the first time provides a comprehensive, systems view of recirculation kinetics of thoracic duct lymphocytes in the whole organism.

Using a unique combination of visual, statistical, and data mining methods, we tested the hypothesis that an immune cell's movement pattern can convey key information about the cell's function, antigen specificity, and environment. We applied clustering, statistical tests, and a support vector machine (SVM) to assess our ability to classify different datasets of imaged flouresently labelled T cells in mouse liver. We additionally saw clusters of different movement patterns of T cells of identical antigenic specificity. We found that the movement patterns of T cells specific and non-specific for malaria parasites are differentiable with 72% accuracy, and that specific cells have a higher tendency to move towards the parasite than non-specific cells. Movements of antigen-specific T cells in uninfected mice vs. infected mice were differentiable with 69.8% accuracy. We additionally saw clusters of different movement patterns of T cells of identical antigenic specificity. We concluded that our combination of methods has the potential to advance the understanding of cell movements in vivo.

Johne's disease (JD), a persistent and slow progressing infection of ruminants such as cows and sheep, is caused by slow replicating bacilli Mycobacterium avium subspecies paratuberculosis (MAP) infecting macrophages in the gut. Infected animals initially mount a cell-mediated CD4 T cell response against MAP which is characterized by the production of interferon gamma (Th1 response). Over time, Th1 response diminishes in most animals and antibody response to MAP antigens becomes dominant (Th2 response). The switch from Th1 to Th2 response occurs concomitantly with disease progression and shedding of the bacteria in feces. Mechanisms controlling this Th1/Th2 switch remain poorly understood. Because Th1 and Th2 responses are known to cross-inhibit each other, it is unclear why initially strong Th1 response is lost over time. Using a novel mathematical model of the immune response to MAP infection we show that the ability of extracellular bacteria to persist outside of macrophages naturally leads to switch of the cellular response to antibody production. Several additional mechanisms may also contribute to the timing of the Th1/Th2 switch including the rate of proliferation of Th1/Th2 responses at the site of infection, efficiency at which immune responses cross-inhibit each other, and the rate at which Th1 response becomes exhausted over time. Our basic model reasonably well explains four different kinetic patterns of the Th1/Th2 responses in MAP-infected sheep by variability in the initial bacterial dose and the efficiency of the MAP-specific T cell responses. Taken together, our novel mathematical model identifies factors of bacterial and host origin that drive kinetics of the immune response to MAP and provides the basis for testing the impact of vaccination or early treatment on the duration of infection.

CD8(+) T cells are specialized cells of the adaptive immune system capable of finding and eliminating pathogen-infected cells. To date it has not been possible to observe the destruction of any pathogen by CD8(+) T cells in vivo. Here we demonstrate a technique for imaging the killing of liver-stage malaria parasites by CD8(+) T cells bearing a transgenic T cell receptor specific for a parasite epitope. We report several features that have not been described by in vitro analysis of the process, chiefly the formation of large clusters of effector CD8(+) T cells around infected hepatocytes. The formation of clusters requires antigen-specific CD8(+) T cells and signaling by G protein-coupled receptors, although CD8(+) T cells of unrelated specificity are also recruited to clusters. By combining mathematical modeling and data analysis, we suggest that formation of clusters is mainly driven by enhanced recruitment of T cells into larger clusters. We further show various death phenotypes of the parasite, which typically follow prolonged interactions between infected hepatocytes and CD8(+) T cells. These findings stress the need for intravital imaging for dissecting the fine mechanisms of pathogen recognition and killing by CD8(+) T cells.