Motility in the Immune System: From Microscopic Movement to Macroscopic Function - Abstracts

From Santa Fe Institute Events Wiki

Working Group Navigation

From individual memory to environmental memory in interaction processes: animals as territorial random walkers

Luca Giuggioli

Bristol University

Recent efforts in analysing spatio-temporal trajectories of higher level organisms aim at uncovering the role of memories in animal movement processes. It is natural to think about animals exploiting some form of memory to associate locations in the environment with high or low rewards. Examples of use of an animal's internal memory are foraging tasks when individuals return to past profitable locations, or during animal confrontations whereby individuals tend to avoid areas where they have had unsuccessful contests with a neighbour. There is however another form of memory that animals can exploit, the so-called external, or environmental, memory. It refers to situations in which individuals do not themselves retain information of past events but simply react to features that they encounter in the environment. In this scenario animal interactions may occur in the absence of internal memory with individuals reacting to stimuli found on the terrain due to the passage or marks left by another individual. This type of indirect interaction processes so far observed in eusocial insects, and coined by biologists as stigmergy, has been formulated mathematically to study the formation of territories whereby animals in a population collectively tessellate the environment in regions of exclusive use by avoiding one another rather than being attracted to specific locations over the terrain, e.g. a den or burrow. I will show that a combination of mathematical and analytical tools can be used to link some of the microscopic characteristics of the animals to those of the territories they live in.

T cell motility and the fibroblastic reticular network: monorails or monkey bars?

Mark J. Miller, Ph.D.

Washington University School of Medicine, Department of Medicine, Infectious Diseases.
660 S. Euclid Ave., St. Louis MO 63110

Two-photon imaging is widely used to study cellular immune responses in mouse models of infection and inflammatory disease. In particular, the technique has led to breakthroughs in understanding the cell dynamics of leukocyte trafficking, antigen presentation and T cell activation and leukocyte effector function. In a series of papers (Miller, 2002; Miller, 2003; Miller, 2004a; Miller, 2004b, Aoshi, 2008), it was proposed that T cell receptor (TCR) repertoire scanning is a stochastic process (or agent based system) that occurs through the dynamic behavior of DC dendrites and robust T cell motility that mediates numerous random TCR-pMHC interactions. This hypothesis assumes that naïve T cell migration is relatively unconstrained in 3D and provides a foundation for agent based models (ABMs) of T cell activation kinetics and immune effector responses. Subsequently, Bajenoff et al. proposed that T cell migration is restricted to the fibroblastic reticular conduit (FRC) network. The authors propose that because the FRC is random in structure, it provides a substrate that guides T cell migration, while also generating the "random migration" pattern characteristic of naïve T cell in vivo (Bajenoff et al., 2006). This idea has been widely adopted as dogma by immunologist, yet there are several observations that contradict this hypothesis. Moreover, T cell motility analysis methods and computer simulations are often based on the assumption that T cells can move freely within the lymph node. Whether T cells are constrained to the FRC network or not is an important issue that has ramifications for understanding the cellular mechanisms that initiate the immune response.

T cell motility

Judy Cannon

The University of New Mexico Health Sciences Center

Naïve T cell move within lymph nodes (LNs) in order to maximize the likelihood of detecting pathogen infection. Dendritic cells carry antigenic material from pathogens in peripheral tissues like the lung that have been infected to LNs. T cells must interact with antigen-bearing DCs in LNs to initiate the T cell response, which is required for clearance of pathogenic infection and immunological memory. We are interested in quantitatively understanding the type of motility taken by T cells in LNs and how different types of search impacts T cell interaction with DCs. While many studies have demonstrated the functional outcome of T-DC interactions, there has been little work done to quantitate and define the precise localization of DCs. As target distribution has been shown in many biological searches to be a key factor in determining efficiency of search, we are quantitatively measuring the level of clustering, as well as total area and volume that is covered by DCs within the LN. To do this, we have imaged CD11c-YFP to label DCs in intact LNs and are measuring DC distribution. We find that DCs occupy greater volume and surface area than would be expected from the estimated DC cell size. We are also in the process of analyzing whether DCs might actively attract T cells to locations of greater DC density. We anticipate that more careful quantitation of DC targets in LNs will help to generate more accurate models of T-DC interactions.​

Chemotaxis of Cytotoxic T cells in the Skin

Ioana Niculescu3, Silvia Ariotti1, Joost Beltman2, Ton Schumacher1, Johannes Textor3 & Rob J. de Boer3

Netherlands Cancer Institute (Amsterdam)1, Leiden University2 & Utrecht University3

We quantify the migration of cytotoxic T lymphocytes (CTLs) within Herpes Simplex Virus-1 (HSV-1) infected epidermis in vivo. Activated T cells display a subtle distance-dependent chemotaxis towards clusters of infected cells, which is mediated by CXCR3 and its ligands. Although the chemotactic migration is weak, “bootstrapping the data” indicates that this behavior is crucial for efficient localization of CTL at the foci of infection (Ariotti et al, in revision). We develop a computational model that includes healthy epidermis, an infection focus, CTLs that realistically squeeze between epidermal cells, follow chemotactic gradients, and kill HSV-1 infected cells. We quantify the role of chemotactic entry into the epidermis, chemotaxis within the epidermis, and chemotactic sensitivity during synapses. Controlling the infection requires chemotaxis, but too strong chemotaxis can be detrimental because (1) CTL migrate too deep into the infection focus, allowing the infection to grow at its border, and (2) chemotaxis can break functional cytotoxic synapses (Niculescu et al, in preparation).

T Cell Migration: Biology, Analysis, and Clinical Relevance

Johannes Textor

Utrecht University

The organs of the immune system are unique in that they largely consist of constantly moving cells. My talk will focus on the motility of T cells, whose migration patterns has been described as random-walk-like. I will present published and modelling results showing how this kind of migration helps to find antigen, and how networks of fibroblastic reticular cells (FRCs) could act to speed up this process. Next I will introduce MotilityLab, a project in which we are building a toolkit designed to help both experimentalists and computational biologists analyze migration data. I will end by showing some data illustrating how understanding T cell migration could help optimize cancer immunotherapy.

T cell Motility and the Cure of HIV

Alan S. Perelson

Theoretical Biology & Biophysics Group
Los Alamos National Laboratory
Los Alamos, NM 87545

During chronic infection, CD8+ T cells frequently become exhausted and express PD-1 as a cell surface marker characteristic of exhaustion. Recently, Zinsenmeyer and Dustin et al (JEM 210:757 2013) showed that exhausted CD4+ and CD8+ T cells showed prolonger motility paralysis, which could be reversed by therapeutic blockade of PD-1 PD1L interactions. Further, Hosking and Whitten et al. (JI 191:4211 2013) showed that during acute LCMV infection CD8+ T cells became exhausted 18-24 hr after infection. Motivated by these findings, Jessica Conway and I developed a model of HIV infection and treatment that includes an effector cells response that can become exhausted. I will show how this model (Conway and Perelson PNAS 112:5467 2015) provides insights into the phenomenon of post-treatment control of HIV-infection in which some patients treated with suppressive antiviral therapy have been taken off of therapy and then spontaneously control HIV infection such that the amount of virus in the circulation is maintained undetectable by clinical assays for years.

Modeling T cell activation in the Lymph Node as a collective foraging problem

Melanie E. Moses

Department of Computer Science University of New Mexico

Cooperative foraging by ants provides a conceptual model for studying the search for antigen by a population of T cells. Ant foraging speed depends on ways that ants move, communicate, and react to the distribution of food in their environment. We investigate whether similar behaviors are used by T cells to search for antigen. Specifically, we describe the efficiency of T cell search and how it depends on T cell movement patterns. We observe the movement patterns of naïve T cells in lymph nodes using ex vivo 2-photon microscopy and describe the statistical distribution of those movements using maximum likelihood methods. We find that while T cells move with features of a Lévy walk, Brownian and Lévy walks are both poor descriptors of T cell motion. Instead, distribution fitting and efficiency simulations indicate that T cells in lymph nodes move using a correlated random walk with a heavy-tailed distribution of step lengths. We find that a lognormal distribution of step lengths, motion that is directionally persistent over short time scales, and heterogeneity in movement patterns among T cells all increase search efficiency. In contrast to Brownian motion and Lévy walks, the observed T cell pattern of motion balances the need for repeated dendritic cell contact and discovery of rare dendritic cells bearing cognate antigen. We also observe specific locations, which we term hotspots, that are visited by T cells more frequently than can be explained by random movement. We show that T cells move differently in hotspots compared with non-hotspots, suggesting that T cells may adapt their movement in response to features of the lymph node environment.

Statistical Mechanical Techniques for Reaction Diffusion: Dynamics of Tethered Random Walkers

V.M. (Nitant) Kenkre

Department of Physics University of New Mexico

Reaction-Diffusion is a term that describes phenomena in which entities move (loosely described by ‘diffusion’) and when, as a result of the motion, they arrive under the influence of either one another or other entities, undergo some noticeable phenomenon (represented by the term ‘reaction’). Examples abound, e.g., in photosynthesis, sensitized luminescence, exciton annihilation, receptor-coalescence, infection spread, foraging, and target-finding. Suited perfectly to the practices of statistical mechanics, the subject has been tackled over decades in multiple ways in physics. In illustrating work I have done in this field since 1980 in the first five of the areas mentioned, I will touch upon our very recent analysis of the spread of infection. This is work done in collaboration with my students, primarily with Satomi Sugaya and some with Matt Chase, both of whom are participating in this workshop. I will discuss techniques and show an interesting counterintuitive result that emerges. The specific context is the dynamics of tethered random walkers for the spread of infection in epidemics. Of the several publications we have prepared or are preparing, Bulletin of Mathematical Biology 76, 3016-3027, (2014) is most relevant. I think application to the subject of our workshop by modifying that formalism could be straightforward and a lot of fun.