(B-69) A data-driven approach to identify emergent behaviors of immune cells within germinal centers

Introduction:: The formation of germinal centers in secondary lymphoid tissues are critical for adaptive immune responses. Germinal centers arise within B-cell follicles and serve as the site for mature B-cells to generate antibodies with high-affinity binding to antigen. Emergent spatial organization of immune cells from the “dark zone” to the “light zone” of germinal centers drive this process. The dark zone contains dense subpopulations of B-cells undergoing rapid proliferation and somatic hypermutation to generate a diverse pool of antibodies. The light zone serves as a site for B-cell selection through dynamic cell-cell interactions with follicular dendritic and T helper cells. Depending on the affinity of these interactions, B-cells may either return to the dark zone for further rounds of somatic hypermutation, undergo apoptosis, or differentiate into plasma or memory B-cells that exit the germinal center. While prior research suggests that the spatial patterning of the germinal center is critical for its ability to facilitate B-cell affinity maturation, the dynamics by which the dark and light zones are established remain an open question. Indeed, the spatial patterning of the germinal center emerges from the individual movements and interactions between T- and B-cells. Differential migration patterns between these cell subpopulations could influence spatial confinements and interactions of these cells, giving rise to functional patterns on the multi-cell level. Here, we analyze intravital microscopy videos of the germinal center light zone in immunized mice (day 10) to derive cell type-specific motility patterns that suggest how differential migration drives self-assembly within the germinal center.


Materials and Methods:: Using two-photon microscopy, three-dimensional (3D) intravital microscopy movies were acquired from cells within the popliteal lymph nodes of previously immunized mice. 3D trajectories were extracted from these movies using Imaris software. Tracked cells include fluorescently labelled CD4 T-cells (TFH) and wildtype, germinal center B-cells (GC B). Each cell was tracked for 10 minutes at a frequency of 30 seconds. A total of 7,890 trajectories were analyzed. For each trajectory, statistical features (average, standard deviation, skewness, and kurtosis) of the displacements, speeds, and angle distributions were extracted. Additionally, each trajectory was rotated along its primary axis of migration to minimize effects on rotational variations in movements. The mean squared displacements (MSD) as a function of time lags (ranging from 30 seconds to 10 minutes) were computed along each axis of migration, and the statistical features of this function were recorded. These motility parameters were then clustered using unsupervised k-means clustering to define eight emergent patterns of cell movements.


Results, Conclusions, and Discussions:: For ease of interpretation, we visualized the motility features quantified for each trajectory in a dimensionally reduced uniform manifold approximation (UMAP) space. To understand which motility features captured the variance across trajectories, we projected the magnitude of individual motility features onto the UMAP (Fig.1A). We found that the MSD was highest at the top left of the UMAP. Simultaneously, we found that the frequency and number of peaks on the MSD is highest on the top right of the UMAP. Using k-means clustering, we identified eight distinct motility patterns (Fig.1B). Each pattern can be characterized with a unique combination of net displacement and direction. TFH and GC B cells show differences in their signatures of these motility patterns (Fig.1C). TFH cells tend to take longer trajectories whereas GC B cells travel shorter and less directed paths. Indeed, the variation in motility behavior between these cell types implies differences in their emergent ability to interact with each other and drive patterning in the germinal center. We hypothesize that TFH cells take on longer trajectories to survey a greater space in the germinal center and their degree of tortuosity may be influenced by interactions with GC B cells. Conversely, GC B cells may take on shorter trajectories with greater tortuosity to maximize interaction but conserve energy for mutation and proliferation. Our identification of differential migration patterns between TFH and GC B cells form the foundation for understanding how dynamic behaviors of these cells could lead to functional interactions that drive self-organizing properties of the germinal center.


Acknowledgements (Optional): :

References (Optional): :