br Results br Discussion The origin and
Discussion The origin and fate of Tfh cells has been intensely studied since their first description 14 years ago (Breitfeld et al., 2000, Schaerli et al., 2000). Although mice engineered to report BCL6 (Kitano et al., 2011, Liu et al., 2012) and interleukin-21 (IL-21) (Lüthje et al., 2012) expression have provided powerful tools to analyze Tfh cells, their usefulness has been limited by Tfh cell heterogeneity and plasticity (Cannons et al., 2013). In this regard, the development of methods for the optical marking and tracking of cells based on their microanatomical location have created further opportunities for more precise delineation of Tfh cell dynamics and the molecular cues that underpin their behavior. Here we have used optical marking by TPP to link Tfh cell location to their behavior, phenotype, and gene expression. Our studies show remarkable differences in the migration pattern and single cell gene-expression signatures between primary and secondary Tfh cells. In addition, we report a subpopulation of “follicular memory T cells” that reside in the follicle where they scan SCS macrophages to initiate the secondary immune response upon antigen re-exposure. This temporospatial dissection of Tfh cell dynamics offers multiple new insights into regulation of GC responses in naive and antigen-experienced animals. Imaging of primary Tfh cells at the peak of the GC response revealed clear spatial segregation in the FM and GC compartments. This confinement was confirmed by TPP and discontinuous cell tracking 24 hr later, which showed retention of the majority of photoconverted GC Tfh cells in the original GC and follicle. Furthermore, NMF analysis of single cell gene expression signatures of FM and GC Tfh cells support the notion that they represent molecularly distinct cell populations. Thus, we conclude that the primary GC is a closed structure designed to partition responding GC b ng australia and restrict their access to CD4+ T cell help. At face value, these data contrasts with the findings of Shulman et al. who concluded that the GC is an open structure designed to broaden the diversity of the available CD4+ T cell help (Shulman et al., 2013). However, the preliminary experiments in their paper only examined polychromatic responses in naive animals that demonstrated initial colonization by multiple clones of red, green, or cyan T cells with the same TCR specificity and not interfollicular exchange as claimed. Furthermore, their subsequent experiments involved prime-boost immunization protocols that involved repeated exposure to antigen. This is a critical point of difference as they do not show any equivalent photoactivation data from naive responses (Shulman et al., 2013). Hence, their data is more consistent with our memory responses. In fact, we not only observed the migration of Tfh cells out of the GC in the secondary response but also their transport in the lymphatic flow of the SCS. This passive transport mechanism whereby cells “surf” the lymph appears to be an efficient and rapid mechanism for dissemination of cells that bypasses the need to traverse multiple anatomical compartments across disparate chemokine gradients. Regardless, it will be interesting to determine what role factors enriched in lymph, such as S1P, play in driving secondary Tfh cells to enter and leave the follicle. Why are Tfh cell dynamics so fundamentally different in naive and immune animals? Initially, GC B cells must pass stringent affinity and specificity checkpoints to ensure only high-affinity non-self-reactive cells are selected. Therefore, restricting primary Tfh cells with the greatest helper capacity to the GC might serve to direct help to cognate B cells and avoid the activation of bystander B cells in the follicle. Accordingly, expression of the genes encoding the B helper cytokines IL-21 and IL-4 were restricted to primary GC Tfh cells. In the secondary response, memory B cells have already passed these checkpoints and therefore have less stringent activation thresholds. Correspondingly, secondary Tfh do not express the same high amounts of Il21 and Il4 transcripts as primary GC Tfh cells. Furthermore, memory B cells are widely distributed in persistent GC remnants within the follicle (Dogan et al., 2009, Talay et al., 2012) and extrafollicular sites including the bone marrow (Dogan et al., 2009, Paramithiotis and Cooper, 1997), tonsillar mucosal epithelium (Liu et al., 1995) and splenic marginal zone (Liu et al., 1988). Therefore, protective secondary antibody responses may depend on the rapid extrafollicular export of secondary Tfh cells to these sites. Thus, unlike primary responses where it takes 7 days or more for Tfh cells to mature, the stereotypic expansion of CXCR5hiPD-1hi cell with a “mature” Tfh cell phenotype that peaks by day 3 in our system may be a part of a pre-wired memory program (Hale et al., 2013). Nevertheless, some memory B cells do enter GCs (Dogan et al., 2009, Pape et al., 2011) and Tfh cells are still required in this location in the secondary response. In this respect, it is notable that the NMF analysis of single cell gene expression by secondary Tfh cells showed that there was a hidden subpopulation of cells with high expression of Bcl6 and Pdcd1 that might be destined to later colonize and persist in secondary GCs. Thus, follicular memory T cells also appear to bifurcate into two responding populations upon rechallenge. However, at the peak of the secondary Tfh response these responding cells are equally likely to be in the FM or GC.