Motility signatures of native T cells within lung tumor microenvironment using the murine crystal ribcage

Introduction: Characterizing the motile behaviors of CD8+ cytotoxic T cells is of particular interest in the context of solid tumors, whereby increased infiltration of T cells is associated with better survival and therapeutic response. Thus, studying T cell motility in the tumor microenvironment contributes towards understanding the mechanisms of native immune response (Nature Communications 2021) and in improving responses of engineered T cells (i.e., CAR-T cells; Science 2022). Current methods for studying T cell motility and distribution rely heavily on simplified 2-D and 3-D in vitro models which do not represent the complexities of the tumor microenvironment, or on histological analysis, which only provides information at a snapshot in time and fails to capture dynamic processes.

To overcome these limitations, we choose to instead image native, actively migrating T cells responding to primary and metastatic tumors in the intact, functional murine lung. We do so by leveraging the crystal ribcage platform (Nature Methods 2023), a transparent ribcage that allows real-time imaging of the lung while maintaining the circulation and respiration functions of the lung. Using this platform, coupled with in vivo labeling of the mouse’s own T cells, we capture timelapses of T cell migration across regions of the tumor microenvironment including the tumor, healthy tissue, and the intermediate region between healthy and normal tissue termed the peritumor: a physically and immunologically unique and understudied region (Trends in Cancer, 2023). Our approach offers substantial advantages over existing methods by providing dynamic information on native T cell activity within the functional lung.

Materials and

Methods: Primary and metastatic cancer models are initiated via either intraperitoneal injection of urethane (primary), or tail vein injection of GFP-positive murine cancer cells (e.g., YUMM1.7, B16-F10; metastatic). All models are induced in a transgenic C57/BL6 mouse line with a tdTomato membrane reporter. Tumor models are allowed to establish for 5-15 days (metastatic) or 20+ weeks (primary) to ensure tumor formation. One hour prior to sacrifice, the mouse’s native CD8+ T cells are labeled in vivo by tail vein injection of anti-CD8 nanobodies conjugated to Alexa-647 fluorophores. Following nanobody incubation, the mouse is intubated via the trachea, and the lung-heart bloc is removed from the chest cavity and placed inside the crystal ribcage imaging device (Nature Methods 2023; Fig. 1A, Fig 1B). Air pressure/flow are precisely controlled via external ventilator. The lung is continuously ventilated in the crystal ribcage under a confocal fluorescence microscope while 4D timeseries images are captured, allowing for continuous imaging of the same region of interest. Analysis is performed using a combination of semi-automated methods in ImageJ software, as well as custom-written Matlab scripts. Cells are categorized as belonging to tumor interior (Tumor), the interface between tumor and healthy tissue (Peritumor), or healthy tissue (Healthy) (Fig. 1C).

Results, Conclusions, and Discussions: Using the crystal ribcage (Fig. 1A) we image tumors in the lung (Fig. 1B) at highly resolved (Fig. 1C) scales. Fluorescent nanobodies bind to the mouse’s own CD8+ T cells (Fig. 1C(ii)), capturing the distribution and behavior of T cells in the lung tumor microenvironment. Examining both metastatic (Fig. 1D) and primary (Fig. 1E) tumors, we track CD8+ T cell infiltration and motility via timelapse fluorescent confocal microscopy. The mean-squared displacement (MSD), an index for the motility signature of T cells, is tracked with time. Within the primary cancer (Fig. 1F), there is a notable difference depending on the region observed: peritumor cells exhibit a higher relative MSD than in healthy tissue, and both are higher than the comparatively stationary cells within the tumor core. This location-specific MSD trend is not as strong within the metastatic tumor model (Fig. 1G): cells exhibit similar patterns in MSD across all 3 locations, though intratumor and peritumor T cells switch to re-localized motion at larger timescales.

These differences in CD8+ T cell motility may be due to the distinct growth patterns of the tumors: the nodular primary tumor rearranges surrounding tissue and may prevent effective CD8+ T cell infiltration, while the infiltrative, co-optive growth pattern of the metastatic tumor leaves native structures in place that may allow easier CD8+ T cell entry and activity. Indeed, in the metastatic model we are more likely to observe higher infiltration and direct damage to cancer cells (Fig. 1H).

CD8+ T cells play a key role in anti-tumor activity, and deciphering these different motility signatures may be key to improving cytotoxic T cell infiltration and killing activities in solid tumors in the lung, thereby improving patient response and outcome. Using the crystal ribcage, we image actively migrating CD8+ T cells within the intact tumor microenvironment in primary and metastatic tumors in the functional murine lung, examining location- and cancer-specific motility styles, and highlighting differences across cell location in the tumor core, peritumor, or healthy adjacent tissue. Our findings and method represent a valuable step towards characterizing and improving cytotoxic T cell behavior and response.

Acknowledgements (Optional):