Multinuclear Hodgin and Reed-Sternberg cells are formed through cytokinesis failure and refusion
Multinucleation is a hallmark feature observed across many cancers, yet its biological implications remain poorly understood. In Hodgkin lymphoma (HL), the malignant post germinal-center B-cell derived Hodgkin and Reed-Sternberg (HRS) cells are often large and multinucleated. Despite their defining role in HL, much of the underlying biology of HRS-cells is unknown, including the mechanisms leading to multinucleation. Potential causes include defects in cytokinesis and/or fusion of mononuclear cells, but these processes have yet to be clarified. Moreover, the dynamics of cell cycle progression and transcriptional activity within individual nuclei of multinucleated HRS cells remain unexplored. In this project, we use HL cell lines combined with fluorescent reporters and live cell imaging to attempt to delineate the origin of multinuclear HRS-cells.
To track nuclei and chromatin dynamics in real time, we generated a HL cell line (L428) that expresses green fluorescent protein conjugated to the chromatin marker histone-2B (H2B-GFP). L428 is a well characterized HL cell line that harbors both mono- and multinuclear cells. Live cells were imaged for up to twenty-four hours. Time-to-mitosis was scored manually for mono- and multinuclear cells. To assess whether multinuclear cells could form by fusion of mononuclear cells, we cocultured cells expressing GFP or mCherry in a 1:1 ratio for twenty-eight days, analyzing them by FACS and microscopy. To enable unbiased and high-throughput readouts, we trained an AI object detection algorithm through human-in-the-loop training.
Live-cell imaging demonstrates that multinuclear HRS-cells are capable of entering mitosis, displaying striking multipolar spindle formation. Interestingly, these cells are able to resolve multispindle mitoses, raising questions about the mechanisms of DNA content allocation between daughter cells. The time-to-mitosis did not differ between multinuclear and mononuclear cells, suggesting that multinucleated HRS cells have self-propagating properties. We furthermore show that multinuclear cells form through incomplete cytokinesis of daughter cells, followed by refusion. In our coculture experiment, we did not detect any GFP/mCherry double-positive cells, confirming that multinuclear cells are not formed through fusion of mononuclear cells.
We show that multinuclear HRS-cells form through incomplete cytokinesis and refusion of daughter cells. We hypothesize that cytokinesis failure results from a defect in abscission, the last step of cytokinesis. One cause of abscission failure is the presence of DNA in the abscission plane as a result of DNA-damage. Next in this project, we will employ a fluorescent DNA-damage sensor containing the double-strand DNA break binding protein 53BP1 to assess the role of DNA-damage in the formation of multinuclear HRS-cells. Furthermore, we will investigate the consequences of multinucleation on cell cycle progression using the FUCCI4 system and transcriptional activity of individual nuclei in multinucleated cells using fluorescent antibodies for euchromatin, and RNA synthesis reporters.