Back

ULK1 drives NDP52-mediated selective autophagic degradation of MHC-I to promote immune evasion in HPV-positive head and neck cancer

Vu, L.; Giacobbi, N. S.; Khalil, M. I.; Yang, C.; Eckerman, W. J.; Gomez Recinos, E.; Garber, J. D.; Son, H.; Chahal, P.; Villa, D. M.; Srivastava, T.; Bennett, A. Z.; Martin, K. R.; Welbon, C.; Williamson, C.; Spanos, W. C.; MacKeigan, J. P.; Olive, A. J.; Pyeon, D.

2026-03-18 cancer biology
10.64898/2026.03.14.711071 bioRxiv
Show abstract

Antigen presentation by major histocompatibility complex class I (MHC-I) is critical for tumor cell killing by CD8+ T cells. Accordingly, tumor cells downregulate MHC-I expression through multiple mechanisms, thereby evading the immune response. Importantly, lower levels of MHC-I are associated with poor responses to immune checkpoint inhibitor therapy. Our recent study has shown that the human papillomavirus (HPV) oncoproteins induce MHC-I protein ubiquitination by membrane-associated Ring-CH-type finger 8 (MARCHF8) in HPV-positive head and neck cancer (HPV+ HNC). However, the mechanism by which ubiquitinated MHC-I is degraded remains elusive. By performing genome-wide CRISPR screens, we identified components of the ULK1 and PIK3C3 complexes for autophagy initiation complexes among the top negative regulators of cell-surface MHC-I expression in HPV+ HNC cells. We show that MHC-I is recruited from the ER to autophagosomes by the cargo receptor NDP52, decreasing MHC-I levels. Further, inhibiting the initiation or nucleation steps of autophagy before autophagosome formation is critical for restoring MHC-I levels on the cell surface. Finally, genetic inhibition of autophagy initiation suppresses HPV+ HNC tumor growth in vivo and enhances the CD8+ T cell-mediated antitumor response. Our findings suggest that autophagic degradation of MARCHF8-ubiquitinated MHC-I is a key immune evasion mechanism in HPV+ HNC.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.