Back

A high-throughput zebrafish screen identifies novel candidate treatments for Kaposiform Lymphangiomatosis (KLA)

Bassi, I.; Jabali, A.; Farag, N.; Egozi, S.; Moshe, N.; Leichner, G. S.; Geva, P.; Levin, L.; Barzilai, A.; Avivi, C.; Long, J.; Otterstrom, J. J.; Paran, Y.; Barr, H.; Yaniv, K.; Greenberger, S.

2024-03-25 developmental biology
10.1101/2024.03.21.586124 bioRxiv
Show abstract

Kaposiform Lymphangiomatosis (KLA) is a rare, aggressive, and incurable disease caused by a somatic activating NRAS mutation (p.Q61R) in lymphatic endothelial cells (LECs). The development of new therapeutic avenues is hampered by the lack of animal models faithfully replicating the clinical manifestations of KLA. Here, we established a novel zebrafish model of KLA by driving conditional expression of the human NRAS mutation in venous and lymphatic ECs. We find that mutant embryos recapitulated clinical features of KLA, including pericardial edema and a dilated thoracic duct, and that the phenotypes were reverted by Trametinib, a MEK inhibitor used for KLA treatment. We further leverage this model in combination with an AI-based high-throughput drug screening platform to search for small compounds selectively reverting the mutant phenotypes and identify Cabozantinib, an FDA-approved tyrosine kinase inhibitor, and GSK690693, a competitive pan-Akt kinase inhibitor, as leading hits. Finally, we test these drugs in cultured cells derived from KLA patient and demonstrate their ability to normalize LEC sprouting and block NRAS downstream pathways, underscoring the potential of GSK690693 and Cabozantinib as potential KLA treatments. Overall, our novel zebrafish model provides a valuable tool for research into the etiology of KLA and for identifying new therapeutic avenues.

Matching journals

The top 6 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.