How Do Antisense Oligonucleotides Enter Cancer Cells? CD44 and EPHA2 Reveal a KRAS-Targeted Delivery Pathway

Antisense oligonucleotides can be designed according to gene sequences and, in theory, can switch off cancer-driving signals. The real challenge, however, is often not “what to target,” but how to cross the cell membrane, avoid degradation, and ultimately reach the target RNA hidden inside the cell. A study targeting KRAS has now mapped the key stages of this delivery journey, offering new engineering clues for improving nucleic acid drugs for pancreatic cancer.

The research team found that the antisense oligonucleotide first interacts with the scavenger receptor CD44 on the cell surface, activating the ERK–RSK signaling axis and causing serine phosphorylation of the receptor tyrosine kinase EPHA2. This change helps the drug enter the cell through endocytosis and travel in vesicles to endosomes near the nucleus, rather than becoming aimlessly trapped in cytoplasmic vesicles.

Endosomes are both vehicles for transport and cages that the drug must escape. The study showed that these endosomes near the nucleus become leaky, allowing some antisense oligonucleotides to escape into the cytoplasm, where they reach and inhibit KRAS messenger RNA. In other words, CD44 and EPHA2 not only increase cellular uptake but may also determine where the drug is delivered and how much of it can leave the endosome in an active form.

The team also identified another lever in the cell’s damage-repair response. Cells use a mechanism associated with stress granules to repair leaky endosomes; when researchers inhibited this repair process, more antisense oligonucleotides were able to escape, and KRAS inhibition increased accordingly. This suggests that efficacy could be improved from two directions in the future: strengthening receptor-directed transport on one side, and temporarily modulating endosomal repair or escape efficiency on the other.

KRAS mutations are important drivers of pancreatic ductal adenocarcinoma, but different mutation types, tumor tissue barriers, and variations among cells may all affect how efficiently nucleic acid drugs reach tumors and enter cells. Even if cell experiments show that a pathway can improve target inhibition, it cannot be directly assumed that human tumors will exhibit the same drug distribution, efficacy, or safety range; intervening in ERK–RSK, EPHA2, or endosomal repair may also affect signaling and membrane stability in normal cells.

Currently available information identifies this work as a bioRxiv preprint that has not undergone peer review and provides no evidence from human trials that would allow its clinical benefit to be assessed. In addition to validating delivery and antitumor effects in animal tumor models, the next steps must clarify which patients’ tumors have sufficient CD44 and EPHA2 activity, and whether increasing endosomal leakage can improve drug efficacy while avoiding unintended toxicity.