Back

FMRP prevents Me31B/DDX6-associated repression of long neurodevelopmental mRNAs

Ma, K.; Hossain, A. R.; Judson, K. L.; Cheng, T. Q.; Greenblatt, E. J.

2026-07-24 molecular biology
10.64898/2026.07.23.739932 bioRxiv
Show abstract

Long, dosage-sensitive mRNAs encoding neurodevelopmental regulators are selectively dependent on fragile X messenger ribonucleoprotein (FMRP) for efficient translation across animal systems, but the mechanism underlying this length-dependent requirement remains unclear. Here, we show that FMRP maintains translation of long target mRNAs by preventing their inappropriate sequestration into a Me31B/DDX6-dependent P-body repression pathway. In Drosophila oocytes, loss of FMRP caused target mRNAs, but not bulk polyadenylated RNA, to accumulate in Me31B-marked P-bodies. Improved individual-nucleotide resolution crosslinking and immunoprecipitation (iiCLIP) revealed that FMRP and Me31B co-occupy many long coding sequences, suggesting that FMRP targets are intrinsically vulnerable to repressive Me31B-associated machinery. A screen targeting [~]120 candidate genes, followed by proteomic analysis, identified an FMRP-associated ribonucleoprotein (RNP) assembly containing the stress granule-linked proteins Rin/G3BP, Lig/UBAP2L, and Capr/Caprin1. These factors supported translation or localization of distinct FMRP target subsets. Finally, inhibition of Me31B-dependent repressive complex assembly restored translation of the majority of FMRP targets in FMRP-deficient oocytes, supporting a model in which translational failure results from excessive repression. Together, these findings reveal a cytoplasmic RNP assembly that safeguards mRNAs by antagonizing promiscuous P-body repression and provide a mechanistic explanation for the selective vulnerability of long neurodevelopmental transcripts to FMRP loss.

Matching journals

The top 3 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.