Back

RNA methylation controls stress granule formation and function during erythropoiesis

GUNAGE, R. D.; Lin, S.; Wiley, D.; Choudhuri, A.; Smith, M.; Han, T.; Shah, A.; Coyne, S.; Koczirka, K.; Zhi Ming Chen, K.; Yu, C.; Drake, K.; Yang, X.; Yang, S.; Zhou, Y.; Bauer, D.; Qian, Z.; Calo, E.; Gregory, R.; Zon, L.

2024-11-08 cell biology
10.1101/2024.11.07.622459 bioRxiv
Show abstract

Stress granules (SGs) are crucial in RNA regulation, affecting cell fate and function. SGs contain RNAs, some of which can be methylated. We studied m6A RNA modifications during the human CD34+ HSPCs (hCD34+) differentiating into erythroid cells and found that mRNAs encoding many erythroid-specific proteins had decreased methylation during differentiation. Increased levels ALKBH5 demethylase during erythropoiesis controls the levels of the 3UTR methylation of these mRNAs. hCD34+ carrying ALKBH5 mutations demonstrated a block in erythropoiesis, and mass-spectrometry studies of the mutant cells showed decreased levels of SG proteins, including the core granule protein ATXN2. ALKBH5 directly regulates the methylation of the mRNA of ATXN2. ATXN2 overexpression accelerated the erythroid differentiation of HSPCs and rescued the erythroid differentiation of ALKBH5 mutant cells. Very few SGs are found in normal human erythroid progenitors. SGs accumulated substantially in ALKBH5 mutant cells, and surprisingly overexpression of ATNX2 reduced SG numbers to normal. Polysome analysis demonstrated m6A-modified RNAs to be enriched in the pre-polysome fractions that were less translated. This work establishes a mechanism by which during stress, ATXN2 facilitates the release of SG-stored m6A-modified RNAs including erythroid-specific and SG-enriched RNAs that are loaded onto functional ribosomes, allowing better translation and accelerated erythroid differentiation during stress.

Matching journals

The top 4 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.