Gracias!!

Thanks Shubhra!

Thanks!

12/ So thankful to everyone who made this possible — Maria for leading the project; the bioinformaticians Gerard Martínez-Cebrián and Lucía Lorenzi; and all our collaborators and funding agencies!🙏🙏🙏

11/ Indeed, p30-expressing cells were more sensitive to ER stress — revealing a potential vulnerability in CEBPA-mutant AML

10/ Surprisingly, co-IPs revealed that p30 fails to interact with ATF4, while p42 does — impairing downstream signaling. Given ATF4’s key role in the ER stress response, we tested how p30 cells respond to tunicamycin-induced stress

9/ We then turned to another AP-1 factor: ATF4, whose binding motifs were enriched in p30-downregulated enhancers. ATF4 itself wasn’t downregulated, but its binding sites were less active — why?

8/Zooming in, p30-expressing cells failed to activate interactions between the FOS gene and nearby regulatory regions.
Also, forced FOS expression rescued some of the downstream transcriptional programs and partially restored inflammatory fitness

7/ H3K27ac ChIP-seq revealed many differentially active enhancers, and these were enriched for AP-1 motifs.
We analyzed expression of the major AP-1 members and noticed that FOS and its paralogs were consistently downregulated in all our experimental models

6/ Still, p30 binds chromatin and is transcriptionally active — so why the defective inflammatory response?
ChIP-seq of both isoforms showed similar binding on inflammatory genes. Was there something else?

5/This had a protective effect: since prolonged inflammation can be detrimental to hematopoietic progenitors, expressing p30 instead of p42 conferred resistance to inflammatory stress

4/How does this affect the inflammatory response? Upon LPS stimulation, p30 cells failed to robustly activate early response genes, though late responses were less affected.

3/ Was this just a mouse-specific effect? Not at all. RNA-seq from CEBPA-mutant AML patients confirmed the same: a marked reduction in inflammatory gene expression.
And when we reintroduced p42 into mutant AML cells, it partially rescued inflammatory gene expression.

2/Using undifferentiated and differentiated mouse primary cells and an inducible HSC line, we found that p30 cells showed a strong downregulation of inflammatory genes compared to p42 cells — and this wasn’t due to differences in maturation.

1/Mutations in CEBPA, a key myeloid transcription factor, are common in AML. Very often these mutations result in the expression of a shorter isoform (p30) instead of the full-length p42.
But what are the transcriptional consequences of expressing p30 instead of p42?

Congrats!!