Back

Widespread discordance between mRNA expression, protein abundance and de novo lipogenesis activity in hepatocytes during the fed-starvation transition

Landgraf, A.; Okada, J.; Horton, M.; Liu, L.; Solomon, S.; Qiu, Y.; Kurland, I. J.; Sidoli, S.; Pessin, J. E.; Shinoda, K.

2025-04-20 bioinformatics
10.1101/2025.04.15.649020 bioRxiv
Show abstract

The mammalian liver plays a critical role in maintaining metabolic homeostasis during fasting and feeding. Liver function is further shaped by sex dimorphism and zonation of hepatocytes. To explore how these factors interact, we performed deep RNA-sequencing and label-free proteomics on periportal and pericentral hepatocytes isolated from male and female mice under fed and starved conditions. We developed a classification system to assess protein-mRNA relationship and found that gene products (mRNA or protein) for most zonation markers showed strong concordance between mRNA and protein. Although classical growth hormone regulated sex-biased gene products also exhibited concordance, [~]60% of sex-biased gene products showed protein-level enrichment without corresponding mRNA differences. In contrast, transition between feeding and starvation triggered widespread changes in mRNA expression without significantly affecting protein levels. In particular, key lipogenic mRNAs (e.g. Acly, Acaca, and Fasn) were dramatically induced by feeding, but their corresponding proteins (ACLY, ACC1, and FAS) showed little to no change even as functional de novo lipogenic activity increased [~]28-fold in the fed state. To facilitate further exploration of these findings, we developed Discorda (https://shinoda-lab.shinyapps.io/discorda/), a web database for interactive analysis. Our findings reinforce the principle that mRNA changes do not reliably predict corresponding protein levels (and vice versa), particularly in the context of sex and acute metabolic regulation of hepatocytes, and that de novo lipogenesis activity can be completely uncoupled from changes in protein expression.

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.