¹H NMR metabolomics reveals increased glutaminolysis upon overexpression of NSD3s or Pdp3 in Saccharomyces cerevisiae

Abstract

NSD3s, the proline‐tryptophan‐tryptophan‐proline (PWWP) domain‐containing, short isoform of the human oncoprotein NSD3, displays high transforming properties. Overexpression of human NSD3s or the yeast protein Pdp3 in Saccharomyces cerevisiae induces similar metabolic changes, including increased growth rate and sensitivity to oxidative stress, accompanied by decreased oxygen consumption. Here, we set out to elucidate the biochemical pathways leading to the observed metabolic phenotype by analyzing the alterations in yeast metabolome in response to NSD3s or Pdp3 overexpression using 1H nuclear magnetic resonance (NMR) metabolomics. We observed an increase in aspartate and alanine, together with a decrease in arginine levels, on overexpression of NSD3s or Pdp3, suggesting an increase in the rate of glutaminolysis. In addition, certain metabolites, including glutamate, valine, and phosphocholine were either NSD3s or Pdp3 specific, indicating that additional metabolic pathways are adapted in a protein‐dependent manner. The observation that certain metabolic pathways are differentially regulated by NSD3s and Pdp3 suggests that, despite the structural similarity between their PWWP domains, the two proteins act by unique mechanisms and may recruit different downstream signaling complexes. This study establishes for the first time a functional link between the human oncoprotein NSD3s and cancer metabolic reprogramming.

NMR: Unique Strengths That Enhance Modern Metabolomics Research

“…In a study using genetically edited yeast to overexpress the human oncoprotein NSD3, Rona and co-workers used NMR to quantify small, but statistically significant, differences in valine, glutamate, and phosphocholine between NSD3 and the structurally similar overexpressed yeast Pdp3. 102 No significant differences were observed for aspartate and arginine, both with low concentrations, a finding that is consistent in light of the structural similarity of the two proteins. Yet, together with alanine, these metabolites were significantly different from the Saccharomyces cerevisiae wild-type strain and consistent with cancer phenotypes.…”

“…This large dynamic range allows for reliable quantitative measurements of metabolites over a wide range of concentrations. In a study using genetically edited yeast to overexpress the human oncoprotein NSD3, Rona and co-workers used NMR to quantify small, but statistically significant, differences in valine, glutamate, and phosphocholine between NSD3 and the structurally similar overexpressed yeast Pdp3 . No significant differences were observed for aspartate and arginine, both with low concentrations, a finding that is consistent in light of the structural similarity of the two proteins.…”

“…Yet, together with alanine, these metabolites were significantly different from the Saccharomyces cerevisiae wild-type strain and consistent with cancer phenotypes. This suggested that despite the near identical metabolic profile of these modified organisms, phosphocholine, valine, and glutamate indicate distinct pathways that ultimately recruit different downstream signaling complexes …”

“…They collected 2D NMR on 71 Escherichia coli cultures, each overexpressing a different gene, and were able to identify 17 metabolites that reflected the optimal conditions for protein expression. These metabolites, with changes between groups ranging from an order of magnitude to as low as 0.2 fold, were used to reduce the number of trial-and-error iterations necessary to achieve high protein yields and set indicators for external stresses that could modulate metabolism to an optimal protein production environment. ,, …”

“…This suggested that despite the near identical metabolic profile of these modified organisms, phosphocholine, valine, and glutamate indicate distinct pathways that ultimately recruit different downstream signaling complexes. 102 Microbial protein production systems rely heavily on overexpression methods to produce large quantities of proteins of interest. 103 These can be hard to optimize or troubleshoot when yields are lacking or no protein is produced.…”