An economic demand-based framework for prioritization strategies in response to transient amino acid limitations

The study addresses how cells prioritize restoring distinct amino acid pools when exogenous supply is transiently disrupted. While cells regulate growth through global metabolic programs and sense amino acid sufficiency (e.g., TORC1 as a demand sensor and Gcn4/ATF4 as a supply-restoring regulator), it remains unclear whether and how cells differentially prioritize individual amino acids that vary in metabolic origins, chemical properties, uses, and intracellular concentrations. Using yeast in glucose- and nitrogen-replete conditions, the authors ask whether restoration responses are uniform across amino acids or hierarchically organized based on supply costs and demand, and seek to define the principles governing this prioritization.

Prior work has framed metabolism as a cellular economy with coordination of supply and demand. TORC1 signaling reflects amino acid sufficiency and growth demands, while Gcn4/ATF4 mediates transcriptional programs to restore amino acid supply during starvation or increased demand. Most studies employ complete amino acid starvation, which treats all amino acids uniformly, leaving open whether cells prioritize specific amino acids under selective disruptions. Amino acids have diverse metabolic origins and routes, and glutamate/glutamine serve as central nitrogen donors via transamination, suggesting potential differences in restoration strategies. The study builds on these insights and the known regulatory logic of TORC1 and Gcn4 to investigate amino acid-specific prioritization.

• Organism and growth conditions: Prototrophic Saccharomyces cerevisiae (CEN.PK background) grown at 30°C in glucose- and ammonium-containing defined media. Rapamycin (200 nM) used as control for TORC1 inhibition. Amino acid dropout media formulated by adding all amino acids at 2 mM (Cys at 0.5 mM) except the specified dropout; tyrosine excluded due to poor solubility.
• Amino acid grouping: Amino acids organized into seven groups considering metabolic origins and chemical structure: PPP, aromatic (Phe, Tyr, Trp); sulfur-containing (Met, Cys); glycolytic/TCA, BCAA (Leu, Ile, Val); TCA, polar (Asp, Asn, Thr; plus Glu, Gln as α-KG derived); glycolytic, uncharged (Ala, Ser, Gly); glutamate-derived (Arg, Pro, Lys); PPP, polar, basic (His). Pathway mapping based on KEGG and known biosynthetic routes in glucose/nitrogen-replete conditions.
• Reporter of supply–demand mismatch: Gcn4-luciferase reporter used to monitor activation upon shifting from amino acid-replete (+AA) to minimal (-AA) or selective dropout media. Time-course sampling at 10–240 min established a maximal response window at 60–90 min; subsequent single-time-point assays performed at 75 min post-shift.
• Protein-level readouts: Western blots to quantify stabilized Gcn4 (sGcn4-T105A,T165A expressed in gcn4Δ) and native chromosomally HA-tagged Gcn4 after 75 min in each dropout. Direct Gcn4 transcriptional outputs quantified by qRT-PCR of known targets.
• Growth assays: Wild-type and gcn4Δ strains grown in complete or dropout media; OD600 tracked up to 11 h with comparisons at 6 h to assess growth sensitivity to each dropout.
• Biosynthetic supply cost calculations: A comprehensive relative costing scale devised for each amino acid (and grouped totals) in glucose/ammonium conditions. Costs incorporate net high-energy phosphate bonds (ATP) produced/consumed, NADH converted to ATP (1 NADH = 3 ATP), and NADPH consumption for reductive biosynthesis, plus costs of metabolic precursors and cofactors. Composite group costs summarized as relative scores; lower numbers indicate higher supply cost in the presented heat map, with higher score reflecting lower cost. Full details in Supplementary Appendix 1 and Tables S4–S5.
• Demand estimation: • Proteome/translation demand estimated using ribosome profiling datasets (GSE91068; GSE122039) to identify top 500 highly translated mRNAs and enriched GO categories (notably ribosomal proteins). Amino acid composition (e.g., Lys, Arg, Pro) quantified in these sets and, separately, in top 500 most abundant proteins from a whole-proteome dataset. • Metabolic demand estimated qualitatively by listing major metabolic outputs and steady-state cellular metabolite amounts (order-of-magnitude using literature and BioNumbers), ranking amino acids into high, moderate, low metabolic demand (Appendix 1; Table S6).
• Dissecting arginine demand components: In arginine auxotroph arg1Δ, established arginine add-back kinetics after -Arg exposure to define Gcn4 attenuation. Tested whether supplementing spermidine (a major polyamine) reduces Gcn4 levels to quantify contribution of polyamine synthesis to arginine demand.
• TORC1 activity assays: Sch9 phosphorylation status assessed via NTCB cleavage mobility shift in HA-tagged Sch9 as a readout of TORC1 activity after dropouts of Arg, Lys, Pro. qRT-PCR of ribosomal protein genes (e.g., RPL30, RPL32, RPS6A, RPS26A) provided an independent TORC1 output measure.
• Statistics and visualization: Mean ± SEM reported; two-tailed Student’s t-test for significance. Box-plot construction and GO enrichment performed with standard web tools as noted.

• Gcn4 activation kinetics: Gcn4-luciferase activity increased within ~15 min of shifting to -AA, peaking at 60–90 min, then declining as supply matched demand; 75 min chosen for comparative assays.
• Hierarchical responses across amino acid groups: All group dropouts induced Gcn4 activity above +AA, but to differing extents. Strong to very strong responses for dropouts of glutamate-derived (Arg/Pro/Lys), sulfur-containing (Met/Cys), glycolytic/TCA BCAA (Leu/Ile/Val), and PPP, polar, basic (His) in the initial reporter; lower responses for PPP, aromatic (Phe/Tyr/Trp), TCA, polar (Asp/Asn/Glu/Gln/Thr), and glycolytic, uncharged (Ala/Ser/Gly). Protein-level analyses refined this: sGcn4 increased strongly for sulfur, glycolytic/TCA BCAA, glutamate-derived, and PPP aromatic dropouts, with minimal sGcn4 increase for histidine; native Gcn4 was highest in glutamate-derived dropout. Consolidated readouts (reporter, sGcn4, native Gcn4, Gcn4 target transcripts) defined the strongest overall response to glutamate-derived dropouts, followed by sulfur, glycolytic/TCA, and BCAA; moderate for PPP aromatic; minimal for glycolytic uncharged and TCA polar.
• Growth sensitivity: gcn4Δ cells showed the greatest short-term growth reduction in the glutamate-derived dropout compared to WT, indicating highest reliance on Gcn4-driven restoration for this group.
• Supply cost analysis: Composite per-molecule biosynthetic costs (glucose/ammonium environment) indicated glutamate-derived amino acids have low supply costs at the group level. Within this group, individual costs differed substantially: arginine lowest, lysine intermediate, proline highest.
• Demand analysis: Translational/proteome demand showed lysine and arginine enriched in highly translated ribosomal proteins (approx. Lys ~11%, Arg ~8.5%, Pro ~3.7% in enriched sets). Metabolic demand estimation ranked arginine as high due to its role as the largest cellular nitrogen assimilator and as a precursor for abundant polyamines (millimolar-scale), whereas lysine had low metabolic demand and proline had low demand overall under these conditions.
• Arginine-specific dominance (law of demand): Among glutamate-derived amino acids, arginine dropout elicited the highest Gcn4 protein and Gcn4 target transcriptional responses (comparable to -AA), and the strongest growth defect in gcn4Δ. In arg1Δ cells, arginine add-back for 30–60 min reduced Gcn4 to near-basal; spermidine supplementation reduced Gcn4 by ~40% relative to -Arg, indicating roughly half of arginine demand derives from polyamine (metabolic) needs.
• TORC1 response: TORC1 activity (Sch9 phosphorylation and ribosomal transcript levels) decreased most upon arginine limitation compared with lysine or proline limitation, aligning with the high-demand/low-cost profile of arginine.
• Conceptual framework: A four-quadrant model relates per-molecule supply cost to total demand. Highest restoration responses occur for amino acids with low supply cost but high demand (e.g., arginine). No amino acids populate a high-cost/high-demand quadrant under tested conditions.

The findings show that when amino acid supply is transiently disrupted, cells prioritize restoration according to a demand-driven logic: amino acids with high total demand and relatively low per-molecule supply costs elicit the strongest restorative responses. Multiple orthogonal readouts (Gcn4 reporter and protein, Gcn4 target transcripts, growth dependence on Gcn4, and TORC1 outputs) converge on glutamate-derived amino acids—especially arginine—as the top priority. Dissecting arginine demand demonstrates a substantial metabolic component (polyamine biosynthesis), confirming that metabolic demands can dominate over translational needs in driving restoration. This supports a cellular ‘law of demand’ framework where fluxes prioritize entities with high demand and lower supply prices, given constant supply parameters in a defined nutrient environment. The model explains minimal responses for amino acids with low demand despite low costs (e.g., some glycolysis-derived) and moderate responses where demand or cost is intermediate. It provides a tractable basis to predict context-dependent prioritization as nutrient environments and proteome allocations vary, with implications for identifying amino acid-specific sensing and restoration mechanisms and guiding metabolic engineering strategies.

This work establishes an economic, demand-based framework for how cells prioritize amino acid restoration after transient supply disruptions. By integrating comprehensive per-molecule biosynthetic supply costs with composite demand estimates (translation plus metabolic uses), the study reveals a hierarchical prioritization, led by glutamate-derived amino acids and especially arginine, whose high demand and low supply cost drive the strongest responses. The framework explains TORC1 and Gcn4 behaviors and accounts for growth outcomes upon selective limitation. It provides actionable insights for predicting amino acid reserves, identifying sensing/restoration circuitry, and designing metabolic engineering interventions. Future work should refine quantitative demand estimates (especially metabolic allocations), extend the framework across organisms and nutrient regimes, and dissect amino acid-specific sensing mechanisms in more complex, multi-supply systems.

• Demand estimates, particularly for metabolic allocations, are qualitative/order-of-magnitude and rely on assumptions and literature-derived abundance ranges; precise quantification remains challenging.
• The costing framework is context-dependent (glucose as carbon source, ammonium as nitrogen source) and would change with different nutrient states and active pathways.
• Experiments are in yeast and in short-term, batch conditions; generalizability to other organisms or complex tissues with multiple supply routes may be limited.
• Some assay constraints exist: tyrosine excluded from media due to solubility; Gcn4 regulation involves both translation and protein stability, complicating direct comparison across readouts; growth effects assessed over short time windows.
• No amino acids observed in a high-cost/high-demand quadrant under these conditions; different environments may alter quadrant occupancy and prioritization.