Extended Data Fig. 2: Pulmonary aspartate increases lung metastasis aggressiveness via NMDA receptor activity.

a. Hypusine levels in 4T1 (top) and MCF10A HRAS^V12 (bottom) spheroids grown in lung-like medium (LLM) supplemented with or without aspartate. A representative image of n = 3 independent experiments is shown. Quantification of hypusine signal normalized over total eIF5A signal is indicated on top of each lane. b. Total spheroid areas for 4T1 (left) and MCF10A HRAS^V12 (right) cells grown in LLM supplemented with or without aspartate (n = 14 4T1 no aspartate, n = 13 MCF10A HRAS^V12 no aspartate, n = 12 4T1 aspartate, n = 8 MCF10A HRAS^V12 aspartate). Data are presented as mean ± s.d. (n indicates independent samples). Unpaired two-tailed t-test with Welch correction. Representative images are shown on the right. Scale bar = 250 μm. c. Total spheroid areas for HUH7 (left) and B16F10 (right) cells grown in LLM supplemented with or without aspartate. N = 11 HUH7 no aspartate, n = 9 B16F10 no aspartate, n = 10 HUH7 aspartate, n = 9 B16F10 aspartate. Data are presented as mean ± s.d. (n indicates independent samples). Unpaired two-tailed t-test with Welch correction. Representative images are shown on the right. Scale bar = 250 μm. d. Hypusine levels in HUH7 (left) and B16F10 (right) spheroids grown in LLM supplemented with or without aspartate. A representative image of n = 3 independent experiments is shown. Quantification of hypusine signal normalized over total eIF5A signal is indicated on top of each lane. e. Top: SUnSET assay showing puromycin incorporation in 4T1 and MCF10A HRAS^V12 spheroids grown in LLM supplemented with or without aspartate. Spheroids were treated with cycloheximide (CHX, 100 nM) or vehicle, followed by puromycin (10 μg/ml). Bottom: Ponceau red staining showing equal protein loading. f. Left: GSEA normalized enrichment scores (NES) for the top 15 upregulated gene sets based on total RNA sequencing data for 4T1 spheroids grown in LLM supplemented with aspartate vs no aspartate. Dot colors and areas indicate FDR-adjusted P-values and gene-set sizes, respectively. Gene sets related to translation are highlighted in bold. Right: Volcano plot based on differential expression analysis of total RNA sequencing data for 4T1 spheroids grown in LLM supplemented with or without aspartate. Genes in Reactome’s translation gene set are highlighted in blue. Log₂ Fold Changes (FC) and negative log₁₀-transformed P-values are indicated in the x and y axes respectively. Both axes were further subject to inverse hyperbolic sine transformation for improved visualization. The horizontal dashed lines represent (from bottom to top) raw and BH-adjusted P-values of 0.05, while the vertical ones represent absolute fold-changes of 1.5. P-values based on fgsea’s adaptive multilevel splitting Monte Carlo approach, subject to FDR adjustment using the Benjamini-Hochberg (BH) approach. g. Hypusine levels in 4T1 (top) and MCF10A HRAS^V12 (bottom) spheroids silenced for eIf5a1, eIf5a2, or scramble shRNA, grown in LLM supplemented with or without aspartate. A representative image of n = 3 independent experiments is shown. Quantification of hypusine signal normalized over actin signal is indicated on top of each lane. h. Total spheroid areas for 4T1 cells silenced for eIf5a1 (n = 8 4T1 no aspartate, n = 8 4T1 aspartate), eIf5a2 (n = 10 4T1 no aspartate, n = 10 4T1 aspartate) or scramble shRNA (n = 9 4T1 no aspartate, n =  10 4T1 aspartate), grown in LLM supplemented with or without aspartate. Data are presented as mean ± s.d. (n indicates independent samples). Two-way ANOVA with Tukey’s multiple-comparison tests. i. Total spheroid areas for MCF10A HRAS^V12 cells silenced for eIF5a1 (n = 12 MCF10A HRAS^V12 no aspartate, n = 11 MCF10A HRAS^V12 aspartate), eIF5a2 (n = 11 MCF10A HRAS^V12 no aspartate, n = 10 MCF10A HRAS^V12 aspartate) or scramble shRNA (n = 10 MCF10A HRAS^V12 no aspartate, n = 11 MCF10A HRAS^V12 aspartate), grown in LLM supplemented with or without aspartate. Data are presented as mean ± s.d. (n indicates independent samples). Two-way ANOVA with Tukey’s multiple-comparison tests. j. ¹³C₄-labeled fractions for various metabolites downstream of aspartate (included) in 4T1 spheroids (n = 3) grown in LLM supplemented with ¹³C₄-aspartate. Data are presented as mean ± s.d. A representative graph of n = 3 experiments is shown. k. Average fractions of total carbon corresponding to ¹³C isotopes in different metabolites, in 4T1 spheroids (n = 3) grown in LLM supplemented with ¹³C₄-aspartate. l. Histograms of CD44 surface expression in 4T1 spheroids dissociated with Trypsin (n = 3) or Accutase (n = 3). Data are normalized to mode (y axis is scaled so that the maximum of each curve is at 100) and further smoothed for display purposes. m. Intracellular vs cell surface levels of ¹³C-glutamate in 4T1 spheroids silenced for Grin2d (n = 3) or scramble shRNA (n = 3), grown in LLM supplemented with ¹³C₅-glutamate. Data are presented as mean ± s.d. (n indicates independent samples). Two-way ANOVA with Tukey’s multiple-comparison tests. n. Total spheroid areas for 4T1 cells grown in LLM supplemented with or without L- or D-aspartate. N = 15 no aspartate, n = 16 L-aspartate, n = 13 D-aspartate. Data are presented as mean ± s.d. (n indicates independent samples). One-way ANOVA (P < 0.001) with Tukey’s multiple-comparison tests. Representative images are shown on the bottom. Scale bar = 250 μm. o. Hypusine levels in 4T1 spheroids grown in LLM supplemented with or without L- or D-aspartate. A representative image of n = 3 experiments is shown. Quantification of hypusine signal normalized over eIF5A signal is indicated on top of each lane. p. Hypusine levels in 4T1 spheroids grown in LLM supplemented with or without aspartate or glutamate. A representative image of n = 3 experiments is shown. Quantification of hypusine signal normalized over eIF5A signal is indicated on top of each lane. q. Grin2d detected in lung metastases 14 days after i.v. injections with 4T1 cancer cells, in mice pre-treated with daily injections of aspartate (20 mM, i.p., 10 days) or PBS, assessed by multiplex immunohistochemistry. Representative images from n = 5 4T1 shScr PBS, n = 5 4T1 shScr aspartate, n = 5 4T1 shGrin2d PBS, n = 5 4T1 shGrin2d aspartate independent mice are shown. Green = membrane marker ATP1A1; Red = Grin2d; Blue = DAPI nuclear staining. Scale bars = 25 μm. Arrows indicate membrane colocalization of Grin2d and ATP1A1. This staining is part of a multiplex staining and only relevant stains are shown. r. Relative fractions of Grin2b and Grin2d copies quantified by droplet digital PCR in 4T1 lung metastases, n = 3. Data are presented as mean ± s.d. (n indicates independent samples). Unpaired two-tailed t-test with Welch correction. s. Average ionomycin-normalized calcium response traces in 4T1 cells overexpressing Grin2b. The shaded ribbon represents standard error of the mean at each time point, while the grey rectangles indicate sequential addition of aspartate, glutamate and ionomycin at the times indicated by the arrows. A representative experiment (n = 38 cells) out of 2 independent experiments is shown. Grin2b overexpression stimulated a calcium response following glutamate but not aspartate addition in 4T1 cells. This shows that the NMDA receptor subunit expression determines the agonist to which breast cancer cells respond. t. Ionomycin-normalized calcium response upon sequential addition of aspartate and glutamate (n = 79 cells). Unpaired two-tailed t-test with Welch correction.

Source Data