Using a functional model of breast cancer heterogeneity, we previously showed that clonal sub-populations proficient at generating circulating tumour cells were not all equally capable of forming metastases at secondary sites1. A combination of differential expression and focused in vitro and in vivo RNA interference screens revealed candidate drivers of metastasis that discriminated metastatic clones. Among these, asparagine synthetase expression in a patient’s primary tumour was most strongly correlated with later metastatic relapse. Here we show that asparagine bioavailability strongly influences metastatic potential. Limiting asparagine by knockdown of asparagine synthetase, treatment with l-asparaginase, or dietary asparagine restriction reduces metastasis without affecting growth of the primary tumour, whereas increased dietary asparagine or enforced asparagine synthetase expression promotes metastatic progression. Altering asparagine availability in vitro strongly influences invasive potential, which is correlated with an effect on proteins that promote the epithelial-to-mesenchymal transition. This provides at least one potential mechanism for how the bioavailability of a single amino acid could regulate metastatic progression.
Gene Expression Omnibus
This work was performed with assistance from CSHL Shared Resources, which are funded, in part, by the Cancer Center Support Grant 5P30CA045508. We thank M. Mosquera, M. Cahn, J. Coblentz, and L. Bianco for support with mouse work; K. Cheng, J. Bourbonniere, D. Hoppe, A. Nourjanova, and R. Puzis for support with histology; E. Hodges and E. Lee for support with next-generation sequencing; and J. Johnson for assistance with HPLC. This work was also performed with the assistance of the Cancer Research UK, Cambridge Institute Proteomics Core Facility. S.R.V.K. is supported by a fellowship from The Hope Funds for Cancer Research. E.W. is supported by a long-term fellowship from the Human Frontier Science Program. L.A.C. is supported by the Susan G. Komen Foundation (SAC110006) and the NCI Breast SPORE program (P50-CA58223-09A1). J.C.H. and C.M.P. are supported by funds from the NCI Breast SPORE program (P50-CA58223-09A1), the Breast Cancer Research Foundation, and the Triple Negative Breast Cancer Foundation. H.G. is supported by a grant from the National Institutes of Health (NIH) (NCI R00 CA194077). Work in the G.P. laboratory is supported by the Institute of Cancer Research, London and a Cancer Research UK grand challenge award (C59824/A25044). G.J.H. is the Royal Society Wolfson Research Professor and is supported by core funding from Cancer Research UK, by a Program Project grant from the NIH (5 P01 CA013106-44), and by a grant from the Department of Defense Breast Cancer Research Program (W81XWH-12-1-0300).
Extended data figures
Genes identified as over-expressed in 4T1-T. Genes identified as over-expressed in 4T1-T cells as compared to 4T1-E cells by differential expression analysis of RNA sequencing data. For each line, cells were grown in vitro and two biological replicates were processed for RNA sequencing. Genes were identified in 4T1-T using a fold-change cutoff of 2 and a DESeq FDR cut-off of 0.05.
Gene Ontology analysis of genes over-expressed in 4T1-T. The genes listed in Supplementary Table 1 were analysed to identify enriched cellular processes, functions and components. Each Gene Ontology term that was associated with the genes in the subset was compared in its number of associates within the list to its number of associations to the entire murine Refseq gene list via a hypergeometic test. Ontology terms that were enriched with a q-value < 0.05 are listed.
Raw RNAi screening data and shRNA depletion scores. Represented are the Illumina sequence reads that were assigned to each of ~6 shRNAs targeting protein coding members of the genes listed in Supplementary Table 1, both in the pre-injection infected cell population and in the cell populations that were removed from the lungs of mice in the in vivo screen or were removed from the matrigel invaded cells in the in vitro screen. Also listed are the log-fold enrichment and depletion scores of each shRNA and false-discovery rates as assigned using an Emperical-bayes moderated t-test.
Expression changes induced by shRNA silencing or cDNA induced over-expression. Relative expression values for each cell line produced using shRNAs for silencing or cDNAs for enforcing expression.
Amino acid composition of serum with and without ʟ-asparaginase treatment. 4T1-T cells harbouring the non-targeting Renilla shRNA were injected into immunocompromised mice. Five mice each were either injected with 60 U l -asparaginase or PBS 5 days per week. After blood collection and serum isolation, free amino acids were quantified using High Performance Liquid Chromatography (HPLC) and a fluorometric detector.
Mouse orthologues of human genes that were identified as differentially expressed during EMT. Listed are genes that are up- or down-regulated when cells were enforced for the expression of Tgf-ß, Twist, Gsc or Snail or when E-cadherin was silenced (EMT-up and –down genes, respectively). EMT-up genes whose protein-level log-fold changes in Asns-silenced cells fell within the bottom or top 10% are annotated as Down-regulated and Up-regulated, respectively.