Metabolic profiling stratifies colorectal cancer and reveals adenosylhomocysteinase as a therapeutic target.

The genomic landscape of colorectal cancer (CRC) is shaped by inactivating mutations in tumour suppressors such as APC, and oncogenic mutations such as mutant KRAS. Here we used genetically engineered mouse models, and multimodal mass spectrometry-based metabolomics to study the impact of common genetic drivers of CRC on the metabolic landscape of the intestine. We show that untargeted metabolic profiling can be applied to stratify intestinal tissues according to underlying genetic alterations, and use mass spectrometry imaging to identify tumour, stromal and normal adjacent tissues. By identifying ions that drive variation between normal and transformed tissues, we found dysregulation of the methionine cycle to be a hallmark of APC-deficient CRC. Loss of Apc in the mouse intestine was found to be sufficient to drive expression of one of its enzymes, adenosylhomocysteinase (AHCY), which was also found to be transcriptionally upregulated in human CRC. Targeting of AHCY function impaired growth of APC-deficient organoids in vitro, and prevented the characteristic hyperproliferative/crypt progenitor phenotype driven by acute deletion of Apc in vivo, even in the context of mutant Kras. Finally, pharmacological inhibition of AHCY reduced intestinal tumour burden in ApcMin/+ mice indicating its potential as a metabolic drug target in CRC.

The amino acid transporter SLC7A5 is required for efficient growth of KRAS-mutant colorectal cancer.

Oncogenic KRAS mutations and inactivation of the APC tumor suppressor co-occur in colorectal cancer (CRC). Despite efforts to target mutant KRAS directly, most therapeutic approaches focus on downstream pathways, albeit with limited efficacy. Moreover, mutant KRAS alters the basal metabolism of cancer cells, increasing glutamine utilization to support proliferation. We show that concomitant mutation of Apc and Kras in the mouse intestinal epithelium profoundly rewires metabolism, increasing glutamine consumption. Furthermore, SLC7A5, a glutamine antiporter, is critical for colorectal tumorigenesis in models of both early- and late-stage metastatic disease. Mechanistically, SLC7A5 maintains intracellular amino acid levels following KRAS activation through transcriptional and metabolic reprogramming. This supports the increased demand for bulk protein synthesis that underpins the enhanced proliferation of KRAS-mutant cells. Moreover, targeting protein synthesis, via inhibition of the mTORC1 regulator, together with Slc7a5 deletion abrogates the growth of established Kras-mutant tumors. Together, these data suggest SLC7A5 as an attractive target for therapy-resistant KRAS-mutant CRC.

Implications of Peak Selection in the Interpretation of Unsupervised Mass Spectrometry Imaging Data Analyses.

Mass spectrometry imaging can produce large amounts of complex spectral and spatial data. Such data sets are often analyzed with unsupervised machine learning approaches, which aim at reducing their complexity and facilitating their interpretation. However, choices made during data processing can impact the overall interpretation of these analyses. This work investigates the impact of the choices made at the peak selection step, which often occurs early in the data processing pipeline. The discussion is done in terms of visualization and interpretation of the results of two commonly used unsupervised approaches: t-distributed stochastic neighbor embedding and k-means clustering, which differ in nature and complexity. Criteria considered for peak selection include those based on hypotheses (exemplified herein in the analysis of metabolic alterations in genetically engineered mouse models of human colorectal cancer), particular molecular classes, and ion intensity. The results suggest that the choices made at the peak selection step have a significant impact in the visual interpretation of the results of either dimensionality reduction or clustering techniques and consequently in any downstream analysis that relies on these. Of particular significance, the results of this work show that while using the most abundant ions can result in interesting structure-related segmentation patterns that correlate well with histological features, using a smaller number of ions specifically selected based on prior knowledge about the biochemistry of the tissues under investigation can result in an easier-to-interpret, potentially more valuable, hypothesis-confirming result. Findings presented will help researchers understand and better utilize unsupervised machine learning approaches to mine high-dimensionality data.