Reviews & Analysis

We performed a proof-of-concept study showing that single-cell RNA sequencing, a method for capturing rich tumor information (not yet in clinics owing to high costs), can be used to identify patients likely to respond to targeted therapy and to monitor the emergence of resistance.

A CRISPR dropout screen for tRNA regulators identified YRDC as the top essential gene in glioblastoma stem cells. Threonine acts as a substrate of YRDC to facilitate the N⁶-threonylcarbamoyladenosine (t⁶A) tRNA modification and shift translation toward mitosis-related genes with a codon bias. Our findings support targeting glioblastoma growth by a well-tolerated dietary therapy.

Pancreatic cancer liver metastases show high levels of immunosuppressive macrophages, which promote tumor growth. We found that blocking the efferocytosis pathway during early-stage metastasis in a mouse model inhibited the development of immunosuppressive activity in monocyte-derived macrophages, which restored T cell activation and reduced the metastatic tumor burden.

Our study reveals that the LGR4–Wnt signaling pathway dictates both ferroptosis and stemness traits to confer drug resistance to tumor cells. We thus generated a monoclonal antibody against LGR4 that blocks LGR4–Wnt signaling and sensitizes chemotherapy-resistant colorectal cancer tumors via selective promotion of ferroptosis.

Using CRISPR–Cas9 screens, we found that cancer-cell-intrinsic loss of Pip4k2c conferred liver-metastatic organotropism in melanoma through hypersensitization to insulin-mediated PI3K–AKT signaling via co-optation of distinct hepatic metabolic cues. Additionally, we showed that combinatorial therapies that abolished physiological and drug-induced changes in glucose and insulin levels specifically reduced liver metastasis.

Mitochondrial DNA mutations are present in over 50% of all cancers, and truncating mutations in several genes encoding components of complex I of the respiratory chain are most recurrent. We found that these mutations are a source of Warburg-like metabolic shifts that promote a pro-inflammatory immunological state, enhancing sensitivity to checkpoint blockade.

Senescent cancer cells, which are characteristically present in tumors after genotoxic therapies, upregulate the immune checkpoint ligand programmed cell death 1 ligand 2 (PD-L2). We show that genetic or pharmacological ablation of PD-L2 prevents the accumulation of intratumoral senescent cells, reducing the recruitment of immunosuppressive myeloid cells and facilitating tumor clearance by T cells.

Imetelstat is a first-in-class telomerase inhibitor with efficacy in a number of blood cancers. Intriguingly, telomere lengths do not predict patient responses to imetelstat. We now show that imetelstat causes cell death by a mechanism that involves two regulators of fatty acid metabolism (FADS2 and ACSL4), driving excessive lipid reactive oxygen species formation and ferroptosis.

We applied an artificial intelligence (AI) approach to a dataset of clinical and advanced multi-omic molecular features from patients with pancreatic adenocarcinoma to predict survival. The results reveal a tumor-type-agnostic platform that can identify parsimonious and robust clinical prediction biomarkers, catalyzing the vision to democratize precision oncology worldwide.

Chromosomal instability (CIN) (a hallmark of human cancer) is caused by persistent errors in chromosome segregation during mitosis. Pharmacological inhibition of the mitotic kinesin KIF18A selectively exploits a mitotic vulnerability for which cancer models with CIN are enriched, which leads to robust anti-cancer effects and durable tumor regression in mice.

Microbiome diversity has been associated with improved outcomes after allogeneic stem-cell transplantation in patients with hematological cancers. Multimodal analysis of intestinal microbiome and metabolome data helped identify immunomodulatory microbial metabolites that were predictive of survival, transplant-related mortality and cancer relapse. These metabolites were products of short-chain-fatty-acid-synthesis pathways, and their associated genes were expressed by both bacterial species and bacteriophages.

Genomic features of de novo metastatic prostate cancer can clarify prognosis and direct therapy. Using multi-region profiling of synchronous primary and metastatic patient tissues, we reveal the complex evolutionary histories of this lethal disease and identify strategies to better capture the genomic features of dominant metastatic populations.

We developed a method for generating dendritic cell progenitors (DCPs) from hematopoietic stem and progenitor cells isolated from bone marrow or blood. When engineered to express IL-12 and FLT3L, these DCPs reprogram the tumor microenvironment and elicit anti-tumor immunity without the need for ex vivo antigen loading.