Publication highlights

Previous work has found that, surprisingly, olfaction is a high-frequency sense: mice can discriminate odour fluctuations at 40 Hz or more. What could this high-frequency acuity be used for? By analyzing simulations of two-dimensional air flow containing multiple odour sources, researchers at the Crick show that high-frequency odour fluctuations contain more information about how far apart two odour sources are than low frequencies. This suggests that the high-frequency acuity helps mice build accurate olfactory maps of their environments, a sort of passive SONAR, but using smells instead of sound.

Researchers at the Crick have shown that over 40% of variation in kidney cancer behaviour is due to non-genetic factors. The team analysed the DNA and RNA of 243 samples from 79 people with kidney cancer in the TRACERx Renal study, to understand both genetic and transcriptional variation (when the genes are read in the cell and converted into proteins). They showed that over 40% of transcriptional variation could not be accounted for by major cancer mutations in the DNA. Instead, it was happening when the DNA was being read in the cells and converted into proteins. Their research identified four types of transcriptional variation which give tumours an advantage, which could be targeted by new treatments or help doctors understand the risk of a person’s cancer spreading.

Loss of the tumour suppressor gene CDKN2A is a common early event in development of the pre-cancerous condition Barrett's oesophagus. Around 1% of Barrett's patients go on to develop oesophageal adenocarcinoma, but rather than enhancing this progression, as would be expected, early CDKN2A loss is actually protective. Having made this striking observation, the team at the Crick and collaborators showed that the reason lies with a second tumour suppressor gene, TP53. Loss of TP53 is a key driver of transformation into oesophageal cancer, but if CDKN2A is also missing, the Barrett's cells are too weakened to progress. CDKN2A changes sides to become a villain later in the process: if it's lost after the cancer has developed, it promotes a more aggressive tumour.

Researchers at the Crick and the UCL Cancer Institute have identified a genetic change which happens early in lung cancer development, that makes cancer cells divide abnormally and become harder to treat. They studied non-small cell lung cancer samples from the Cancer Research UK-funded TRACERx study, to investigate which genetic changes make two hallmarks of cancer, chromosomal instability and whole genome doubling, more likely. They identified that a gene called FAT1 was mutated in lung cancer cells with unstable chromosomes before they doubled their genomes. Cells with a complete loss of FAT1 couldn’t divide properly to produce two new cells. When FAT1 and another gene involved in cell size regulation called YAP1 were removed, the cancer cells no longer doubled their genomes. This suggests that drugs that block YAP1 could be particularly effective against cells with high levels of chromosomal instability.

Nuclear Magnetic Resonance (NMR) spectroscopy helps scientists understand how proteins are structured and behave. While NMR commonly focuses on the backbone and methyl-bearing side chains of proteins, analysing aromatic side chains, often crucial for protein function, is more difficult. To overcome this, researchers at the Crick and UCL developed a deep learning tool named FID-Net-2. By combining innovative developments in biomolecular NMR with advanced deep learning, FID-Net-2 substantially enhances the quality and resolution of NMR data for aromatic side chains. This allows insights into the mechanism of protein dynamics like folding. This approach works across various protein sizes and promises to improve NMR analysis in structural biology.

Scientists at the Crick have generated human stem cell models which, for the first time, contain notochord – a tissue in the developing embryo that acts like a navigation system, directing cells where to build the spine and nervous system (the trunk). The team first analysed chicken embryos to understand exactly how the notochord forms naturally. By comparing this with existing published information from mouse and monkey embryos, they established the timing and sequence of the molecular signals needed to create notochord tissue. With this blueprint, they produced a precise sequence of chemical signals and used this to coax human stem cells into forming a notochord. The stem cells formed a miniature ‘trunk-like’ structure, which spontaneously elongated to 1-2 millimetres in length. The scientists believe this work could help to study birth defects affecting the spine and spinal cord.

The germinal centres (GCs) of the body act as factories where antibody-secreting B cells are fine-tuned to reach their highest antigen affinity. GC-B cells cycle between two GC zones, undergoing antigen-driven selection and initiating cell division in the light zone (LZ), before migrating to the dark zone (DZ), where they vigorously proliferate. Initiation of cell division in the LZ was a puzzle, as the low-oxygen conditions in the LZ normally induce cell cycle arrest. Researchers at the Crick showed that a microRNA called miR-155 metabolically reprogrammes LZ GC-B cells by regulating genes that enhance energy production and prevent cell death. This process is essential for effective immune function in the face of infection.

The NanoBiT Biochemical Assay was created to investigate protein-protein interactions in live mammalian cells. Soly Ismail, Scientific Programme Manager in the Oncogene Biology Laboratory, led by Julian Downward, further developed the assay so it only needed to use parts of a extracts from cells rather than the live cells themselves, allowing it to be scaled up to undertake many tests at once. These protein-protein interactions are often difficult to visualise but could be potential new drug targets. Soly used the assay to detect and block weak interactions between a cancer-causing protein called RAS and an enzyme called PI3kK. The identified compounds that bind with PI3kK will be followed up in further tests to understand the nature of these interactions and how to optimise these compounds for drug development. Soly was awarded the Sir David Cooksey Prize in Translation for this work.

The MCM helicase enzyme separates the two strands of the double helix, enabling DNA replication, with two copies of MCM (a “double hexamer”) marking where replication can start. We have a clear understanding of how double hexamers are loaded in yeast, but not in human cells. Researchers from the Crick reconstituted double hexamer loading with purified human proteins and determined the atomic structures. They found that, unlike in yeast, loading of a human double hexamer is sufficient to start opening DNA. Two alternate loading pathways exist to load human double hexamers. Loading factors that are essential in yeast only play a dispensable, regulatory role in the human system. Thus, while the double-hexamer loaders are conserved throughout evolution, how they function is different. This work begins to unravel how human cells regulate initiation of DNA replication, ensure that their genome is duplicated only once, and prevent chromosome instability and cancer.