To test this, researchers from Dr. Qian Wenfeng’s lab at the Institute of Genetics and Developmental Biology (IGDB) of the Chinese Academy of Sciences studied Arabidopsis. Using a dynamic editing-based lineage tracing system, the researchers tracked cell trajectories. During division, the base editor AID (acting as a “writer”) continuously introduced heritable mutations into a synthetic DNA “readout” sequence, allowing lineage information to accumulate.

Through deep sequencing, the team identified somatic mutations in parental leaves and germline mutations transmitted to the progeny. These were used to reconstruct a cell lineage tree connecting somatic and germline cells. The results revealed two distinct patterns: some germline cells segregated during inflorescence meristem formation (late segregation), while others segregated earlier than branch formation (early segregation).

This early segregation provides a plant counterpart to Weismann’s theory, showing that germline segregation in Arabidopsis occurs earlier than previously believed. Ultimately, the study illustrates how plants balance developmental plasticity with safeguarding the genetic integrity of their offspring.

_________________________________________________________________________________________________________

By Liu Jia Chinese Academy of Sciences

The article can be accessed on: Phys.org

Image credit: Current Biology (2026)

The biochemical basis of this selectivity lies in the enzyme’s interaction with the Protospacer Adjacent Motif (PAM). Structural analysis led by Dr. Hong Li revealed that ThermoCas9’s binding affinity is governed by the presence of methyl groups within the PAM sequence. In healthy cells, the methyl group acts as a physical protrusion that prevents the enzyme from seated correctly, akin to a screwdriver unable to fit into a blocked screw head. Because cancer cells exhibit altered methylation patterns, the enzyme can successfully bind and cleave tumour DNA while leaving healthy, methylated DNA intact.

While this study marks the first instance of a CRISPR system responding to the most abundant type of human DNA methylation, the authors characterise the findings as an early-stage proof-of-concept. Current results demonstrate successful DNA cleavage in human cell cultures but have not yet confirmed the induction of apoptosis. Future research will focus on maximising DNA damage to ensure therapeutic efficacy. Furthermore, this mechanism holds potential for treating other conditions defined by aberrant methylation, including autoimmune disorders and paediatric neuroblastoma.

_________________________________________________________________________________________________________

By Van Andel Research Institute

Article can be accessed on: Phys.org

Image Credit: Nature (2026).

In a mouse model, scientists led by Oleh Taratula and Yoon Tae Goo demonstrated that nanocarriers loaded with follistatin messenger RNA accumulate in tumours. Once there, the mRNA directs cells to produce the follistatin protein, which is crucial for both inhibiting tumor growth and promoting muscle tissue development.

These LNPs can be administered intravenously and are guided to the lungs by vitronectin, a blood serum protein. Vitronectin interacts with integrin receptors, which are overexpressed on tumor surfaces, effectively directing the nanoparticles to lung cancer sites. Taratula noted that this approach offers a promising solution for systemic delivery of mRNA therapeutics to lung tumors, a challenge in the field. This method achieved an approximately 2.5-fold greater reduction in tumour burden compared to conventional LNPs, which tend to accumulate in the liver.

Lung cancer is a prevalent and deadly disease, and cachexia, a severe muscle-wasting syndrome, affects many patients, contributing to up to 30% of cancer deaths. Patients with cachexia experience significant weight and muscle loss despite adequate nutrition.

By using follistatin mRNA-loaded LNPs, the researchers created a therapy that targets both lung cancer and cachexia without adverse effects. While more preclinical work is needed, the early results are encouraging, and the team hopes for future human testing.

_________________________________________________________________________________________________________

By Steve Lundeberg

The article can be accessed on: Phys.org

Image Credit:  Parinaz Ghanbari

By sequencing the entire genome of tumours, researchers were able to identify a wide range of genetic alterations, including mutations, structural variants and mutational patterns. These patterns, often referred to as mutational signatures, reflect the biological and environmental processes that contribute to cancer, such as DNA repair defects, cellular metabolism, and external exposures.

The scale of the dataset allowed the identification of both known and previously unrecognised mutational signatures. In particular, the study reported dozens of new signatures that had not been detectable in smaller cohorts, highlighting additional and still poorly understood mechanisms involved in cancer development.

Importantly, the findings demonstrate the value of WGS in capturing the full genomic landscape of tumours in a single analysis. Compared to targeted sequencing approaches, WGS provides a more complete and integrated view of cancer genomes, which can inform diagnosis, prognosis, and treatment decisions.

The study also highlights the importance of large, well-annotated datasets that link genomic information with clinical outcomes. Such resources enable more accurate interpretation of genomic changes and support the translation of genomic data into clinical care.

________________________________________________________________________________________________________________

By Danielle Gerhard, PhD

Article can be accessed on: TheScientist

Image credit:©iStock, ArtemisDiana

A recent study from the University of Warwick, published in Nature Biomedical Engineering, evaluates the reliability of artificial intelligence (AI) tools designed to predict cancer biology directly from histopathology images. These tools are often positioned as enabling faster and more cost-effective diagnostics. The study analysed over 8,000 patient samples across multiple cancer types, including breast, colorectal, lung and endometrial cancers. While several AI models demonstrated high predictive accuracy, the findings indicate that this performance is frequently driven by “shortcut learning” where models rely on indirect correlations rather than biologically meaningful signals.

For example, instead of identifying a specific mutation such as BRAF, models may detect associated features like micro-satellite instability and use these as proxies. This approach can produce accurate predictions in controlled datasets but may fail when such correlations do not hold in different clinical settings. The study highlights that current AI pathology tools may therefore lack robustness for routine clinical use. It also underscores the limitations of relying on headline accuracy as a primary performance metric, as this may mask underlying biases and confounding factors in model behaviour.

The authors emphasise the need for more rigorous and bias-aware evaluation frameworks to ensure that AI systems capture genuine biological relationships. Strengthening validation approaches will be essential to support the safe and effective integration of AI into clinical oncology and to ensure that these tools provide meaningful value beyond existing diagnostic methods.

________________________________________________________________________________________________________________

By University of Warwick

Article can be accessed on: MedicalXpress

Image credit: Dr. Fayyaz Minhas / University of Warwick.

Scientists from Penn Medicine and Children’s Hospital of Philadelphia have created a CRISPR-based functional genomics platform that can efficiently edit genes in primary acute myeloid leukaemia (AML) cells taken directly from patients. This marks a significant advance over traditional CRISPR screens performed in cancer cell lines or preclinical models, which often fail to capture the full diversity of patient tumours.

The new platform enables researchers to test hundreds of gene perturbations at once to identify which genes and regulatory elements are essential for cancer cell survival and growth. By applying CRISPR tools to heterogeneous tumour samples, the study reveals both known and previously unrecognised genetic dependencies specific to individual patient leukemias. These insights could help prioritise targets for future therapeutic development.

In this research, the team optimised CRISPR delivery into patient-derived cells to achieve high gene-editing efficiency and performed screens both in vitro and in preclinical models. They confirmed many established leukaemia “dependency” genes and also uncovered vulnerabilities that vary between patients or subtypes.

Combining CRISPR perturbations with single-cell RNA sequencing provided high-resolution understanding of how different cells within AML samples respond, showing cellular changes in gene activity, state and behaviour

Image Credit: Molecular Cell (2026). DOI: 10.1016/j.molcel.2026.02.003

________________________________________________________________________________________________________________

By Meagan Raeke, Perelman School of Medicine at the University of Pennsylvania

Article can be accessed on: MedicalXpress

Colibactin is a potent and highly unstable toxin produced by certain strains of Escherichia coli and other gut bacteria. It is known to cause DNA mutations linked to colorectal cancer, but its instability has made it challenging to study. Researchers in the U.S. have now determined exactly how colibactin attacks DNA by using advanced analytical tools, including mass spectrometry and nuclear magnetic resonance spectroscopy. To overcome the toxin’s rapid breakdown, they grew colibactin-producing bacteria directly next to DNA strands, allowing them to capture the toxin’s effects immediately.

The scientists discovered that colibactin targets DNA sequences rich in adenine and thymine bases. Its mode of attack involves forming an interstrand cross-link, a bridge-like structure that binds the two DNA strands together. This permanent damage prevents cells from properly reading or copying their genetic material, ultimately causing mutations associated with cancer development. They also found that the toxin attacks the minor groove of DNA, the narrower region where the DNA backbones are closest. This specificity is driven by colibactin’s unstable, positively charged core, which is naturally attracted to the negatively charged, AT-rich minor groove.

Understanding the structure of the DNA cross-link and the mechanism of damage may support the development of diagnostic tools, therapeutic strategies to neutralize colibactin, and potential approaches to reduce cancer risk.

Image Credit: Science (2025). DOI: 10.1126/science.ady3571 (MedicalXpress)

By Paul Arnold

The therapy uses “universal” CAR T-cells engineered from healthy donor cells. Key base-editing steps included: removing receptors to make the cells “universal”; removing the CD7 flag to prevent “friendly-fire” self-destruction; removing the CD52 flag to resist strong antibody drugs; and adding a Chimeric Antigen Receptor (CAR) that targets the CD7 flag on leukemic T-cells.

The results of the clinical trial, published in the New England Journal of Medicine, include data from 10 patients (eight children and two adults). Key findings show that 82% of patients achieved very deep remissions, allowing them to proceed to a stem cell transplant, and 64% remain disease-free. Side effects were tolerable. The first patient, Alyssa, is now three years disease-free and off treatment, highlighting the therapy’s potential to provide hope for patients who do not respond to standard care.

Image credit: Unsplash/CC0 Public Domain (MedicalXpress)

By University College London

Article can be accessed on: MedicalXpress

Lead researcher Dr. Andres Cifuentes-Bernal explains that the system analyses how genes influence each other over time, providing a clearer picture of the biological processes that allow tumours to grow, spread, and resist therapy. Traditional cancer studies focus mainly on common mutations, but this overlooks rare changes and the complex interactions that give malignant cells momentum. The new framework fills this gap by capturing the dynamic nature of cancer.

Using large breast cancer datasets, the AI method identified both well-known cancer genes and previously hidden ones that while not mutated still affect other genes and shape tumour progression. It accurately detected many cancer drivers listed in the Cancer Gene Census, confirming its reliability. Several newly discovered gene candidates are linked to cell signalling, immune response, and metastasis.

Associate Professor Thuc Le says the approach highlights cooperative networks rather than isolated genes, offering deeper insight into tumour evolution. The adaptable method may also assist in understanding other diseases involving shifting gene regulation, including neurodegeneration and autoimmune disorders.

Image Credit: Unsplash/CC0 Public Domain (MedicalXpress)

By University of South Australia