CRISPR CRITTERS
CRISPR enabled precision oncology: from gene editing to tumor microenvironment remodeling
image:
This figure illustrates the evolution of CRISPR technology from 1987 to 2019, presented in a horizontal timeline format and categorized into four generations, each denoted by a distinct color: The first generation (blue) represents DNA double-strand break editing technology, initiated with the establishment of the SpCas9 system in 2007 and subsequently refined for precision through advances such as the Cas9-D10A nickase and the high-fidelity SpCas9-HF1 in 2015; The second generation (orange) refers to targeted DNA cutting technology, including Cas12, developed in 2015, and Cas14, invented in 2018, which overcame PAM sequence restrictions, thereby enabling a broader target range and expanding applications from gene editing to molecular-level diagnostics; The third generation (green) involves targeted RNA regulation technology, encompassing the CRISPRi/a system for transcriptional regulation, the Cas13 system, invented in 2016, for the specific degradation and modification of cancer-related mRNA, and dCas9-DNMT3A/TET1-mediated DNA methylation editing developed between 2022 and 2023; The fourth generation (yellow) encompasses precise editing technologies that avoid double-strand breaks, including base editors (CBE/ABE), developed in 2016, and the Prime Editor, introduced in 2019, which have broadened applications to include the repair of diverse point mutations and the correction of tumor driver genes such as EGFR.
view moreCredit: Kailai Li，Peixin Huang，Yue Qian，Anqi Lin，Jingjun He，Junyi Shen，Li Chen，Kai Miao，Jian Zhang
CRISPR gene-editing technology, renowned for its precision, is revolutionizing oncology research and treatment. Its evolution spans from the initial DNA double-strand break-inducing Cas9 to more advanced systems like Cas12, Cas13, base editors, and prime editors. These tools have expanded the scope of CRISPR beyond simple gene knockout to include RNA editing, transcriptional regulation, and epigenetic reprogramming, establishing a versatile platform for precision cancer interventions.
The generational progression of CRISPR systems has directly fueled innovations in cancer research. First-generation Cas9 enabled foundational gene knockout studies. Subsequent generations introduced Cas12 for DNA targeting with different PAM requirements, Cas13 for RNA degradation, and base/prime editors for precise nucleotide changes without inducing double-strand breaks. This functional expansion allows researchers to systematically investigate cancer mechanisms, from identifying driver genes to modeling drug resistance.
A primary application of CRISPR in oncology is the systematic identification of tumor driver genes and synthetic lethal targets through high-throughput screening. Genome-wide CRISPR libraries, such as GeCKO, have been instrumental in pinpointing genes essential for cancer progression, metastasis, and drug resistance across various cancer types. Furthermore, technologies like Perturb-seq, which combines CRISPR perturbation with single-cell RNA sequencing, enable the mapping of gene regulatory networks and heterogeneous cellular responses at single-cell resolution. CRISPR is also pivotal in dissecting the tumor microenvironment (TME) and immune evasion mechanisms. It is used to study metabolic reprogramming by targeting enzymes like LDHA, modulate angiogenesis via genes like VHL, and disrupt immune checkpoints such as PD-L1 and CD47. By editing these components, CRISPR helps reveal how tumors evade immune surveillance and suggests strategies for TME remodeling to enhance anti-tumor immunity.
In therapeutic development, CRISPR enables precise targeting of oncogenes and reactivation of tumor suppressor genes. Strategies include inactivating fusion oncogenes like BCR-ABL and EML4-ALK, or restoring the function of TP53 and PTEN. A major focus is engineering immune effector cells; CRISPR enhances CAR-T and NK cell therapies by knocking out inhibitory receptors (e.g., PD-1, TGFBR2) to improve persistence and cytotoxicity within the immunosuppressive TME, and facilitates the development of universal allogeneic cell products.
The effective delivery of CRISPR components remains a critical challenge. Viral vectors, such as AAV and lentivirus, offer high efficiency but face issues like immunogenicity and limited packaging capacity. Non-viral vectors, particularly lipid nanoparticles (LNPs), provide a safer alternative with lower immunogenicity and reduced risk of genomic integration, though their transfection efficiency needs improvement. Innovations in smart delivery systems, including microenvironment-responsive and spatiotemporally controlled nanocarriers, are being developed to enhance targeting specificity and safety.
In summary, the field is advancing through the development of next-generation tools like compact Cas enzymes (e.g., CasĪ¦, Cas12f) for easier delivery, and AI-guided sgRNA design platforms (e.g., DeepCRISPR) to optimize efficiency and minimize off-target effects. Clinical trials are already evaluating the safety and efficacy of CRISPR-edited CAR-T cells and PD-1 knockout T cells. The future of CRISPR in oncology lies in integrating it with combination therapies, multimodal editing approaches, and personalized treatment strategies informed by genomic and single-cell data, ultimately driving the transition towards smarter, safer, and more precise cancer therapeutics.
Journal
Med Research
Method of Research
Literature review
Subject of Research
People
Article Title
CRISPR Enabled Precision Oncology: From Gene Editing to Tumor Microenvironment Remodeling
Article Publication Date
5-Nov-2025
Temporal evolution of large language models in oncology: performance trends of ChatGPT-3.5, ChatGPT-4, and Gemini
FAR Publishing Limited
image:
(A) Comparison of MD with 95% CI for subjective question accuracy across different LLMs over various time periods. The x-axis represents the magnitude of MD, with blue squares indicating the MD value of each study, the square size being proportional to study weight, and horizontal lines showing 95% CI. Diamonds represent the pooled MD values at the bottom of the figure. (B) Comparison of Risk Ratio (RR) with 95% CI for objective question accuracy across different LLMs over various time periods. Sample size (No. of Studies) and p-values are indicated for each subgroup. The x-axis represents the magnitude of the Risk Ratio, with blue squares indicating the RR value of each study, the square size being proportional to study weight, and horizontal lines showing 95% CI. Diamonds represent the pooled RR values at the bottom of the figure. The legend selectively displays results from fixed-effect and random-effects models based on I², including corresponding heterogeneity index I² values.
Abbreviations: LLMs, Large Language Models; MD, Mean Difference; RR, Risk Ratio; CI, Confidence Interval.
view moreCredit: Zilin Qiu, Aimin Jiang,Chang Qi, Wenyi Gan, Lingxuan Zhu, Weiming Mou, Dongqiang Zeng, Mingjia Xiao, Guangdi Chu, Shengkun peng, Hank Z.H. Wong, Lin Zhang, Hengguo Zhang, Xinpei Deng, Quan Cheng, Bufu Tang, Yaxuan Wang, Jian Zhang, Anqi Lin, Peng Luo
Large language models (LLMs) have emerged as transformative tools in healthcare, offering potential value in oncology for information retrieval, clinical decision support, and patient communication. However, the dynamic nature of oncological knowledge—including evolving treatment guidelines and diagnostic standards—raises questions about how LLMs’ performance holds up over time, especially as these models are relied on for increasingly nuanced clinical tasks.
This study, conducted in adherence to PRISMA guidelines, systematically collected relevant literature through 2025 from PubMed, Google Scholar, and Web of Science databases. The research focused on three prominent LLMs: ChatGPT-3.5, ChatGPT-4, and Gemini. Researchers analyzed 614 oncology questions spanning common malignancies (e.g., lung, breast, colorectal cancer) and rare tumors (e.g., glioma, multiple myeloma), using both original study scoring criteria and a standardized five-point Likert scale to assess response accuracy.
Key findings reveal clear divergent temporal trends across the models:
- ChatGPT-3.5 showed a consistent decline in performance (subjective questions: MD=-3.30; objective questions: RR=0.92). A notable turning point occurred between March 14 and April 26, 2023, where the model’s responses to new questions shifted from outperforming baseline queries to underperforming, with this performance gap continuing to widen over subsequent months.
- ChatGPT-4 exhibited a more pronounced drop in accuracy, with statistically significant declines observed (subjective questions: MD=-7.17; objective questions: RR=0.93), even as a more advanced iteration of the ChatGPT series.
- In contrast, Gemini demonstrated steady and significant improvement in oncology question-answering (subjective questions: MD=11.48; objective questions: RR=1.15), outpacing the ChatGPT models as time progressed.
Subjective questions—those requiring complex analysis, integration of clinical context, and nuanced judgment—were far more susceptible to temporal performance degradation than objective, fact-based queries. This disparity highlights the unique challenges LLMs face in applying evolving clinical knowledge to real-world oncology scenarios, where flexibility and alignment with the latest standards are critical.
The study’s results provide vital guidance for the responsible deployment of LLMs in oncology. As healthcare systems increasingly adopt these AI tools to support patient care and clinical decision-making, ongoing performance monitoring, standardized evaluation protocols, and strategies to integrate up-to-date clinical data will be essential to ensure safety and reliability.
Journal
Journal of Translational Medicine
Method of Research
Meta-analysis
Subject of Research
People
Article Title
Temporal Evolution of Large Language Models (LLMs) in Oncology
Article Publication Date
4-Nov-2025
Keeping engineered cells on script with nature’s playbook
Technique shields synthetic modifications from being washed away by the tide of cell growth
Genetic engineers can design and assemble sophisticated gene circuits to program cells with new functions, but important signaling molecules can become diluted as these cells grow and divide, causing the synthetic gene circuits to lose their new functions.
Xiaojun Tian, an associate professor in the School of Biological and Health Systems Engineering, part of the Ira A. Fulton Schools of Engineering at Arizona State University, and his team have discovered a way to protect these fragile genetic programs using a principle borrowed straight from nature.
The project is powered by interdisciplinary expertise in synthetic biology, modeling and metabolic engineering, as provided by David Nielsen, a chemical engineering professor in the School for Engineering of Matter, Transport and Energy, part of the Fulton Schools at ASU, and Wenwei Zheng, an associate professor of chemistry in the School of Applied Sciences and Arts, also part of ASU’s Fulton Schools.
In a new paper published in Cell, the researchers have outlined a technique that can stabilize synthetic gene circuits by forming small, droplet-like compartments inside cells through a process called liquid-liquid phase separation.
These microscopic droplets, called transcriptional condensates, act like molecular safe zones around key genes, shielding synthetically engineered modifications from being washed away by the tide of cell growth.
“When we try to program cells to perform useful tasks, such as diagnostics or therapeutic production, the genetic programs often fail because cell growth dilutes the key molecules needed to keep them running,” Tian says. “We addressed this challenge by tapping into the cell’s own strategy of phase separation to protect engineered systems.”
Borrowing from nature’s playbook
Cells use phase separation to organize their inner environment, creating compartments for essential biochemical reactions without the use of membranes. Tian’s team realized that by engineering similar condensates around synthetic genes, they could mimic this natural organization and maintain genetic stability across various cell generations.
“We discovered that by forming tiny droplets called transcriptional condensates around genes, we can protect genetic programs and keep them stable even as cells grow,” Zheng says. “It’s a simple physical solution that prevents dilution and keeps circuits running reliably.”
This approach represents a major shift from traditional strategies in synthetic biology, which have largely focused on tweaking DNA sequences or regulatory feedback loops to keep engineered systems functioning.
Instead of more complex control systems, Tian’s team introduced a physical design principle that leverages the existing spatial organization of molecules inside cells.
A new design for self-stabilizing, programmable cells
While natural cells have evolved to use condensates as a built-in protective system for regulating gene circuit activity, this study is among the first to show how it can be repurposed to stabilize synthetic programs.
“Cells already use these droplets to regulate themselves,” Tian says. “We’re now harnessing the same strategy for synthetic biology.”
Adopting this methodology could help researchers build more reliable biological systems that maintain predictable, productive functions.
“This opens a new way to build more reliable living systems, from stable cell factories to future medical applications,” Tian says. “Our strategy can become a new design principle for researchers who need their engineered cells to work consistently.”
Images taken via microscope from the study show bright, glowing clusters of transcriptional condensates inside cells, which serve as visual proof that droplets can form precisely where needed to stabilize gene activity.
“It’s exciting to see how these droplets can be used to boost bioproduction yields,” Nielsen says. “This kind of collaboration bridges fundamental biological insights with real metabolic engineering applications.”
Sourcing stability through collaboration
Tian’s group is already exploring how to engineer different kinds of condensates to control different genes, effectively turning them into programmable control hubs inside cells.
“We want to program different condensates to control different genes, creating smart cells that can adapt and function long-term,” he says. “We’re learning how to design with the cell, not against it.”
This approach to designing in accordance with nature rather than trying to override it represents a key turning point in the field. The next step is to demonstrate the technique’s applications for more diverse implementations to determine resilience and scalability, though researchers see no shortage in potential applications.
“Researchers in synthetic biology who struggle with unstable circuits will see this as a new way to make their systems more reliable,” Zheng says. “Bioprocess engineers who want a consistent yield can also use it. For biophysicists like me, it’s exciting to see physical principles like phase separation turned into practical engineering tools.”
“This work reflects a new direction in synthetic biology,” Tian says. “By using the cell’s own organizing principles, we can build systems that are both powerful and inherently stable.”
Journal
Cell
Method of Research
Experimental study
Subject of Research
Cells
Article Title
Phase separation to buffer growth-mediated dilution in synthetic circuits
Article Publication Date
7-Nov-2025
