Biology subjects

Fraser, K.

Publications and source records attributed to Fraser, K..

5 recordsLinked to original sources

bioRxiv2026

Diet-induced macrophage-driven inflammation fuels pancreatic plasticity and aggressiveness

High-fat diet (HFD) and obesity are increasingly recognized as risk factors of pancreatic ductal adenocarcinoma (PDAC), yet the mechanisms by which dietary fat contribute to oncogenic transformation remain elusive. Using an inducible acinar-specific KrasG12V/Trp53-loss genetically-engineered mouse model of PDAC, we established early-and late-onset protocols to assess age-dependent susceptibility to HFD. Specifically, HFD accelerated tumorigenesis with poorer prognosis in early-onset mice and, strikingly, enabled full PDAC development in late-onset adult mice otherwise resistant to oncogenic transformation. Tumors arising under HFD activated a distinct transcriptional and epigenetic state enriched in pathways or genes related to stemness, plasticity, and metastatic competence, which was maintained even in tumor-derived cell lines. Mechanistically, fatty acid-educated macrophages secreted the cathelicidin antimicrobial peptide (CAMP), activating P2X purinoceptor 7 (P2RX7) signaling in tumor cells to drive a highly plastic, immune-evasive phenotype reinforced by the expression of the peptidoglycan recognition protein 1 (PGLYRP1), further shielding tumor cells from macrophage phagocytosis. Functionally, HFD-induced tumors displayed enhanced metastatic potential independent of host context. Analysis of 164 human PDAC samples revealed that elevated body-mass index (BMI) was associated to a conserved CAMP-P2RX7-CXCR4 signature, maintained despite weight loss during disease progression. Together, these findings uncover a diet-imprinted macrophage-tumor cell circuit that promotes transformation and accelerates PDAC progression, positioning it as a therapeutic vulnerability in obesity-associated pancreatic cancer. Statement of significanceHigh-fat diet induces pancreatic tumorigenesis in adult tissue, driving metastatic competence and a plastic state, and ultimately engages a macrophage-derived CAMP-P2RX7 circuit that reinforces immune evasion and accelerates PDAC progression.

cancer biology↗

bioRxiv2026

AptaBLE: A Deep Learning Platform for Aptamer Generation and Analysis

Aptamers are single-stranded oligonucleotides that bind molecular targets with high affinity and specificity. However, their discovery remains time-consuming, expensive, and susceptible to experimental biases. Here we present AptaBLE, a deep learning framework for predicting aptamer-protein binding. Additionally, we demonstrate two de novo generation methods that produce novel aptamers with desired specificity profiles and Kds as low as 31 nM to-date. AptaBLE represents a significant advance towards therapeutic and diagnostic aptamer development.

bioinformatics↗

bioRxiv2025

Lipid Nanoparticles with Aptamers Enable Targeted mRNA Delivery to CD4⁺ T Cells

In vivo genetic engineering of T cells could overcome the logistical, biological, and safety challenges of ex vivo modification, but effective and safe delivery systems remain limited by a lack of cellular specificity. Here, we developed aptamer-functionalized lipid nanoparticles (LNPs) for targeted mRNA delivery to CD4+ T cells, employing both a validated CD4-binding aptamer (Apt62) and novel aptamers generated using our proprietary transformer-based AI language model, AptaBLE. LNPs formulated with ionizable lipid SM102 or MC3 and conjugated with aptamers at controlled densities were physiochemically characterized and assessed for binding, in vitro transfection, in vivo biodistribution, and safety evaluation. Aptamer-functionalized LNPs demonstrated selective nanomolar binding to recombinant CD4, achieved enhanced transfection of CD4+ versus CD4- T cells in vitro, and significantly enriched mRNA delivery to immune-rich tissues in vivo, achieving up to 70-fold spleen signal enhancement with SM102 formulations compared to non-targeted controls, while maintaining suitable safety profiles. Overall, these findings demonstrate aptamer-functionalized LNPs, augmented by AI-guided aptamer design, as a tunable, non-immunogenic platform for in vivo T cell engineering. HighlightsO_LIAptamer-functionalized LNPs enable selective mRNA delivery to CD4+ immune cells rich organs. C_LIO_LIAptamer-functionalized LNPs maintain a favorable systemic safety profile in vivo. C_LIO_LIAI-guided AptaBLE platform generated functional aptamers validated in nanoparticle delivery. C_LI

molecular biology↗

bioRxiv2025

EVOFLOW-RNA: GENERATING AND REPRESENTINGNON-CODING RNA WITH A LANGUAGE MODEL

RNA plays a critical role across numerous biological functions. Recent advances in language modeling show promise with representing RNA, but the possibility of large-scale RNA design and optimization has not been fully explored. We propose EvoFlow-RNA, a bidirectional non-coding RNA language model leveraging a masked discrete diffusion model (MDM) formulation for both generative modeling and representation learning. EvoFlow-RNA bridges the gap between RNA sequence representation and design. It outperforms leading RNA models on three BEACON tasks critical to understanding RNA function, spanning from structure prediction to gene editing. For unconditional generation, it synthesizes diverse RNA sequences with native-like structural and binding properties. Additionally, EvoFlow-RNA can globally redesign aptamer sequences around preserved binding recognition sites with enhanced functionality. Our results demonstrate the effectiveness of EvoFlow-RNA in RNA modeling, highlighting the capability and potential of masked discrete diffusion for both recapitulating and enhancing existing RNAs.

synthetic biology↗

bioRxiv2019

Designer DNA architecture offers precise and multivalent spatial pattern-recognition for viral sensing and inhibition

DNA, when folded into nanostructures of customizable shapes, is capable of spacing and arranging external ligands in a desired geometric pattern with nanometer-precision. This allows DNA to serve as an excellent, biocompatible scaffold for complex spatial pattern-recognizing displays. In this report, we demonstrate that a templated designer DNA nanostructure achieves multi-ligand display with precise spatial pattern-recognition, representing a unique strategy in synthesizing potent viral sensors and inhibitors. Specifically, a star-shaped DNA architecture, carrying five molecular beacon-like motifs, was constructed to display ten dengue virus envelope protein domain-III (ED3)-binding aptamers into a 2D pattern precisely matching the pentagonal arrangement of ED3 clusters on the dengue viral surface. The resulting spatial pattern recognition and multivalent interactions achieve high dengue-binding avidity, conferring direct, highly-sensitive, facile, low-cost, and rapid sensing as well as potent viral inhibition capability. Our molecular-platform design strategy could be adapted to detect and combat other disease-causing pathogens, including bacteria and microbial-toxins, by generating the requisite ligand patterns on customized DNA nanoarchitectures.

bioengineering↗