Biology subjects

Lu, W.-J.

Publications and source records attributed to Lu, W.-J..

4 recordsLinked to original sources

bioRxiv2026

A Dynamic Stem Cell State Links Environmental Cues to Skeletal and Regenerative Fates

Stem cells faithfully generate tissue architecture during development, yet how their identity and regenerative output are reshaped as anatomical niches mature and environmental signals change with aging and injury remains poorly understood. Rare skeletal stem cells (SSCs) across diverse skeletal sites are defined by combinations of cell-surface markers, precluding their genetic interrogation in vivo. Here, using spatially informed single cell transcriptomics, we identify Pannexin-3 (Panx3) to mark a dynamic functional state within bona fide SSCs and establish Panx3-CreERT2 mice for its longitudinal interrogation. Panx3 SSCs exhibit long-term clonal persistence, self-renewal and multipotency, whereas Panx3- SSCs can acquire this state. Successive SSC cohorts engage the Panx3 state during skeletal development, osteoanabolic stimulation expands pre-existing Panx3 SSC clones and injury induces it locally in previously unlabeled progenitors. Thus, anatomically distinct progenitors can converge on a common stem cell state in response to different environmental inputs. Three-dimensional reconstruction of chondrogenic niches further identifies signals that redirect Panx3-lineage cells towards stable cartilage, enabling durable regeneration in aged and osteoarthritic joints and with human SSCs. These findings establish anatomical origin, stem cell state and lineage fate as separable dimensions of stem cell identity and provide genetic access for interrogating and directing endogenous SSC function.

cell biology↗

bioRxiv2025

Polymyxin B lethality requires energy-dependent outer membrane disruption

Polymyxin antibiotics target lipopolysaccharide (LPS) in both membranes of the bacterial cell envelope, leading to bacterial killing through a mechanism that remains poorly understood. Here, we demonstrate that metabolic activity is essential for polymyxin lethality and leverage this insight to determine its mode of action. Polymyxin B (PmB) efficiently killed exponential phase E. coli but was unable to eliminate stationary phase cells unless a carbon source was available. Antibiotic lethality correlated with surface protrusions, LPS loss, and a significant reduction in outer membrane (OM) barrier function, processes that required LPS synthesis and transport. While the energy-dependent OM disruption was not directly lethal, it facilitated PmB access to the inner membrane (IM), which the antibiotic permeabilised in an energy-independent manner, leading to cell death. Finally, we show that the polymyxin resistance determinant MCR-1 prevents PmB-mediated LPS loss and OM protrusions and thereby renders the antibiotic ineffective.

microbiology↗

bioRxiv2020

A transferrable and interpretable multiple instance learning model for microsatellite instability prediction based on histopathology images

BackgroundMicrosatellite instability (MSI) is a negative prognostic factor for colorectal cancer (CRC) and can be used as a predictor of success for immunotherapy in pan-cancer. However, current MSI identification methods are not available for all patients. We propose an ensemble multiple instance learning (MIL)-based deep learning model to predict MSI status directly from histopathology images. DesignTwo cohorts of patients were collected, including 429 from The Cancer Genome Atlas (TCGA-COAD) and 785 from a self-collected Asian data set (Asian-CRC). The initial model was developed and validated in TCGA-COAD, and then generalized in Asian-CRC through transfer learning. The pathological signatures extracted from the model are associated with genotypes for model interpretation. ResultsA model called Ensembled Patch Likelihood Aggregation (EPLA) was developed in the TCGA-COAD training set based on two consecutive stages: patch-level prediction and WSI-level prediction. The EPLA model achieved an area-under-the -curve (AUC) of 0.8848 in the TCGA-COAD test set, which outperformed the state-of-the-art approach, and an AUC of 0.8504 in the Asian-CRC after transfer learning. Furthermore, the five pathological imaging signatures identified using the model are associated with genomic and transcriptomic profiles, which makes the MIL model interpretable. Results show that our model recognizes pathological signatures related to mutation burden, DNA repair pathways, and immunity. ConclusionOur MIL-based deep learning model can effectively predict MSI from histopathology images and are transferable to a new patient cohort. The interpretability of our model by association with genomic and transcriptomic biomarkers lays the foundation for prospective clinical research.

bioengineering↗

bioRxiv2019

Hedgehog pathway activation through conformational blockade of the Patched sterol conduit

Activation of the Hedgehog pathway may have therapeutic value for improved bone healing, taste receptor cell regeneration, and alleviation of colitis or other conditions. Systemic pathway activation, however, may be detrimental and therapeutic application has been difficult for lack of agents amenable to tissue targeting. We have developed a novel agonist, a conformation-specific nanobody against the Hedgehog receptor Patched1. This nanobody potently activates the Hedgehog pathway in vitro and in vivo by stabilizing an alternative conformation of a Patched1 \"switch helix\", as revealed by cryo-EM structure determination. Although this conformation likely constitutes part of the transport cycle, nanobody-trapping disrupts the cycle and prevents substrate movement through the Patched1 sterol conduit. Our conformation-selective nanobody approach provides a new route to the development of transporter-related pharmacologic agents and may be generally applicable to the study of other transporters.

biophysics↗