Picture

From Computational Design to Clinical Translation

Welcome to the Butch/Wang Joint Research Group in the Department of Biomedical Engineering at Nanjing University. 

​Integrating computational design, delivery engineering, and experimental validation to accelerate therapeutic translation

Latest Group News:

Published on
Picture

We've published work in Chemical & Biomedical Imaging covering our development of pemefolacianine (NY-07) to address a critical failure mode in fluorescence-guided surgery. Current FDA-approved probes generate false-positive rates of 25-68% because they accumulate in both tumor and inflammatory tissue, leading to unnecessary resections and surgical complications. NY-07, designed through ML optimization integrated with CADD modeling of antifolate drugs, achieved 8-fold selectivity for tumor over inflammatory tissue. The compound is currently being advanced through Phase II clinical trials by our industrial collaborator Nanjing Nuoyuan Medical Devices following IND approval from both FDA and NMPA.

Published on
Picture

We've demonstrated that treatment timing fundamentally determines therapeutic efficacy in combination cancer therapies. Published in Advanced Science, this work establishes that a 36-hour interval between siRNA delivery and photothermal treatment doubles tumor reduction compared to non-optimized protocols. This finding has immediate implications for how we schedule any gene silencing combination therapy.

Published on
We've published an important review led by Prof Sheng of Shenzhen Institute for Advanced Technology in the Journal of Nanobiotechnology, which Professor Wang and I are happy to have contributed to. The review systematically examines cell membrane-engineered nanoparticles for brain-targeted drug delivery, providing guidance for future research in this challenging field.

The blood-brain barrier remains one of the most significant obstacles in treating neurological diseases, blocking over 98% of potential therapeutics. While various nanocarrier approaches have emerged, the field lacks a systematic comparison of different cell membrane sources and how different membrane sources cross this barrier.
We systematically compared six membrane types—red blood cells, platelets, tumor cells, macrophages, neutrophils, and NK cells—each exploiting distinct transport mechanisms. Red blood cell membranes provide immune evasion, tumor cell membranes enable disease-specific targeting, and immune cell membranes leverage natural inflammatory pathways. These work through receptor-mediated transport, tight junction modulation, or direct membrane penetration, with each approach offering complementary strengths and limitations.

Our analysis reveals significant performance variations across membrane types. Red blood cell carriers show excellent circulation times but require additional targeting modifications, while tumor cell membranes provide inherent disease specificity but face scalability challenges. Immune cell membranes offer natural inflammatory targeting but have shorter circulation half-lives. Hybrid membrane approaches attempt to combine advantages but remain understudied.
What makes this review especially valuable is the identification of critical translation barriers and future research directions. We outline key challenges including standardized manufacturing protocols, comprehensive safety characterization, and the need for mechanistic validation studies. The integration of artificial intelligence for rational design and the development of personalized membrane sources represent promising avenues for advancement.
We believe this systematic analysis provides an important roadmap for researchers working to overcome blood-brain barrier delivery challenges and advance these promising biomimetic platforms toward clinical translation.

Published on
Picture
We've published breakthrough research in Advanced Functional Materials that could transform how we prevent cancer from coming back after surgery. Our team developed a personalized cancer vaccine that can be produced in just 72 hours using the patient's own tumor cells—a dramatic improvement over current cell therapies that take 6-8 weeks.

Here's how it works: we take tumor cells removed during surgery and engineer them with an attenuated bacteria strain to increase the response they trigger when exposed to immune cells. These modified cells are then combined with L-arginine and delivered through dissolving micro-needle patches that patients can receive right at the surgical site.

The system works through three mechanisms: it amplifies presentation of the patient's specific tumor antigens, reprograms immune-suppressive macrophages into tumor-fighting cells, and generates nitric oxide that triggers inflammatory cell death in any remaining cancer cells. Specifically, the bacterial components activate the NF-κB signaling pathway, which drives macrophages to produce high levels of iNOS—an enzyme that converts the supplied L-arginine into toxic nitric oxide that kills tumor cells.

When we tested this approach in mouse models of incomplete melanoma resection, the results were remarkable. The vaccine reduced tumor regrowth by 92.7% without any other treatment, including complete remission in 20% of cases. Even more importantly, we show these vaccines have a strong abscopal effect meaning that the immune system learns to fight cancer cells anywhere in the body, not just at the micro-needle application site. 

What makes this especially promising is the speed and personalization. Unlike generic cancer vaccines, this approach uses each patient's actual tumor cells, ensuring coverage of their unique cancer signature. The rapid 72-hour production timeline means treatment can begin while the immune system is still primed from surgery.

We are excited about expanding this work out to other cancer types in the future. 

Published on
Picture
I'm excited to share that our team's latest paper on statistical image processing for cancer margin detection during surgery has been accepted in Analytical Chemistry.

Our work reveals a strong correlation between tumor and surrounding tissue fluorescence variance, enabling stable tumor boundary identification for up to 24 hours post-injection - a key finding that could help to standardize the window for guided resections.

By stratifying fluorescence signal by standard deviations above background, we demonstrated that SNR thresholds correlate strongly with cancer probability. The approach was validated first in xenograft models and then successfully translated to clinical samples through our collaboration with China Medical University, achieving millimeter scale diagnosis with 87.5% accuracy with no false negatives.

This method offers potential for precise, real-time, quantitative guidance during cancer resection to help surgeons ensure complete tumor removal while preserving healthy tissue.
Published on
Picture
I'm excited to share our latest research published in Cell Reports Physical Science on a game-changing advancement in photothermal therapy for cancer treatment. Our team discovered that the dimer of indocyanine green (ICG), ICG-II, can completely eliminate tumors in just 2 minutes, compared to regular ICG which doesn't work even after 5 minutes.

The key breakthrough centers around what we call "butterfly oscillation." ICG-II is essentially two ICG molecules linked together, and this pairing creates a constrained molecular motion that dramatically improves how efficiently light energy converts to heat. While regular ICG only converts 47% of laser energy to heat, ICG-II achieves an incredible 95.6% conversion efficiency - the highest ever reported for a small-molecule photothermal agent.

Here's what makes this so exciting: the molecular structure of ICG-II forces it to undergo a "butterfly" vibration where parts of the molecule flap toward and away from each other. This motion facilitates non-radiative energy release as heat rather than fluorescence, making it vastly more effective at destroying cancer cells through targeted heating.

In our mouse studies, direct injection of ICG-II followed by just 2 minutes of near-infrared light completely eliminated xenograft tumors in some cases, with tumor temperatures reaching over 60°C. The treatment was so effective that two mice experienced complete remission. Meanwhile, equivalent doses of regular ICG couldn't achieve tumor clearance even with longer treatment times.

What's particularly promising is that ICG-II maintains excellent biocompatibility and is actually a known, FDA-permitted compound that forms naturally as an impurity during ICG production. This could significantly accelerate the path toward clinical applications, potentially offering cancer patients a faster, more effective photothermal treatment option.

Published on
Picture
We've developed a promising new tool that could make oral cancer surgery more precise. Our team created a specialized fluorescent dye called LP-S that lights up metastatic lymph nodes, making them easier for surgeons to identify and remove.
Right now, surgeons rely on a standard dye called ICG, but it's not very specific - it highlights everything, making it hard to tell which lymph nodes actually contain cancer cells. Our new dye is much more precise,  specifically binding to EGFR receptors, which are found in high concentrations on oral cancer cells.

When we tested LP-S in mouse models, the results were impressive. The dye produced clearer, longer-lasting fluorescence in cancerous lymph nodes compared to ICG, giving surgeons a much better view of what they're working with. This means they can more accurately remove the correct lymph nodes while leaving healthy tissue in place.

If this technology makes it through clinical testing, it could be a game-changer for oral cancer patients. Better lymph node mapping means more accurate cancer staging and ultimately better treatment outcomes.

                   
Published on
Picture
We've been working with our colleagues at IceKredit on an exciting project for the  First Global AI Drug Design Competition sponsored by Baidu and Tsinghua University to help identify potential COVID-19 treatments. Using advanced machine learning, we developed a model that can predict whether new compounds might be effective against SARS-CoV-2.

What makes our approach special is that it looks at molecules from multiple angles - both the fine details of their structure and their broader chemical fingerprints. This multi-scale view helps us make more accurate predictions about which compounds are worth testing in the lab.
This work builds on our recent publication in the Journal of Computer-Aided Molecular Design and shows how our computational methods can be applied to tackle urgent health challenges like the pandemic.

Author

Chris

Categories