Unexpected Finding: Hydration-based Shape Memory of Bioadhesive
Experiments can often be messy. They involve long preparations, numerous apparatus, and hardworking machines (and humans) continuing for hours if not a whole day. Hectic and exhausting experiments sometimes push tired researchers to be lazy on one very important thing – cleaning up after experiments. Around 3 years ago, it was a late-night after whole-day mechanical tests on adhesive hydrogels in the lab when such laziness overwhelmed tired hands to leave a poor adhesive hydrogel behind stretched by a mechanical tester overnight.
The next morning, together with other mess to clean up, the poor adhesive hydrogel was still at the mechanical tester elongated a few times of the original length and stayed overnight. When the sample was unloaded to clean up, it was fully dried. But the dried adhesive hydrogel also showed something peculiar – drying froze the adhesive in the stretched state and unloading it did not recoil back to the original length. It was interesting enough to play with and it shrunk back to the original length when it was hydrated again with water. It was very fun, and curiosity pushed further to try a little crazy thing: cutting a small piece of the dried bioadhesive (in the stretched state) and putting on a piece of pork meat in the lab freezer. Surprisingly, the adhesive shrunk and contracted the pork meat as it adhered to wet tissue & hydrated!
After digging more, we learned that it was a kind of shape-memory effect driven by the rubbery-to-glassy transition of hydrogels based on water contents. When hydrogels are hydrated (or wet), swollen polymer networks of hydrogels exhibit rubbery behavior (like elastic rubbers). But, when hydrogels dry and lose water, dried polymer networks of hydrogels become less mobile and exhibit glassy behavior (like stiff plastics). While this has been a well-known phenomenon, what makes it very interesting is when such a hydration-based shape-memory effect is combined with our bioadhesive hydrogel.
Our previously developed bioadhesive takes a dry form to implement the dry-crosslinking mechanism to rapidly remove interfacial water and form near-instant robust adhesion to wet tissues (for details, please see Nature 575, 169-174 (2019)). Interestingly, the hydration-based shape-memory mechanism makes a perfect synergy with the dry-crosslinking mechanism: a dry bioadhesive can achieve rapid robust adhesion to wet tissues while hydration of the dry bioadhesive during and after adhesion provides controlled mechanical modulation by releasing the programmed strain. It’s like two birds with one stone and we were very excited that this could be really useful for something.
Joint Force: Collaboration Toward Better Treatment for Diabetic Wounds
After thinking hard about where this interesting strain-programmable bioadhesive could be used, wound healing experts from Beth Israel Deaconess Medical Center (BIDMC) across the Charles River joined the team to share wisdom. Boosted by the BIDMC team’s expertise, we started to see our strain-programmable bioadhesive as a smart wound dressing that could provide not only protective mechanical cover to wounds but also active mechanical modulation to promote healing. We were in the course of completing a detailed characterization of the diabetic wound healing micromilieu with single-cell transcriptomics and had discovered a unique healing-associated fibroblast subpopulation (for details, please see Nature Communications 13, 181 (2022)). As fibroblasts are key players in extracellular matrix remodeling and contraction through myofibroblast differentiation and this process is severely dysregulated in diabetes, we hypothesized that manipulating wound mechanics could be beneficial for the acceleration of tissue repair.
Deeper Understanding: Benchtop, Finite Element & Pre-clinical Studies
With a promising technology and shared vision for better treatment of diabetic wounds, our team was super motivated to embark on a transformative scientific journey to best develop, optimize, and validate a therapeutic solution. With a unique and strong combination of broad expertise in our team, ranging from materials science and mechanics to biology and clinical treatment, we planned highly ambitious multi-disciplinary studies that have closely informed each other to build a deeper scientific understanding of our technology platform for diabetic wound healing.
First, we designed and performed an extensive set of benchtop tests to characterize mechanical properties, adhesion performance, and strain-programming & release of the strain-programmed patch. As we could see the potential of the strain-programmed patch as a versatile platform not only for diabetic wounds but also for other applications, we tried to provide great details of science and data from our benchtop studies in our work. This ended up in lengthy Supplementary Information sections longer than the paper!
Second, we built theoretical and finite element (FE) models of our system to evaluate the strain-programmed patch’s capability to mechanically modulate diabetic wounds in silico. One remarkable insight we got from FE analyses was that the presence of native pre-strain (or pre-tension) in the skin generated the initial enlarging of the wound and substantial hoop stress concentration around the wound edge (imagine making a defect in stretched rubber sheet). Considering the reduced contractility of diabetic wounds and the importance of mechanoresponsive mesenchymal cells (such as fibroblasts) in diabetic wound healing, we hypothesized that alleviating the wound enlarging and hoop stress concentration at the wound edge by the strain-programmed patch could be a promising (and uniquely explorable) strategy to promote healing of diabetic wounds. These analyses greatly boosted our deeper understanding of the platform and allowed us to rationally guide the optimization of the strain-programmed patch.
Lastly, we employed multiple pre-clinical models to demonstrate the efficiency of the patch. First, we showed remarkable wound closure with the patch application in the db/db mouse model and revealed the mechanisms of action with detailed flow cytometry, immunofluorescence, and transcriptomics analyses. The effect was similar in an ex vivo human skin model treated with the patch, where surgically discarded skin is kept in culture for a few days and a small wound is inflicted with a biopsy punch. Importantly, we also used highly clinically relevant models such as the diabetic pig. Porcine skin bears great resemblance to human skin in terms of architecture, cell populations, and mechanical properties and is thus the model of choice for studies of increased translational value. We found that the patch also exhibited robust wound closure in swine wounds. Finally, we also confirmed our findings in a humanized diabetic mouse model, where we grafted human skin on the back of nude mice, then induced diabetes and inflicted wounds on the human skin. Collectively, our results indicate the effectiveness of the patch across different models of increasing clinical relevance and suggest that the patch drives diabetic wound repair by facilitating new blood vessel formation, promoting keratinocyte migration, and enriching specific mechanoresponsive fibroblast subpopulations.
Outlook: Toward Clinical Translation & Beyond
With the comprehensive platform development and the strong pre-clinical data from multiple clinically relevant models, we were super excited about the potential impact of our solution to help patients with debilitating chronic wounds. Our team is working toward clinical translation of the strain-programmed patch that would require commercialization steps in the next few years including GMP manufacturing, clinical trials, and regulatory clearance. Furthermore, as one of the reviewers suggested during the review process, we will explore other applications of the strain-programmed patch including other acute and chronic wounds such as burn injuries, venous ulcers, and pressure sores. Seeing the publication of the years-long collaboration project filled with countless hours of hard work gives us an emotional finale, but bringing our solution to patients makes us excited about the next journey to come.
For more details, check out our paper “A strain-programmed patch for the healing of diabetic wounds” on Nature Biomedical Engineering.