Multi‐Bioinspired Functional Conductive Hydrogel Patches for Wound Healing Management

Abstract Many hydrogel patches are developed to solve the pervasive and severe challenge of complex wound healing, while most of them still lack satisfactory controllability and comprehensive functionality. Herein, inspired by multiple creatures, including octopuses and snails, a novel muti‐functional hydrogel patch is presented with controlled adhesion, antibacterial, drug release features, and multiple monitoring functions for intelligent wound healing management. The patch with micro suction‐cup actuator array and a tensile backing layer is composed of tannin grafted gelatin, Ag‐tannin nanoparticles, polyacrylamide (PAAm) and poly(N‐isopropylacrylamide) (PNIPAm). In virtue of the photothermal gel‐sol transition of tannin grafted gelatin and Ag‐tannin nanoparticles, the patches exert a dual anti‐microbial effect and temperature‐sensitive snail mucus‐like features. In addition, as the “suction‐cups” consisting of thermal responsive PNIPAm can undergo a contract‐relax transformation, the medical patches can adhere to the objects reversibly and responsively, and release their loaded vascular endothelial growth factor (VEGF) controllably for wound healing. More attractively, benefiting from their fatigue resistance, self‐healing ability of the tensile double network hydrogel, and electrical conductivity of Ag‐tannin nanoparticles, the proposed patches can report multiple wound physiology parameters sensitively and continuously. Thus, it is believed that this multi‐bioinspired patch has immense potential for future wound healing management.

Bovine serum albumin (BSA) labeled by fluorescein isothiocyanate (FITC), tryptone soy broth, and liquid sabourand medium were bought from Solarbio, China.VEGF was obtained from PeproTech.E. coli (ATCC8739), S. aureus (ATCC9144), and C. albicans (ATCC10231) were obtained from Shanghai Bioresource Collection Center, China.HUVECs and endothelial cell medium were purchased from ScienCell, China.The CCK-8 and live/dead cell viability kit were derived from Thermo Fisher Scientific, USA.SYTOX Green and propidium iodide were achieved from Keygen Biotech, China.Ecoflex was purchased from Smooth-On, Inc., USA.Arduino board (UNO Rev3) was obtained from Allchips Ltd., China.

Preparation of the mucus
The mucus referred to the Ag-TA nanoparticles grafted gelatin.To obtain it, under the atmosphere of nitrogen protection and 50℃ water bath heating, gelatin (200 mg mL -1 ) and tannic acid (1 mg mL -1 ) were dissolved in ultrapure water (10 mL).AgNO 3 (1 mL, 0.1 mol L -1 ) solution was added under vigorous stirring.The pH was titrated by NaOH for maintaining at 8.5 during the 3 h reaction and finally adjusted to 7.4 with hydrochloric acid for termination.
The prepolymer of PAAm-DN could self-crosslink without N,N,N',N'-Tetramethylethylenediamine due to the catalytic performance of Ag-TA nanoparticles.Both the two DN hydrogels stood under the nitrogen atmosphere until complete gelation.

Preparation of the heterogeneous patches
A step-by-step template perfusion method was used to obtain the heterogeneous patch.Firstly, the prepolymer solution of PNIPAm-DN was filled into the suction-cup array of the mold under the vacuum and the prepolymer solution of PAAm-DN was perfused again as the backing layer.After complete gelation, the patch was carefully rolled over.The backing layer was 2 mm.Mucus was added in an amount of 50 μL cm -2 .Unless otherwise mentioned, all patches contain added mucus.

Mechanics performance test
In compressive tests, the compression started from the original height to 50% strain.The modulus was the ratio of stress to strain at this time.In the tensile tests, the tension started from the original length to break.In the cyclic tests, strains of hydrogels were cycled between 0-50% for compression and 0-100% for tension.

Photothermal performance demonstration
The hydrogel patch of 2.0 cm × 2.0 cm was put under the NIR laser of 808 nm.The temperature was recorded.The NIR irradiation was modulated by an Arduino board.Specifically, in the photothermal cycling experiment, unless otherwise mentioned, the NIR was modulated to be 1 W cm -2 for 1 min and then turned off for 3 min as a cycle.The test was repeated for 4 cycles.

Contract-relax experiment of the suction-cups
Micro suction-cup arrays of the patch were labeled with Rhodamine B. The initial diameter was recorded as Φ 0 .Then the patch was irradiated with NIR and kept at 45℃ .The diameter (Φ) at this time was also recorded.Then NIR was turned off, and the diameter (Φ) was recorded again after the size of the suction-cups became stable.The contraction rate was equal to Φ/Φ 0 .The diameter was measured by ImageJ.

Drug release experiment
The model drug FITC-BSA at a total amount of 50 μg was uniformly loaded into the mucus and suction-cups of the patch.The total mass of the patch was 1g.Then the patch was immersed in 5 mL PBS solution under 37℃ .For NIR(+) groups, the same NIR irradiation model as the previous photothermal cycle was applied additionally.The solution was sampled and drug release was measured by multimode microplate reader.

Adhesion experiments
A patch of 2.0 cm × 2.0 cm was pre-pressed under a force of 4 N to the same size pigskin while 200 μL of sol mucus was added before the force was withdrawn under 37℃ .Then the patch was exposed to the NIR irradiation for 1 min to maintain the temperature at 45 ℃ .After the temperature returned to 37 ℃ , tensile tests were performed.For removal, the patch was again exposed to the NIR irradiation for 1 min to recover the temperature to 45℃ and 10 mL of PBS at 25℃ was used to wash mucus away from the gap.For cyclic adhesion and detachment experiments, the test preparation in the adhesive state was the same as above and stopped when the stress reached 25 kPa.The nonadhesive state test was performed immediately after PBS washing.

Physiology indicator monitoring
For the temperature-resistance change rate (ΔR/R 0 ) curve, the patch attached to the surface of the pigskin was heated and the resistance was recorded.For temperature cycle monitoring, the temperature of the hot stage was cycled between 42℃ and 30℃ , and the resistance was recorded.
For the strain-resistance change rate (ΔR/R 0 ) curve, in order to obtain more standard results, the patch was attached to the neatly shaped and highly elastic rectangular Ecoflex film surface, which was used to simulate the biological tissue.The resistance of the patch was then recorded during stretching.For repeated motion monitoring, a patch was attached to the phalangeal joint of the index finger, and resistance was continuously recorded.All the tests were carried out with a bench digital multimeter, and all the individual sampling points of resistances were read after stabilization.
When serve as a non-invasive flexible electronic sensor applied on human skin, it does not require approval according to local laws.

In Vitro biocompatibility assay
For leachate biocompatibility assay, the HUVECs of initial density equal to 10 4 cells ml -1 were cultured in three days.For the control group, the endothelial cell medium was used.For the patch group, the patch was soaked at a ratio of 0.2 g patch into 1 ml endothelial cell medium during 1 day to obtain leachate, which is referred to the ISO 10993-12 standard.Then filtered leachate was used for cell culture.For co-culture assay, the 3T3 cells of initial density equal to 10 4 cells ml -1 were cultured in three days.In the experimental group, extra 0.1 or 0.2 g patch was added to each 1 ml of culture medium and co-cultured with cells.The live/dead staining and CCK-8 assay were carried out strictly according to the kit instructions every day.
Dual anti-microbial experiment E. coli, S. aureus and C. albicans liquid were prepared (OD 600 = 0.5).For the control group, 50 μL bacterial/fungal liquid was inoculated in 2.5 mL of the corresponding liquid medium in a 6-well plate.
For the patch group, 50 μL bacterial/fungal liquid was inoculated evenly on a 0.5 g patch and interacted for 15 minutes.For the patch+NIR group, additional NIR irradiation consistent with the previous photothermal cycle was applied.Then the patches were put into a 2.5 mL liquid medium (namely each 0.2 g patch add into 1 ml medium).Bacteria and fungi were cultured for 8 and 24 h respectively.Finally, OD 600 test and fluorescent staining were performed on each group.The following equation is used for the calculation: Inhibition rate=(1-OD experimental /OD control )*100% Herein, OD experiment and OD control represent OD 600 of experimental group and control group respectively.For fluorescent staining, live bacteria were stained green by SYTOX Green, fungi were stained blue by fluorescent brightener 28, and dead bacteria and fungi were stained red by propidium iodide.

In Vivo animal experiment
Male Sprague-Dawley rats of 200-250 g (Sun Yat-sen University) were equivalently and randomly divided into 5 groups.The backs of rats were shaved, and pieces of full-thickness skin with 1.0 cm diameter on the back were excited to create a skin wound model, then S. aureus (20 µL, 10 9 CFU mL -1 ) was added to the round wound bed.For the control group, the rats were only treated with PBS solution.For the patch group, a patch with a diameter of 1.2 cm adhered to the wound.For the  This reflected the interaction between networks.More specifically, for mucus, the -NH peak became weaker and the peak of tetrasubstituted\pentasubstituted benzene ring was on 822 cm -1 , indicating that the Michael addition occurred as expected.

Figure S3. (a)
Tyndall effect of the particles persisted after 24 hours stably, suggesting that the particles were well and stably dispersed in the solution.Scale bar is 1 cm.(b) Zeta potential of the particles was negative, suggesting that the Coulomb repulsion prevented agglomeration.In order to make the results observable and measurable, the mucus used in the experiment was diluted 5 times.
(c) SEM images of (i) mucus, (ii) PAAm-DN, (iii) PNIPAm-DN and (iv) Ag-Tannin NPs at low magnification.The hydrogel network were all uniform and porous.Comparatively, the network of both DN hydrogels were relatively denser than the mucus.Scale bar are 10 μm for (i)-(iii) and is 1.5 μm for (iv).(d) Schematic illustration of the main interactions in the double network hydrogel.Rigid PAAm/PNIPAm were cross-linked by covalent bonds and non-covalently entangled with gelatin chains.Rich hydrogen bonds were generated mainly between tannic acid and molecular chains.The particles provided nano-toughening.In addition, there were coordination bonds between Ag and TA, π-π interactions between TA and other aromatic groups, etc.For (b), data are shown as mean ± SD, n = 3.

Figure S2 .
Figure S2.(a) Schematic diagram of the patch composition.The backing layer was constructed with the PAAm-DN hydrogel.The micro suction-cup actuator array was constructed with the PNIPAm-DN hydrogel.The mucus was composed of Ag-TA NPs grafted gelatin and filled the gaps in the interfaces.(b) The FTIR spectrum and characteristic peak of PAAm-DN, PNIAPm-DN, and mucus.For PNIAPm-DN and PAAm-DN, the superposed characteristic peak could be observed, indicating the co-existence of double network components.Compared with PAAm and PNIPAm of a single network, the peaks of PAAm-DN and PNIAPm-DN around 3300 cm -1 became broader due to hydrogen bond association.

Figure S4 .
Figure S4.Detailed tensile strain-stress curves of (a) PAAm-DN, (b) PAAm-SN, (c) PNIPAm-DN, and (d) PNIPAm-SN.The hydrogel breaks at the highest point of the curve.The strain and stress at this time were recorded as breaking strain and strength, respectively.

Figure S6 .
Figure S6.(a) Photograph and microscopic image of the suction-cup mold.Scale bars are 0.5 cm in (i) and 500 μm in (ii).(b) The step-by-step template perfusion method.The prepolymer solution of PNIPAm-DN and PAAm-DN was perfused sequentially.After gelation, the patch was carefully rolled over, and the mucus were added when applying.(c) Fluorescence images of the heterogeneous patches from the (i) oblique view and (ii) top view.The structure of the suction-cups were stained red.Scale bars are 1000 μm in (i) and 500 μm in (ii), respectively.(d) 3D reconstruction of the patch via Micro-CT in a larger field of view.The scale bars are 2 mm.

Figure S7 .
Figure S7.(a) Fluorescence image of suction-cups opening and contracted with the surface under pre-pressure.For demonstration, the patch was inverted and pre-pressed using a clear coverslip.The contact area was colored yellow.Scale bar is 500 μm.(b) A liquid film at the contact interface between the suction-cup and the substrate sealing the negative pressure.For demonstration, the patch was inverted, and a fluorescent solution was added on the interface for visualization of the liquid film.

Figure S8 .
Figure S8.Thermal images of patches under NIR irradiation.The temperature of the patch increased from ~25℃ to ~45℃ with spatial uniformity.Scale bar is 2 cm.

Figure S9 .
Figure S9.Customized programmed NIR devices.(a) The control signal of Arduino was programmed by the computer and output to the amplifier.Then the NIR laser was controlled to irradiate the patch sample.(b) Digital photographs of the setup.

Figure S10 .
Figure S10.(a) Scheme of a reversible gel-sol conversion of the non-covalently cross-linked temperature-sensitive gelatin-based network.(b) The temperature-moduli curve of gelatin in the rheology test.G' and G'' represent the storage modulus and loss modulus.(c) Scheme of a volume phase transition (VPT) of PNIPAm.The PNIPAm network shrinks when warming up, whereas swells when cooling down, which is ascribed to its reversible change between the hydrophilic and hydrophobic states.

Figure S11 .
Figure S11.Digital photographs showing the controlled adhesion and detachment of the patch.Scale bar is 1 cm.

Figure S13 .
Figure S13.(a) Comparison of the displacement-stress curve between the non-adhesive and adhesive patches.The adhesive curve obviously showed a bigger area (adhesion energy) and height (adhesion strength) than the non-adhesive one, indicating the effectiveness of the strategy.(b) The adhesion strength at 42℃ and 45℃ .(c) The patch was attached to pig skin and placed vertically.NIR was used for irradiation and kept the temperature at 45°C for ten minutes.During this period, there was no sign of patch detachment.Scale bars are 1 cm.For (b), data are shown as mean ± SD, n = 3.

Figure S14 .
Figure S14.(a) The patch could firmly adhere to a finger joint regardless of the bent-straight motion.Scale bar is 1 cm.(b) The tensile strength in healed-unhealed cycles.(c) The stress-displacement curve of the healed and unhealed patch.(d) The schematic diagram of topological adhesion mediated controlled healing.(e) Photograph of the patch healed by mucus.Scale bar is 0.5 cm.

Figure S15 .
Figure S15.(a) Live/dead staining images of co-cultured cells.(b) Statistical analysis of CCK-8 array.In the experimental group, 0.1 or 0.2 g patch was added to each 1 ml of culture medium.For (b), data are shown as mean ± SD, n = 6.

Figure S16 .
Figure S16.(a) The ROS scavenging efficiency of the patch.Concentration of H 2 O 2 was 10 mmol/L.(b) Qualitative experiment of ROS scavenging.After the addition of mucus, the 1% H 2 O 2 solution decomposed to produce oxygen, proving its catalytic decomposition ability.Scale bar is 1 cm.For (a), data are shown as mean ± SD, n = 3.