Sprayable and biodegradable, intrinsically adhesive wound dressing with antimicrobial properties

Abstract Conventional wound dressings are difficult to apply to large total body surface area (TBSA) wounds, as they typically are prefabricated, require a layer of adhesive coating for fixation, and need frequent replacement for entrapped exudate. Large TBSA wounds as well as orthopedic trauma and low‐resource surgery also have a high risk of infection. In this report, a sprayable and intrinsically adhesive wound dressing loaded with antimicrobial silver is investigated that provides personalized fabrication with minimal patient contact. The dressing is composed of adhesive and biodegradable poly(lactic‐co‐glycolic acid) and poly(ethylene glycol) (PLGA/PEG) blend fibers with or without silver salt (AgNO3). in vitro studies demonstrate that the PLGA/PEG/Ag dressing has antimicrobial properties and low cytotoxicity, with antimicrobial silver controllably released over 7–14 days. In a porcine partial‐thickness wound model, the wounds treated with both antimicrobial and nonantimicrobial PLGA/PEG dressings heal at rates similar to those of the clinical, thin film polyurethane wound dressing, with similar scarring. However, PLGA/PEG adds a number of features beneficial for wound healing: greater exudate absorption, integration into the wound, a 25% reduction in dressing changes, and tissue regeneration with greater vascularization. There is also modest improvement in epidermis thickness compared to the control wound dressing.

Conventional wound dressings-typically adhesive-coated, nondegradable, and thin polymer films-are prefabricated and nonconformal, making them difficult to use on large TBSA wounds or traumatic injuries, which are irregular in shape and depth. To prevent hematoma formation and exudate buildup, these dressings often require frequent changes that are painful, disrupt the healing epidermis, and may increase the risk of infection. 6,7 Hydrogel dressings attempt to address these limitations by providing a moist wound healing environment that can be delivered as soft, moldable material. 8,9 When removal is necessary, they can be dissolved on-demand by exchanging crosslinks. 10 Nanofibrous wound dressings provide excellent absorption of wound exudate and oxygen permeation because of their high porosity, but are typically prefabricated using electrospinning. 11,12 Composites of both approaches have been explored to provide a combination of porous structure and moisture. 13 However, no current wound dressing provides custom sprayable fabrication with no patient contact, exudate absorption, and adhesion to the wound without additional fixation.
Here, we investigate solution blow spinning (SBS) as a sprayable method for the direct deposition of biodegradable polymer fibers containing antimicrobial silver onto wounds. Unlike electrospinning, which uses an applied voltage to drive fiber production and has low production rates, SBS uses a pressurized gas to produce fibers from a polymer solution with high production rates. 14 A polymer blend solution of poly(lactic-co-glycolic acid) and poly(ethylene glycol) (PLGA/PEG) sprayed with a portable airbrush, produces intrinsically adhesive polymer fibers that accumulate on and seal the wound. Silver salts are incorporated to create a sprayable and antimicrobial wound dressing (PLGA/ PEG/Ag) that can release bactericidal silver ions and reduce the risk of infection. These commonly used silver salts have broad-spectrum antimicrobial activity with relatively low minimum inhibitory concentrations (MIC) and minimum bactericidal concentrations (MBC). [15][16][17][18][19][20] In this report, we examined the effect of silver nitrate (AgNO 3 ) on silver ion (Ag + ) release, mechanical properties, and adhesion of PLGA/ PEG. in vitro studies were used to determine the optimal concentration of AgNO 3 loaded into PLGA/PEG spinning solutions. To demonstrate the feasibility of using PLGA/PEG and PLGA/PEG/Ag wound dressings, they were evaluated in an in vivo porcine partial-thickness excisional wound model. The incorporation of biodegradable PLGA/ PEG into the scab and its absorption of wound exudate were examined using histology and fluorescence microscopy. Dressing changes were made as needed and tracked to demonstrate the potential benefits of using an intrinsically adhesive dressing that is biodegradable and can be absorbed by the wound.

| Effect of solvent on fiber morphology and silver content
We first evaluated whether solvent properties could affect the morphology of fibers produced by SBS, which is a critical factor in fiber formation. Acetone has been previously used for SBS processes and does not affect cell viability. 21,22 However, AgNO 3 's solubility in acetone at room temperature is only 4.4 mg/mL. 23 Ethyl acetate has lower toxicity than acetone and is able to dissolve nearly 10 times as much AgNO 3 , up to a concentration of 27 mg/mL. 24 This makes it an excellent alternative solvent for SBS of polymers in situ.
The pure PLGA/PEG samples for both acetone and ethyl acetate resemble each other in morphology (Figure 1a,e). However, SBS from acetone produces fibers that exhibit an AgNO 3 dependent change towards a beads-on-a-string morphology (Figure 1b-d), which is characteristic of polymer solution jets that form spherical structures during the spraying process. 25  Energy-dispersive X-ray spectroscopy (EDS) was used to perform elemental analysis on blowspun samples of PLGA/PEG containing from 1 to 10 mg/mL of AgNO 3 in solution. From three 40 × 40 μm images, as shown in Figure 1k-l with red color superimposed, we estimated the weight percent (wt%) of silver in blow spun fibers produced from three different PLGA/PEG/Ag solutions containing 1, 5, and 10 mg/mL of AgNO 3 . The measured wt% varies from 0.67 to 2.4% when sprayed from acetone and from 0.14 to 1.19% when sprayed from ethyl acetate ( Figure 1m).

| Thermal and mechanical properties
The effect of AgNO 3 on the mechanical and thermal properties of PLGA/PEG was then evaluated. At the highest concentration tested

| Silver release kinetics
The release rate of Ag + was studied by immersing samples of PLGA/ PEG/Ag in 37 C water for up to 1 month. Periodically over 28 days, the supernatant was collected and analyzed using inductively coupled plasma-atomic emission spectroscopy (ICP-AES) to determine the concentration of Ag + . This study tested three different AgNO 3 concentrations in polymer solution (0.5-5.0 mg/mL). The objective of this study was to determine which formulations released an appropriate amount of Ag + , guided by three benchmarks described in the literature: (a) The 24 hr cytotoxic limit, which is approximately 50 μM for fibroblasts, 29,30 (FCC) and higher for keratinocytes, 500 μM (KCC). 31,32 (b) The (MBC, which is approximately 20 μM for a small model bacterial inoculation. 33 (c) The MIC, which is around 5 μM for an inoculation of the same size. 33 Depending on initial AgNO 3 loading concentration, PLGA/PEG/ Ag releases between 0.5 and 5 μmol per gram of wound dressing ( Figure 3a). When loaded with more AgNO 3 , the release profile changes, with silver release extending over 7 days for 1 and 5 mg/mL PLGA/PEG/Ag (Figure 3b). Ag + release is nearly complete within 1 day for PLGA/PEG/Ag with 0.5 mg/mL AgNO 3 . Higher AgNO 3 concentrations had approximately 50% release at 1 day and reached 100% release after approximately 14 days. The differences in release kinetics are related to the size of the burst phase, which occurs during the first hours of drug release and typically accounts for a significant portion of drug release in PLGA devices. [34][35][36] We estimated the Ag + concentration in a partial-thickness wound sprayed with 2 mL of polymer solution, given a penetration depth of Ag + into the wound bed of 7.5 mm. 34 Each data set was fit to a logarithmic model ( Figure S1), which was then used to estimate absorption, distribution, metabolism, and excretion of Ag + in the wound using previously described first order kinetics. 35 This model, which combines in vitro Ag + release data with realistic approximations for wound volume and various clearance rates, is shown in F I G U R E 1 SEM images of PLGA/PEG fibers produced using solution blow spinning with increasing concentrations of AgNO 3 added to the spinning solution. Fibers produced using acetone as the spinning solvent (a-d) create a beads-on-a-string morphology when loaded with AgNO 3 , while those made with ethyl acetate (e-h) have a consistent web-like fiber morphology. Scale bar = 100 μm. When using ethyl acetate as the spinning solvent, there are decreases in fiber diameter with AgNO 3 concentration (i), while porosity (j) is similar (n = 2-4). Energy dispersive x-ray spectroscopy (EDS) shows that PLGA/PEG fibers produced using solution blow spinning with ethyl acetate contains silver. EDS signal (red) superimposed on scanning electron microscopy images of fibers produced from polymer solution containing 1 mg/mL (k) and 10 mg/mL (l) AgNO 3 . Scale bar = 10 μm. (m) Plot of estimated weight percent of silver calculated from EDS for PLGA/PEG spinning solutions varying in AgNO 3 concentration. Asterisks indicate statistical significance: *p < .05; **p < .01; ***p < .001 This model likely overestimates the rate of Ag + release due to the aqueous in vitro environment, which accelerates Ag + release compared to the interface of the healing wound.

| Antimicrobial activity and L929 cytotoxicity
Silver, while antimicrobial, can be cytotoxic at high concentrations.
We aimed to determine a concentration of AgNO 3 to incorporate into the polymer mats that suppresses bacterial growth with minimal toxicity. Bacterial growth inhibition was measured by the method of Kirby-Bauer disk diffusion (Figure 4a). This inhibition was tested using two commonly infectious Gram-negative and Gram-positive Cytotoxicity towards L929 mouse fibroblasts was tested at the antimicrobial concentrations of AgNO 3 ( Figure 4b). As shown, an AgNO 3 concentration of 5 mg/mL led to significant cytotoxicity compared to 95% cell viability at 1 mg/mL. The optimal concentration of AgNO 3 of 1 mg/mL produced antimicrobial efficacy yet also allowed for sufficient cell viability.   Angiogenesis is an important factor in the early stages of wound healing that can be measured by the dermis blood vessel density. 37 Despite initially being similar at PWD 7, blood vessel density is signifi-

| DISCUSSION
An ideal dressing is easy and painless to apply, antimicrobial, keeps a moist wound environment, and requires minimal dressing changes while still protecting the wound. Here, we demonstrate that SBS allows for in situ sprayable wound dressing deposition with minimal wound contact. This ensures consistency with "no-touch" technique, which is used in clinical practice to minimize transfer of infectious microorganisms and protect the wound. 38 (Figure 8c). This study further supports existing evidence that biodegradable polyesters-although relatively uncommon in wound dressings-can produce excellent results as scalable, low cost alternatives for large TBSA wounds. 48,49 Wound healing is a complex and tightly regulated inflammatory process, during which excessive inflammation leads to greater scaring. 50 Skin injury produces a cascade of events that includes infiltration of neutrophils, fibroblasts, and endothelial cells. 51  closer, respectively, to healthy skin than that of Tegaderm. Significant epidermal hypertrophy indicates greater levels of inflammation during the early stages of wound healing. 56 Although similar at PWD 7, at PWD 35, blood vessel density is significantly increased for PLGA/PEG and PLGA/PEG/Ag compared to Tegaderm (Figure 7d). This indicates greater levels of angiogenesis in the healing dermis during the proliferative stage of wound repair which is critical for healing. 57,58 Revascularization could be stimulated by the lactic acid supplied by the biodegradable PLGA scaffold, which has been shown to increase angiogenesis. [59][60][61] High expression of TGF-β leads to excess fibroblast activity, which contributes to the development of hypertrophic scars and keloids. The composition of collagen subtypes can also give insight into the degree of scarring that may ultimately develop in the wounds.
Scarless healing has been linked to greater collagen III deposition, while hypertrophic and keloidal scars have increased collagen production and an increased ratio of collagen I to collagen III. 62  Pigs, which provide the best animal model for human skin, 63 expose the wounds to unpredictable trauma and shear forces.
Tegaderm dressed wounds suffered significantly greater dressing loss than those dressed with PLGA/PEG or PLGA/PEG/Ag (Figure 8c). The most likely reason is the moist environment, which reduces the adhesiveness of the dressing. Ultimately, PLGA/PEG and PLGA/PEG/Ag have better adhesion to wounds because they form a durable protective interface with tissue that is reinforced with fibrin from coagulated blood and can absorb exudate. However, shear forces from animal movement and cage trauma led to some dressing retraction from the wound edges. This likely created a drier and less protected environment at the wound edge, delaying the rate of re-epithelialization seen at the biopsy sites.

| Scanning electron microscopy and energy-dispersive X-ray spectroscopy
A Hitachi SU-70 Schottky field emission gun scanning electron microscope was used to image nanofiber mats sputter coated with gold.
Snapshots were taken across the surface of the fiber mat. Fiber diameter and porosity were determined using the DiameterJ plugin for ImageJ (n = 2-4). 64 EDS was used to measure weight fraction. EDSdetermined abundances were converted to weight fraction based on the three primary elemental components of the polymer fibers being carbon, oxygen, and silver.

| Tensile mechanical testing
Tensile tests were made using a TA Instruments DMA Q800 equipped with a film tension clamp. Samples were stretched under a controlled force ramp from 0 to 5 N at a rate of 0.01 N min −1 . Measurements were made either at room temperature or at 37 C after a 10 min isothermal period.
Elastic modulus was calculated from the linear region of the resulting stress/strain curve. Each sample type was replicated five times (n = 5).

| Differential scanning calorimetry
Approximately 10 mg samples of fiber mats were sealed in aluminum hermetic pans (TA Instruments) using a sample encapsulation press. Differential scanning calorimetry (DSC) measurements were made on a TA Instruments DSC Q100. Samples were held isothermal at −50 C for 5 min and then heated and cooled from −50 to 80 to −50 C, at a rate of 10 C min −1 for two continuous cycles. The inflection point of the heat flow during the T g was used to determine the midpoint.

| Silver release studies
A 2 mL of polymer solution was used to fabricate fiber mats containing various amounts of AgNO 3 , which were then weighed (n = 3).
The fiber mats were then submerged in 4 mL of deionized water, and stored at a constant temperature of 37 C. The supernatant was sampled periodically over 30 days. Samples were analyzed by ICP-AES on a Shimadzu ICPE-9000, measuring at 328 nm for Ag + . Mass of silver released was normalized to the mass of the fiber mat.

| Wound closure adhesion testing
Wound closure adhesion testing was performed on the TA Instruments DMA Q800. 1 cm by 1 cm sections of porcine skin were attached to rectangular clamps using cyanoacrylate glue. The rectangular clamps were brought together end to end, and 1 mL of sealant polymer solution was deposited on this joint, closing the gap between the two skincoated clamps (see ASTM F2458-05). 65 The sealant was carefully applied and trimmed to avoid coating the interface between the ends and edges of the clamps. It was then allowed to set at 37 C for 10 min before testing. A controlled force ramp was used to increase force at a rate of 1 N min −1 until failure. Failure type was recorded as either adhesive or cohesive. Force values were normalized to the surface area of skin coated by the adhesive, which was measured using calipers, giving adhesive strength. Each sample type was replicated five times (n = 5).

| Antimicrobial zone of inhibition
The antibacterial effects of the PLGA/PEG/Ag dressing were tested in vivo against Staphylococcus aureus (ATCC 6538) and Escherichia coli (ATCC 8739) using standard disk susceptibility testing methods. 66,67 PLGA/PEG/Ag antimicrobial effects were tested against a broadspectrum antibiotic, gentamicin sulfate (MilliporeSigma). Blank 10 mm Kirby-Bauer disks (MilliporeSigma) were loaded with gentamicin solution in sterile water at a potency of 10 μg. 68 After air-drying for 15 min, the gentamicin disks were pressed onto the agar. Disks of PLGA/PEG/Ag (1 mg/mL, Ag-M, and 5 mg/mL, Ag-H) were made by spraying 5 mL of PLGA/PEG/Ag solution onto sterile glass coverslips and cutting into 5 mm disks (n = 3-5). The disks were pressed onto the agar surface with at least 2 cm distance between disks. All disks were then wet with 30 μL sterile water to facilitate release of Ag + .
After incubation for 24 hr at 37 C, the ZOI for each disk was measured using a caliper. ZOI was normalized to the disk diameter.

| Statistical analysis
Statistical analysis was performed on Stata (StataCorp) or Origin (OriginLab). Typically, one-way ANOVA was used to compare group variation, followed by post hoc pairwise Tukey comparison to determine significant differences between the groups. This procedure was not used for Figure 7a because the data sets are not normally distributed. Statistical significance is considered for p < .05. Typically, averages were plotted with error bars representing standard error. If no asterisks are shown, there are no significant differences among the groups.

| CONCLUSIONS
Clinically, most partial thickness wounds would be dressed with antimicrobial ointments or dressings, like bacitracin or silver sulfadiazine, and gauze. These dressings are not occlusive and require daily dressing changes. They are often time-consuming and difficult to apply, and painful for the patient. By controllably releasing AgNO 3 from solution blow spun PLGA/PEG/Ag, the requirements for a rapid, broad-spectrum antibiotic treatment and a highly adhesive wound dressing that absorbs exudate are met. Release of Ag + can be tailored for lasting antimicrobial activity without incurring high cytotoxicity. The sprayable SBS process allows for simple application of a conformal dressing for wounds of any shape and size that does not need to be removed or changed. Overall, the use of PLGA/PEGbased dressings is simple, effective, and was without significant wound complication or delay in healing in a porcine partial thickness wound model.

CONFLICT OF INTEREST
The authors have no conflicts of interest.