Chronic wounds that are resistant to antibiotics are a major threat to the healthcare system. Biofilm infections are stubborn and hostile and can cause deficient functional wound closure. We report a clinically relevant swine model of biofilm-infected full-thickness chronic wounds. This model is powerful for mechanistic studies as well as for testing interventions.
Biofilm infection is a major contributor to wound chronicity. The establishment of clinically relevant experimental wound biofilm infection requires the involvement of the host immune system. Iterative changes in the host and pathogen during the formation of such clinically relevant biofilm can only occur in vivo. The swine wound model is recognized for its advantages as a powerful pre-clinical model. There are several reported approaches for studying wound biofilms. In vitro and ex vivo systems are deficient in terms of the host immune response. Short-term in vivo studies involve acute responses and, thus, do not allow for biofilm maturation, as is known to occur clinically. The first long-term swine wound biofilm study was reported in 2014. The study recognized that biofilm-infected wounds may close as determined by planimetry, but the skin barrier function of the affected site may fail to be restored. Later, this observation was validated clinically. The concept of functional wound closure was thus born. Wounds closed but deficient in skin barrier function may be viewed as invisible wounds. In this work, we seek to report the methodological details necessary to reproduce the long-term swine model of biofilm-infected severe burn injury, which is clinically relevant and has translational value. This protocol provides detailed guidance on establishing an 8 week wound biofilm infection using P. aeruginosa (PA01). Eight full-thickness burn wounds were created symmetrically on the dorsum of domestic white pigs, which were inoculated with (PA01) at day 3 post-burn; subsequently, noninvasive assessments of the wound healing were conducted at different time points using laser speckle imaging (LSI), high-resolution ultrasound (HUSD), and transepidermal water loss (TEWL). The inoculated burn wounds were covered with a four-layer dressing. Biofilms, as established and confirmed structurally by SEM at day 7 post-inoculation, compromised the functional wound closure. Such an adverse outcome is subject to reversal in response to appropriate interventions.
Biofilm infection complicates burn and chronic wounds and causes chronicity1,2,3,4,5. In microbiology, biofilm mechanisms are primarily studied, with a focus on the microbes1,6. The lessons learned from these studies are of paramount importance from a biological science standpoint but may not necessarily be applicable to clinically relevant pathogenic biofilms6,7,8. Clinically relevant biofilm structural aggregates should include microbial as well as host factors8,9,10. Such a microenvironment allows for the inclusion of host-microbe iterative interactions, which are critical to developing a clinically relevant biofilm7,8. In such a process, the participation of immune cells and blood-borne factors is critical11,12. The host-microbe interactions underlying clinical pathogenic biofilms, as seen in chronic wounds, occur over a long period of time. Thus, any experimental approach aimed at developing a translationally relevant model of biofilm infection must account for these factors. So, we sought to develop a clinically reproducible swine chronic biofilm infection model.
While human studies clearly represent the best approach to studying healing outcomes, often they are not best suited to addressing the underlying mechanisms and new mechanistic paradigms. Ethical concerns limit the use of study designs requiring the collection of multiple biopsies from a chronic wound at different time points. It is, therefore, critical to have a well-established and reproducible animal model to enable invasive studies for the thorough examination of biofilm fate7,13. The selection of an animal model depends on several factors, including scientific/translational relevance and logistics. The porcine system is widely acknowledged to be the most translationally valuable experimental model to study human skin wounds7. Thus, this work reports an established swine model of biofilm-infected full-thickness burn injury. This work is based on several original publications reported in the literature2,7,13,14,15,16,17. In this study, a clinical isolate of multidrug-resistant Pseudomonas aeruginosa (PA01) was chosen to infect the wound. P. aeruginosa is a common cause of wound infections2,18,19,20. It is a Gram-negative bacterium that can be difficult to treat due to its resistance to some antibiotics11,19,21. None of the swine biofilm models reported so far involved 8 week long-term studies22,23,24,25,26. Chronic wounds are those that remain open for 4 weeks or more14,27,28. There are no other chronic wound biofilm models reported in the literature. This work addresses the notion of functional wound closure2,7,13,15,17,29.
All the animal studies were performed in accordance with protocols approved by the Institutional Animal Care and Use Committee (IACUC) #21147. The study was conducted at the Laboratory Animal Resource Center (LARC), Indiana University. We used a female domestic white pig (70-80 lb) in this protocol.
1. Animal acclimatization
2. Surgery room setup
3. Sedation of the pig
4. Induction of anesthesia
5. Animal preparation for burn wounding
6. Antiseptic preparation and marking of the skin burn site
7. Burn wounding procedure
8. Burn wound assessment and imaging
9. Bandaging and dressing
10. Animal recovery and postoperative care
11. Biofilm preparation and inoculation
12. Biopsy collection
13. Euthanasia and tissue collection
A standardized burn device was used to create full-thickness burn wounds at 150 °C for 1 min, which resulted in a homogenous deep burn with a uniform margin of erythema and inflammation (Figure 3 and Figure 7). Each pig received eight full-thickness burn wounds on their back, as depicted in Figure 3C.
The non-invasive real-time assessment of the burn wounds by B-mode high-resolution ultrasound to confirm the wound depth and progression of wound healing over time showed the destruction of all skin layers up to the subcutaneous fat (Figure 4). Laser speckle imaging (LSI) was used for further characterization of the wound perfusion (Figure 4A).
The burn wounds showed a thick pyogenic membrane on the wound surface by day 7 post-inoculation, thus confirming the infection and the establishment of the burn wound biofilm (Figure 7A). Digital planimetry showed an increased wound area at day 3 post-inoculation with PAO1 due to the inflammatory response at the wound site and margins (Figure 7A,B). Although the wound area started to shrink by day 14 post-inoculation, incomplete healing to approximately 25% of the original wound size was observed at day 56, indicating the chronicity of the wounds (Figure 7B). Wound chronicity and impaired wound healing were further confirmed by the TEWL, which showed high transepidermal water loss. The TEWL results reflected the loss of skin barrier function compared to normal skin at all measured the timepoints, thus indicating functional impairment of the burn wound healing (Figure 7B). This was also confirmed by the suppression of the tight junctional proteins ZO-1 and 213 and the impairment of the restoration of skin barrier function, as reflected in the high TEWL values seen at day 35 (mid) and day 56 (late) despite the visual wound closure (Figure 7I).
The burn depth was further validated by H&E staining, which showed distortion and necrosis of all the histological skin layers, as shown in Figure 7C. The established biofilm of PA01 was further validated at day 7 post-inoculation by CFU (Figure 7E,F), SEM imaging (Figure 7G), and immunofluorescence staining (Figure 7H).
Figure 1: Setup for the procedure. (A) Surgical table preparation. (B) Ear vein cannulation for IV fluids and drug administration. (C) Thermal blanket covering to protect the pig from hypothermia during the procedure. (D) Burner and timer setup. Please click here to view a larger version of this figure.
Figure 2: Surgical site sterilization and marking. (A) Hair clipping and sterilization. (B) Marking of the burn site using a sterile eight-wound standard template (each wound is 2 in x 2 in). (C) Final marking using a sterile skin marker. Please click here to view a larger version of this figure.
Figure 3: Burn wound induction. (A,B) Standardized burner with a pressure gauge and automated controller unit (2 in x 2 in) applied to the pre-marked wound site. (C) The whole back showing the eight full-thickness burn wounds. Please click here to view a larger version of this figure.
Figure 4: Noninvasive burn wound imaging and assessment. (A) Laser speckle imaging (LSI) with proper orientation of the laser beam indicator to the center of the wound is shown in the left-side image; the right-side image shows the LSI device and the real-time skin vascular perfusion map. (B) Transepidermal water loss (TEWL) probe application to the wound site at five different spots (four wound corners and the center demonstrated in lower-right corner image) is shown in the left-side image; the right-side image is a representative real-time captured screen of the TEWL measurement. (C) Harmonic ultrasound scanning of the burn wound using a high-resolution 16 MHz ultrasound probe is shown on the left side; the right-side image shows the ultrasound device and the real-time screen recording. (D) Structural (B-mode images, grayscale ultrasound) and biomechanical (elastography, color ultrasound) images of the burn wound site at the inoculation day and day 7 post-inoculation. The wound depth is indicated by the yellow dashed line. Please click here to view a larger version of this figure.
Figure 5: Wound dressing and bandaging. (A) Application of the transparent film dressing for each wound separately. (B) All the dorsal inoculated burn wounds are covered with the first layer of dressing. (C) A larger transparent film dressing is placed over the entire wound area.(D) Application of the second layer of gauze and a loose layer of stretchy elastic bandage around the entire trunk of the pig to absorb any fluid exudate that comes from the wounds. (E) Covering of the entire wound area with a final layer of 4 in adhesive dressing. Please click here to view a larger version of this figure.
Figure 6: Bacterial inoculation. (A) Setup for the Pseudomonas aeruginosa (PA01) inoculation at day 3 post-burn. (B) Topical application of the inoculum with a pipette using a 500 µL volume for each wound. (C) The inoculum is dispersed across the wound surface evenly using a sterile disposable spreader. Please click here to view a larger version of this figure.
Figure 7: Wound healing progress and biofilm confirmation. (A) Representative images of the wound closure over the timeline of the study. Scale bar = 1 cm. (B) Quantitation of the wound area and TEWL measurements over the timeline of the study (n = 6). The data are represented as mean ± SD. N.S. refers to the TEWL value of normal skin. (C) Schematic diagram showing different wound biopsy sites. D. H&E staining with its corresponding Masson's trichrome staining showing distortion and necrosis of all the skin layers at day 3 post-burn and day 7 post-inoculation. Scale bar = 500 µm. (E) Representative digital images of non-selective agar (Luria-Bertani agar) and selective agar (Pseudomonas Isolation Agar) with bacterial colonies grown from porcine wound bed tissue. The selective medium enables the accurate counting of the PA01 colonies only. (F) A sample colony forming unit (CFU) calculation from the colony counts taken from processed day 7 post-inoculation wound biopsies is shown. (G) Representative scanning electron microscopy (SEM) images of the inoculated burn wounds at day 7 post-inoculation showing the established PA01 biofilm, with a zoomed-in image on the right side. Scale bar = 1 µm. The red arrowheads point to extracellular polymeric substances (EPS). (H) P. aeruginosa on the burn wounds were visualized using anti-Pseudomonas (green) antibody; the immunofluorescence images of the day 7 post-inoculation wound biopsies show heavy colonization of the wound tissues by P. aeruginosa. Scale bar = 100 µm. (I) Representative mosaic (scale bar = 200 µm) and corresponding zoomed-in (scale bar = 50 µm) images of ZO-1- and ZO-2-stained sections on day 35 and day 56 post-inoculation, demonstrating reduced expression of the proteins following the induced infection. The OCT-embedded frozen sections (10 µm) were stained using anti-ZO-1 (green) or anti-ZO-2 (green). The sections were counterstained using DAPI. The bar graphs present the quantitation of the ZO-1 and ZO-2 signal intensity. The data are presented as mean ± SD (n = 3); * p < 0.05 compared to spontaneous ones. Mann-Whitney or Kruskal-Wallis one-way analysis of variance tests were performed to test the significance. Figure 7H,I has been modified from Roy et al.13. Please click here to view a larger version of this figure.
This report provides a detailed protocol for setting up a swine model of chronic wound biofilm infection for experimental studies. Several swine biofilm models have been reported previously22,23,24,25,26, but none of them are swine models involving 8 week long-term studies. Chronic wounds are those that remain open for 4 weeks or more14,27,28. There are no other chronic wound biofilm models reported in the literature. This work addresses the notion of functional wound closure2,7,13,15,17,29. A study conducted in 2014 was the first to report that biofilm-infected wounds may close without the restoration of barrier function7. The measurement of the skin barrier function in the healing wound using transepidermal water loss (TEWL) is reported in this work.
Anatomically and physiologically, the porcine skin, compared to the skin of other small animals, is a closer match to the human skin32,33,34. Both pig and human skin has a thick epidermis33, and the dermal-epidermal thickness ratio ranges from 10:1 to 13:1 in pig, which is comparable to humans34,35. Histologically and biomechanically, the skin of humans and pigs shows similarities in the rete-ridges, subdermal fat, dermal collagen, hair distribution, adnexal structures, and blood vessel size and distribution36,37,38. Functionally, both pigs and humans share similarities in the composition of the lipid, protein, and keratin components of the epidermal layer, as well as comparable immunohistological patterns37,38. The porcine immune system, compared to that of other small animals, shares higher similarities with the human immune system, meaning pigs are an appropriate model for studies on the host interactions that are integral to the complexities of the pathological biofilm in wound infections39. The critical assessment of the pros and cons offered by various animal models has led to the consensus that pigs represent an efficient model for studying wound healing34,38. Additionally, domestic pigs spontaneously develop chronic bacterial infections, as observed in humans10. The burn device used to create the wounds is an advanced and automated burn device that delivers heat energy based on a temperature read out from the targeted skin site22,40. Such an approach improves the rigor and reproducibility of the burn injury. The use of human clinical isolates of bacteria to infect the pig wounds adds value as a pre-clinical model.
Burn injuries are complex and cause several systemic perturbations20,41. Thus, it is important to resuscitate the pig with adequate fluids and prevent hypothermia during anesthesia and recovery. Several factors can interfere with the wound healing, including the post-burn nutrition, fluids, and pain42. Close monitoring of the nutrition and pain assessments is, therefore, of importance. Post-burn pain can be severe and influence the animal's behavior and diet. Interventions to address behavioral concerns must be actively considered. Regular and continuous pain scoring and management is imperative. A thorough pain assessment sheet with a very detailed pain management plan is included in this protocol. To avoid cross-contamination between the wounds, special attention should be made to apply the first layer of the dressing on each wound separately. Critical care should be taken in handling all the biohazardous materials and when performing the thorough disinfection of the equipment, tools, and entire surgical room. The application of multiple layers of the dressing prevents the pig from exposing the wounds during their effort to rub or scratch the itching back.
The pig in the current model was not compromised by underlying metabolic disorders (e.g., diabetes), and, therefore, the effect being studied was purely the impact of the bacterial biofilm infection on wound healing. However, the model lends itself to the induction of diabetes (using streptozotocin for example) and could be used to study biofilm infection in relation to an underlying metabolic disorder. The other limitation of the model is the controlled infection setting using P. aeruginosa, a bacterium. It is expected that the normal skin micro-flora of the pig may also be growing in the wound and could impact healing. Further analysis using NGS or other advanced techniques to delineate the microbial content of the wound is necessary. The current model could also be applied to mixed infections with differing microbial species (e.g., fungal, viral, etc.). This is an important element, as clinically relevant wounds are likely to be populated by mixed microbes, which may impact wound healing differentially.
There are many potential advantages in this model, including the similarity to the complexity and long-term sequelae of human chronic wounds, the automated and reproducible burn process, and the use of clinically isolated bacterial species. The use of multiple non-invasive imaging modalities represents a powerful approach for collecting useful physiological data characterizing the wound. Finally, the assessment of the functional wound healing via the restoration of skin barrier function based on TEWL is critical. In conclusion, a robust, simple, detailed, and easy-to-use protocol to develop a biofilm-infected severe burn injury using a porcine model system is shown in this work.
The authors have nothing to disclose.
We would like to thank the Laboratory Animal Resource Center (LARC), Indiana University, for their support and the veterinarian care of the animals during the study. This work was partly supported by the National Institutes of Health grants NR015676, NR013898, and DK125835 and the Department of Defense grant W81XWH-11-2-0142. In addition, this work benefited from the following National Institutes of Health awards: GM077185, GM069589, DK076566, AI097511, and NS42617.
Sedation | |||
Ketamine | Zoetis | 10004027 | 100mg/ml |
Telazol | Zoetis | 106-111 | 100mg/ml |
Xylazine | Pivetal | 04606-6750-02 | 100mg/ml Anased |
3ml syringe w/ 20g needle | Covidien-Monoject | 8881513033 | |
Winged infusion set 21g | Jorgensen Labs | J0454B | |
Anesthetic | |||
Isoflurane | Pivetal | 21295097 | |
Surgery | |||
Hair clippers | Wahl | 8787-450A | |
Nair | Church and Dwight Co. Inc | 70506572 | |
Chlorhexidine Solution | First Priority Inc. | 179925722 | |
70% Isopropyl Alcohol | Uline | S-17474 | |
0.9% Saline Solution | ICU Medical | RL-7282 | |
Non-woven gauze | Pivetal | 21295051 | |
Paper tape | McKesson | 455531 | |
2" Elastic tape | Pivetal | 21300869 | |
18-22g Intravenous Angiocath | SurVet | (01)14806017512306 | |
Spay hook | Jorgensen Labs | J0112A | |
Sterile lube | McKesson | 16-8942 | |
Laryngoscope | Jorgensen Labs | J0449S | |
Roll gauze | Pivetal | 21295032 | |
Endotracheal tube (7-9mm) | Covidien | 86112 | Shiley Hi-Lo Oral Nasal Tracheal Tube Cuffed |
15gtt/ml IV administration set | ICU Medical | 12672-28 | |
LRS 1000ml bag | ICU Medical | 07953-09 | |
Three Quarter Drape Sheet | McKesson | 16-i80-12110G | |
Analgesia | |||
Buprenorphine | RX Generics | 42023-0179-05 | 0.3mg/ml |
Fentanyl Transdermal | |||
Carprofen | 21294548 | Pivetal | 50mg/ml Levafen |
Bandaging | |||
Transparent film dressing 26×30 | Genadyne Biotechnologies | A4-S00F5 | |
Film dressing 4 x 4-3/4 Frame Style | McKesson | 886408 | |
Vetrap | 3M | 1410BK BULK | |
Elastic tape 4" | Pivetal | 21300931 | |
Kerlix Roll Gauze | Cardinal Health | 3324 | |
Imaging | |||
Canon EOS 80D | Canon | 1263C004 | |
Speedlight 600EX II-RT | Canon | 1177C002 | |
EFS 17-55mm Ultrasonic | Canon | 1242B002 | |
GE Logiq E9 | GE | 5197104-2 | |
ML6-15 Probe | GE | 5199103 | |
PeriCamPSI | Perimed | 90-00070 | |
DermaLab | Cortex Technologies Inc | 4608D78 | |
Biopsy/Tissue Collection | |||
6mm punch biopsy | Integra Lifesciences | 33-36 | |
bupivicaine 0.5% | Auromedics Pharma | 55150017030 | |
Size 10 Disposable Scalpel | McKesson | 16-63810 | |
Dissection scissors | Pivetal | 21294806 | |
Rat tooth thumb tissue forceps | Aesculap | BD512R | |
Non-adherent Dressing | Covidien | 2132 | Telfa |
50ml Conical tube | Falcon | 352070 | |
Eppendorf/microcentrifuge tube | Fisherbrand | 02-681-320 | |
OCT Cassette | |||
Non Woven Gauze 4×4 | Pivetal | 21295051 | |
Inoculum | |||
Low salt LB agar | Invitrogen | 22700-025 | |
Low salt LB broth | Fisher scientific | BP1427-500 | |
Petri plate | Falcon | REF-351029 | |
Polyprophyline round bottom tubes (14 ml) | Falcon | REF-352059 | |
Pseudomonas Agar Base (Dehydrated) | Thermo Scientific | OXCM0559B | |
LB Agar, powder (Lennox L agar) | Thermo Fisher Scientific (Life Technologies) | 22700025 | |
Gibco™ DPBS, calcium, magnesium | Gibco | 14040133 | |
Euthanasia | |||
18-22g Intravenous Angiocath | SurVet | (01)14806017512306 | |
Fatal Plus | Vortech Pharmaceuticals | 9373 |