Effect of Canine Adipose Mesenchymal Stem Cell Secretome on a Model of Second-Intention Wound Healing in the Red-Eared Slider Turtle (Trachemys scripta)

Mesenchymal stem cells (MSCs) participate in tissue regeneration primarily through the secretion of bioactive molecules included under the general concept of secretome. This secretome represents a rich and complex mixture of different soluble factors and a variety of extracellular vesicles (Villatoro et al. 2019). Many of these paracrine soluble factors are proteins that are well conserved phylogenetically, such as the vascular endothelial growth factor (Holmes and Zachary 2005), or show varied grades of homology among vertebrates (Secombes et al. 2016). The use of the MSC secretome as a therapeutic element has shown the same efficacy as the implantation of MSCs in different diseases. The secretome has some advantages compared to the use of MSCs, avoiding some of the undesirable effects associated with the use of stem cells including a null oncogenic potential, lack of immunogenic reaction, and less possibility of infection transmission. Furthermore, the secretome is easier to prepare and store, allowing immediate availability and preparation in different dosage forms for clinical use (Vizoso et al. 2017).

Healing pathologies represent an important problem in veterinary medicine, affecting the quality of life of the patient. The use of the secretome of MSCs in skin healing has raised great expectations by demonstrating accelerated wound closure via promoting angiogenesis, suppressing the immune system, and secreting and remodeling the extracellular matrix (Ahangar et al. 2020).

Turtles are of veterinary interest for conservation purposes (also as pets) and, compared with mammals, have shown longer wound healing times when maintained in similar environmental conditions (Negrini et al. 2016); however, their MSCs have not yet been described. Canine adipose MSC secretome (cS-MSC) has been well characterized (Villatoro et al. 2019) and presents considerable opportunities as a therapeutic element (Lombardi et al. 2019). The goal of our study was to evaluate, in a previously described model of secondary-intention skin wound healing (Negrini et al. 2017), the potential biologic activity of cS-MSC on acute experimental skin wounds left to heal by secondary-intention in red-eared slider turtles, Trachemys scripta, maintained in natural temperature conditions and with free access to water.

Our blinded controlled study lasted 63 days and included 14 adult female T. scripta, weight range 1.2–2.9 kg. Turtles belonging to the Municipal Zoological Park of Córdoba, Spain were kept outdoors with free access to water and were fed ad libitum on a pelletized ration (Aquatic Turtle Monster Diet, Zeigler Bros., Inc., Gardners, Pennsylvania, USA). All turtles were identified with microchips and considered healthy based on physical, hematologic, serum biochemical, and fecal examinations.

Secretome from canine adipose MSCs was supplied by the Laboratory of Bioengineering and Tissue Regeneration (University of Málaga, Spain); its characterization has been previously published (Villatoro et al. 2019) and is summarized in Table 1.

The turtles were anesthetized with butorphanol intramuscularly (Torphasol®, Esteve, Ecuphar Veterinaria S.L.U, Barcelona, Spain; 1 mg/kg) plus propofol intravenously (Propovet™, Abbott Laboratories Ltd., Queenborough, UK; 10 mg/kg); one 6-mm full-thickness wound was made on the dorsal aspect of each rear limb using a disposable circular scalpel. On days 1, 7, and 14, 0.3 mL of cS-MSC were administered subcutaneously around the test wounds with a 25 ga needle. The opposite wound was used as a control, with lactated Ringer's solution (B. Braun VetCare, Barcelona, Spain) applied by the same procedure and investigator. After each application, the turtles were kept out of the water for a period of 1 h. Wounds were allocated randomly by tossing a coin. Clinical and histopathologic evaluations were done by an investigator blind to the treatment. All animal procedures were approved by University of Córdoba Committee on Animal Research and Ethics (project 2017PI/22).

At days 7, 14, 21, and 28 wounds were evaluated clinically and photographed including an internal scale. The wound area was measured digitally (ImageJ, US National Institutes of Health, Bethesda, Maryland, USA) and wound retraction was expressed as the percentage of the initial wound area. From day 21 onward, test and control wounds were compared according to depth of the defect, inflammation, crust consistency, and time to crust peeling off, and were categorized as equal, better, or worse. Complete wounds' biopsies from three randomly chosen animals were obtained under general anesthesia at days 21, 28, 42, and 63. Biopsy sites were closed with 2-0 USP absorbable monofilament thread (B. Braun VetCare). At the end of the study, all turtles were released to their original pond; two animals not biopsied were included to achieve statistical power for comparison of wound retraction. Tissue samples were processed by routine histologic procedures. Histopathologic outcomes were inflammatory response, granulation tissue production and maturation, collagen remodeling, and re-epithelialization (semiquantitative analysis).

Data distribution was analyzed by the Kolmogorov-Smirnov test. The mean retraction of test and control wounds was compared at each time point by a paired t-test in Prism 5.04 software for Windows (GraphPad Software 2010). A value of P<0.05 was considered significant. Sample size was determined based on the SD of a previous experiment (Negrini et al. 2017) to achieve a 95% power to detect mean differences of 25% in wound retraction using GraphPad StatMate 2.00 for Windows (GraphPad Software 2012).

The behavior and general physical condition of the animals were not affected during the study. Clinically, the main variable assessed was wound retraction because it is a numeric variable suitable to evaluate treatment efficacy. Mean wound retraction in the test and control wounds at every time point was statistically nonsignificant (Table 2). Clinical evaluation of the wounds made by two investigators blind to the treatment resulted in a better score to secretome-treated wounds in 43% and 57% of the animals respectively, with only two discrepant results.

Microscopically, there were no significant differences between treated and control wound healing. Overall, in 3/12 treated wounds, healing was improved compared to their internal (same animal) control, whereas 9/12 showed a similar healing stage to that of the controls. No adverse reaction was observed in wounds attributable to cS-MSC application. At days 21 and 28, in one third and two thirds of turtles, respectively, the granulation tissue was more abundant and collagen fibers were better organized in cS-MSC-treated wounds than in controls. At days 42 and 63, the healing features were similar in all wounds samples (Fig. 1).

In our study, neither objective outcomes, such as wound retraction nor the clinical and histopathologic subjective evaluations, detected any significant functional activity of cS-MSC on wound healing of T. scripta. This experimental excisional wound model has previously been shown to be sensitive enough to demonstrate that porcine insulin significantly increased wound retraction and modulated the inflammatory response of T. scripta (Negrini et al. 2017), and in the present study showed a high degree of agreement between the assessments of researchers. Previous studies have demonstrated xenogeneic biologic activity of several components of the secretome (Secombes et al. 2016); however, the lack of biologic activity in this case may be due to the phylogenic distance of some of its components of the secretome between mammals and reptiles. Future work using concentrated secretome with higher concentrations of the components might produce different results by developing different mechanisms of action. In conclusion, we were unable to detect significant biologic activity of cS-MSC, at the dose used, in the healing of cutaneous wounds by second-intention of T. scripta.

This work was funded by the research groups PAIDI AGR262 and BIO307, Junta de Andalucía, Spain. The authors declare that there is no conflict of interests.