Accelerated Wound Healing in Minipigs by On-Site Production and Delivery of CXCL12 by Transformed Lactic Acid Bacteria

Abstract

Non-healing wounds are a growing medical problem and result in considerable suffering. The lack of pharmaceutical treatment options reflects the multistep wound healing process, and the complexity of both translation and assessment of treatment efficacy. We previously demonstrated accelerated healing of full-thickness wounds in mice following topical application of the probiotic bacteria Limosilactobacillus reuteri R2LC transformed to express CXCL12. In this study, safety and biological effects of a freeze-dried formulation of CXCL12-producing L. reuteri (ILP100) were investigated in induced full-thickness wounds in minipigs, and different wound healing evaluation methods (macroscopic, planimetry, 2D-photographs, 3D-scanning, ultrasound) were compared. We found that treatment with ILP100 was safe and accelerated healing, as granulation tissue filled wound cavities 1 day faster in treated compared to untreated/placebo-treated wounds. Furthermore, evaluation using planimetry resulted in 1.5 days faster healing than using 2D photographs of the same wounds, whereas the areas measured using 2D photographs were smaller compared to those obtained from 3D scans accounting for surface curvatures, whereas ultrasound imaging enabled detailed detection of thin epithelial layers. In conclusion, topical administration of the drug candidate ILP100 warrants further clinical development as it was proven to be safe and to accelerate healing using different evaluation methods in minipigs.

Keywords: Limosilactobacillus reuteri; active wound care; chemokines; three-dimensional imaging; topical; wound measurement.

Conflict of interest statement

Ilya Pharma AB is a for-profit SME developing next generation immunotherapies to treat wounds in skin and mucosa. E.V., M.P. and S.R. are co-founders of Ilya Pharma, and MP and SR support the company as experts. E.Ö., H.L. and P.F. are employees at Ilya Pharma.

Figures

Figure A1

Wound healing assessed from 2D photographs for Cohort A. Wound area and wound healing over time were measured from 2D photographs in Cohort A (A,B), and days to 100% (C), 75% (D), and 50% (E) re-epithelization were assessed. (A,B): Untreated (n = 3, N = 18), Wild type R2LC (n = 3, N = 12), ILP100 (n = 5, N = 26). (C–E): Untreated (n = 3, N = 18), Wild type R2LC (n = 3, N = 11), ILP100 (n = 5, N = 25). Kruskal-Wallis test with Dunnett’s multiple comparison test (A–E). * p < 0.05.

Figure A2

Wound healing assessed from 2D photographs for Cohort B. Wound area and wound healing over time measured from 2D photographs in Cohort B (A,B). From these measurements, days to 100% (C), 75% (D), and 50% (E) re-epithelization were assessed. (A,B); Placebo (n = 3, N = 18), ILP100 (n = 6, N = 36). (C–E): Placebo (n = 3, N = 12–17), ILP100 (n = 6, N = 23–35). Mann-Whitney (A–E). * p < 0.05, **** p < 0.0001.

Figure 1

Schematic illustration of the protocol design. The study consists of two cohorts: one in males (Cohort A, blue box) and one in females (Cohort B, orange box). In the illustration, the numbers in the filled blue (Cohort A) and orange (Cohort B) boxes indicate the days where 2D photographs, macroscopic evaluation, treatment, and planimetry (only Cohort A) were performed. No 2D photographs, macroscopic evaluation, or planimetry were performed on the days in parenthesis. Time points for wound induction, collection of blood (for hematology, clinical chemistry, plasma levels of CXCL12 and SppIP, and for CFU counts of ILP100), collection of urine samples, and for additional imaging (only Cohort B) with ultrasound and 3D scanning are indicated with arrowed symbols.

Figure 2

Comparisons of different wound assessment methods. Wound areas from Cohort A were measured by on-site planimetry (A) and from 2D photographs (B), n = 18, N = 93–96), and the areas were compared to reveal method-dependent differences. From the planimetric data, the areas of newly formed epithelia, granulation tissue, and unspecific tissue for all wounds in Cohort A (C) were retrieved (n = 18, N = 96). Furthermore, days to 100% (n = 18, N = 96), 75% (n = 18, N = 91), and 50% (n = 18, N = 91) re-epithelialisation of all wounds were assessed with planimetry and from 2D photographs (D–F) in Cohort A. In Cohort B, method-dependent differences between area measured with 3D scanning and from 2D photographs on day 9 (G), n = 17, N = 42), and day 28 (H), n = 6, N = 17) were assessed. In addition, the formation of granulation tissue was assessed macroscopically by scoring from 0 to 4 (scale bar 1 cm) (I). To assess correlations between macroscopic evaluations and volume measurements from the 3D scans at day 9, granulation score (J) and hypergranulation score (K) was plotted against volumes generated from the 3D scans of the respective wound (n = 17, N = 42), where red dots indicate clot formation in the wound. On days 2, 9, and 28, the wounds in Cohort B were imaged with ultrasound and representative images are shown (L), where the yellow lines delineate measured diameters and depth showing that the wounds are fully epithelialized and considered healed at d28. Method-dependent differences were assessed by comparing the diameter of each wound measured from images acquired by ultrasound or 3D scanning day 2 (M) and by ultra-ound, 3D scanning, or from 2D photographs on day 9 (N). Statistic comparisons were made using Wilcoxon signed rank test (A,D–F,H,I,M), Jonckheere’s trend test (J) and Friedmans test with Dunn’s multiple comparison test (N) (* p < 0.05, ** p < 0.01, **** p < 0.0001).

Figure 3

Wound healing assessed by planimetry and macroscopic evaluation in cohort A. (A) shows representative photographs of healing over time where the inserted numbers depict days post-induction and the scale bar corresponds to 1 cm. Wound healing was assessed by planimetry detecting wound area (B), % of the wound being re-epithelized (C), and area of newly formed epithelia over time (D). Days to 100% (E), 75% (F), and 50% (G) re-epithelization of the wounds were also assessed. In addition, granulation tissue formation was measured as wound area covered by granulation tissue (H) and days to 100% of the wound area being covered by granulation tissue (I) using planimetry, as well as macroscopically as granulation score (0–4) day 3 to 11 (J) and days to granulation score 4 (K). The animals and wounds were divided into the following groups: Untreated (n = 3, N = 18), Wild type R2LC (n = 3, N = 12), ILP100 (n = 5, N = 26). Statistic comparisons between groups were made using Kruskal-Wallis test with Dunn’s multiple comparison test (B–K). For (B–D,H,J) one test was performed for each time point. * Indicates the difference between untreated and ILP100 and # difference between untreated and wild type R2LC. (* p < 0.05, ** p < 0.01, *** p < 0.001).

Figure 4

Wound healing assessed by macroscopic evaluation and from 3D scans in Cohort B. Wounds in Cohort B were macroscopically evaluated for granulation tissue formation by granulation score day 3 to 11 (A) and days to granulation score 4. ** p < 0.01, *** p < 0.001 (B) (Placebo: n = 3, N = 18, ILP100: n = 6, N = 36). The healing was also assessed using 3D scans, where panel (C) shows representative projections of 3D scans from day 3, day 9, and day 28. From the 3D scans, wound area, wound depth, and wound volume were measured (D–F) for placebo and ILP100-treated wounds day 9 (Placebo: n = 3, N = 9, ILP100: n = 6, N = 18). On day 28 when the wounds were healed, scar area, scar height, and scar volume (G–I) were measured from 3D scans, (Placebo: n = 3, N = 9, ILP100: n = 5–6, N = 14–18). Statistic comparison between the groups were made using Mann-Whitney tests (A,B,D–I).