Amniotic fluid-derived multipotent stromal cells drive diabetic wound healing through modulation of macrophages

Abstract

Background: Cutaneous wounds in patients with diabetes exhibit impaired healing due to physiological impediments and conventional care options are severely limited. Multipotent stromal cells (MSCs) have been touted as a powerful new therapy for diabetic tissue repair owing to their trophic activity and low immunogenicity. However, variations in sources and access are limiting factors for broader adaptation and study of MSC-based therapies. Amniotic fluid presents a relatively unexplored source of MSCs and one with wide availability. Here, we investigate the potential of amniotic fluid-derived multipotent stromal cells (AFMSCs) to restore molecular integrity to diabetic wounds, amend pathology and promote wound healing.

Method: We obtained third trimester amniotic fluid from term cesarean delivery and isolated and expanded MSCs in vitro. We then generated 10 mm wounds in Lepr^db/db diabetic mouse skin, and splinted them open to allow for humanized wound modeling. Immediately after wounding, we applied AFMSCs topically to the sites of injuries on diabetic mice, while media application only, defined as vehicle, served as controls. Post-treatment, we compared healing time and molecular and cellular events of AFMSC-treated, vehicle-treated, untreated diabetic, and non-diabetic wounds. A priori statistical analyses measures determined significance of the data.

Result: Average time to wound closure was approximately 19 days in AFMSC-treated diabetic wounds. This was significantly lower than the vehicle-treated diabetic wounds, which required on average 27.5 days to heal (p < 0.01), and most similar to time of closure in wild type untreated wounds (an average of around 18 days). In addition, AFMSC treatment induced changes in the profiles of macrophage polarizing cytokines, resulting in a change in macrophage composition in the diabetic wound bed. We found no evidence of AFMSC engraftment or biotherapy induced immune response.

Conclusion: Treatment of diabetic wounds using amniotic fluid-derived MSCs encourages cutaneous tissue repair through affecting inflammatory cell behavior in the wound site. Since vehicle-treated diabetic wounds did not demonstrate accelerated healing, we determined that AFMSCs were therapeutic through their paracrine activities. Future studies should be aimed towards validating our observations through further examination of the paracrine potential of AFMSCs. In addition, investigations concerning safety and efficacy of this therapy in clinical trials should be pursued.

Keywords: Amniotic fluid multipotent stromal cells; Cellular therapy; Diabetic wounds.

Fig. 1

Topical application of AFMSCs accelerates diabetic wound healing. a Photographs of WT and diabetic wounds at indicated time points with indicated treatments. b Mean time to wound closure. ****p < 0.001, n ≥ 3. c Percentage remaining wound area over time. Shaded area under the curve represents wound burden. d Wound burden, integrated area under the curve in c, in arbitrary units (a.u). e Wound closure rate. Dashed regression lines fit to each treatment group

Fig. 2

AFMSCs do not induce adverse effects in mouse diabetic wounds. a DiI-stained AFMSCs in culture. b IVIS-generated images of diabetic wounds following topical application of DiI stained AFMSCs, until signal could no longer be detected. Images are representative of n ≥ 3. c Representative H&E image of diabetic wound tissue at post-operative days 2, and 3 after vehicle treatment and AFMSC treatment, n ≥ 3. Scale bar 100 µm

Fig. 3

AFMSC treatment induces molecular change in diabetic wound bed. a Representative H&E tissue sections of diabetic mouse wounds from post-op day 7. Black dashed line indicates the epithelial gap. *p < 0.05. b Quantification of epithelial gaps, n ≥ 3. c Relative gene expression of wound healing factors, with mouse models and treatments as indicated at post-op day 7. *p < 0.05. ****p < 0.0001. ns, not significant, n ≥ 3. d Relative gene expression of IL6. *p < 0.05. **p < 0.01, ns, not significant, n ≥ 3. e Representative images of immunodetection of F4/80 and Arg1 expressing cells in diabetic wound tissues with treatments as indicated. Tissues are counterstained with Ho33342. White arrows, F4/80⁺/Arg1⁺ cells. Asterisk, autofluorescence from erythrocytes