Research design and methods
Animals
Ten-week-old male db/db mice (Jackson Laboratories, Bar Harbor, Maine, USA) were purchased and housed for a week in the Brigham and Women’s Hospital vivarium to become acclimatized prior to surgery performed at the age of 11 weeks. All animal experiments were carried out in accordance and approved by the Institutional Animal Care and Use Committee at Brigham and Women’s Hospital (Protocol number 2016N000249).
Surgical procedure and postsurgical monitoring
A total of 47 11-week-old db/db male mice were included in this study (weight: 45.8±2 g). The grouping of the mice is shown in figure 1C.
Thirty-two mice received an intraperitoneal injection of ATZ (Aldrich Chemistry, St. Louis, Missouri, USA) at 0.5 g/kg body weight (22.9±1 mg). The mice dorsa were then prepared for surgery by undergoing hair removal with Nair. Twenty minutes postinjection a full thickness 1.0×1.0 cm wound including the panniculus carnosus was excised on the dorsum of the mice. MSA (Sigma Lifesciences, St. Louis, Missouri, USA), at a dose of 300 mg/kg body weight (13.7±0.6 mg), was topically applied to the wound, left for 5 min and then washed off with 1.0 cc of saline. The mice were then randomly separated into two groups: Chronic Diabetic and Treated Chronic Diabetic.
The Chronic Diabetic group received only occlusive dressing (Tegaderm HP Transparent Dressing; 3 M Health Care, St. Paul, Minnesota, USA).
The wounds in the Treated Chronic Diabetic group were covered with a 1.0×1.0 cm of Integra Dermal Regeneration Template (online supplementary figure 1) that was sutured in place using a 6–0 Prolene suture (Ethicon, Somerville, New Jersey, USA), and were subsequently covered with an occlusive dressing (figure 1C).
The 15 mice that received no injection served as the Diabetic Control group. They underwent a similar excision and the wounds were covered with an occlusive dressing.
In order to prevent the animals from biting and removing the occlusive dressing, the dressing was sutured in place, at four corners on the mice dorsa, using a 4–0 Prolene suture (Ethicon, Somerville, New Jersey, USA). The occlusive dressing covered the wounds throughout the duration of the experiment, which varied from 21 to 28 days, and was only removed for inspection and photographic purposes. Following injury, the animals were monitored for any signs of distress, including piloerection and increased respiratory rates. If signs of severe systemic illness, such as lethargy or weight loss greater than 20%, the animal would be euthanized. To ameliorate pain, the mice received analgesics (Buprenorphine, 0.05 mg/kg, subcutaneously) immediately after surgery, and then every 8 hours, for 48 hours.
Animals were sacrificed using carbon dioxide asphyxiation and the entire wound as well as 5 mm surrounding the wound, was collected at various time points for detailed histological and immunohistochemical evaluation (online supplementary table 1).
Assessment of glycemic control
Tail venous blood samples were obtained by a 1 mm incision on the tail tip on the day of surgery and on postoperative day 5, 10 and 21 or at a specific time. Blood glucose was then measured using the Antsense II Blood Glucose Analyzer (Bayer-Sankyo, Tokyo, Japan).33 Whole blood was collected on the days of sacrifice (online supplementary table 1) for hemoglobin A1c (HbA1c) assessment. Briefly, approximately 10 µL of blood was collected and 5 µL were added to a DCA 2000 analyzer (Bayer, Elkhart, Indiana, USA) which automatically measured Hb A1c as previously described.34 Unpublished historical glucose and HbA1c data of a wildtype, non-diabetic Control group were used for comparison.
Wound area measurement
The wound, including a ruler, was digitally photographed on days 5, 10, 14 and 28 using a Nikon D3100 SLR camera (online supplementary table 1). Wound area was measured by selecting the border using photographs imported into ImageJ software (V.1.52a; Media Cybernetics, Bethesda, Maryland, USA), under double-blinded conditions. Wound contraction was calculated using the following formula: ((wound area on day n)/(wound area on day 0)) ×100, where n was day 5, 10, 14 or 28. Wound area measurements on day 10 were based on 10 animals from each of the three groups as well as unpublished historical data of a Control (non-diabetic) group.
Microscopy
The harvested tissue was fixed in 10% neutral buffered formalin and stored in 70% ethanol to be embedded in paraffin. Five-micrometer-thick cross-sections were cut through the wound center to ensure inclusion of the wound bed and border and were stained with various stains and antibodies in order to assess different parameters. Quantifications of the parameters were done by two individual assessors under blinded conditions.
Standard staining
Slides were stained with H&E and Masson trichrome (MT) according to standard protocol.
The inflammatory tissue thickness on day 10 was determined by taking three high power field (HPF) photos of the wound bed of each H&E stained cross-section at 20× (n=10 per group). Photos were taken using an Olympus BX53 light microscope (Olympus UCMAD3, T7, Tokyo, Japan). The inflammatory tissue was marked out on each HPF by drawing two borders: Border A placed above the subcutaneous fat and Border B placed on the surface of the wound. In cases of scab formation Border B was drawn underneath the layer of scab. The distance between the two borders was measured in ImageJ.
HPF photos of the wound bed of each MT stained cross-section were taken at 20×. The level of collagen deposition was quantified for day 10, based on an n of 10 per group, using ImageJ by performing image thresholding as per previously defined parameters.26
Immunohistochemistry
Sections were deparaffinized, rehydrated and probed with antibodies for Ki-67 (1:200, Abcam, Cambridge, UK) and CD45 (1:50, R&D Systems, Minneapolis, Minnesota, USA) for 16 hours at 4°C. Alexa 488 goat anti-rabbit IgG (1:200, Invitrogen, Carlsbad, California, USA) and Goat IgG VisUCyte HRP Polymer (R&D Systems, Minneapolis, Minnesota, USA) were used as secondary antibodies for the Ki-67 and CD45 slides, respectively. Finally, the samples were counterstained with hematoxylin and the images acquired.
To quantify cellular proliferation, three HPF photos of the wound bed (n=10 per group) of each Ki-67 stained cross-section were taken at 20×. Ki-67-positive cells were counted and expressed as a ratio of proliferating nuclei to total nuclei on day 10.
To quantify leukocyte infiltration, three HPF photos of the wound bed of each CD45 stained cross-section were taken at 20× per mouse on day 10 (n=10 per group). The images for each section were analyzed using the color deconvolution function on ImageJ (V.1.52a) and the percentage of CD45-positive area per section was calculated. The scaffold layer in the Treated chronic diabetic group was excluded from the analysis.
Immunofluorescence staining
A double immunofluorescence procedure was carried out using anti-CD31 (1:400; Abcam, Cambridge, UK) and anti-α-smooth muscle actin antibody (SMA) (1:400; Abcam, Cambridge, UK) antibodies. The sections were blocked with BSA (5%) for 2 hours and incubated with primary antibody at 4°C overnight. After thorough washing, the sections were incubated with secondary goat anti-rabbit secondary antibodies (CD31, AS1111; α-SMA, AS1110; 1:400; Aspen Biotechnology, Hubei, China) for 1 hour in the dark. Last, the nuclei were stained with 4',6-diamidino-2 phenylindole (Aspen Biotechnology, Hubei, China). Images were captured using the immunofluorescence function of Olympus model BX53 light microscope (Olympus UCMAD3, T7, Tokyo, Japan), and merged using ImageJ (V.1.52a).
Blood vessel number was assessed using the microvessel density (MVD) counting technique, whereby the average number of microvessels per HPF is estimated by counting CD31 stained cells.35
Maturity of new blood vessels can be assessed using the microvascular pericyte coverage index (MPI).36 37 Specifically, we analyzed five HPFs to calculate the percentage of CD31-positive microvessels that stained for both CD31 (endothelial cells) and α-SMA (pericytes). The MPI was expressed as the α-SMA+CD31:CD31 ratio.
Statistical analysis
One-way analysis of variance was performed to detect statistically significant differences between the groups. Data are expressed as mean±SD. Statistical significance was set at a p<0.05. In the case of multiple comparisons, p values were adjusted via Bonferroni-Correction. All statistical analyses and visualization of the results were performed using GraphPad Prism V.8.00 for MacOS (GraphPad Software, La Jolla, California, USA).
Data and resource sharing and availability
The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request. Details on the resources used in this study, including the rodent model, drugs, antibodies, software and hardware, are provided throughout the manuscript.