Identification of Specific miRNAs in Neutrophils of Type 2 Diabetic Mice: Overexpression of miRNA-129-2-3p Accelerates Diabetic Wound Healing

Diabetes can delay the healing of wounds and cause complications such as foot ulcers (1). Effective tissue repair requires the recruitment of immune cells from bone marrow (BM) to injured sites. In chronic wounds, the continuous influx of neutrophils and macrophages to the wound site can be maintained by stimuli such as tissue hypoxia, bacterial components, foreign bodies, and fragments of necrotic tissue (2). Chronic inflammation is predominantly characterized by excessive and prolonged infiltration of neutrophils and macrophages (3), which is frequently in diabetic foot ulcers (4).

Skin tissue repair consists of three phases: inflammation, proliferation/migration, and maturation/resolution. Previously, we showed that the inflammatory phase is aberrant in diabetes, and the numbers of myeloid cells, including monocytes, granulocytes, and precursors, in cutaneous wounds were shown to be significantly raised on day 2 after wounding (D2W), D7W, and D10W in diabetic mice (db) compared with those in control mice (non-db); moreover, the recruitment and/or accumulation kinetics of these cells were altered (5). In many bacterial and autoimmune inflammatory diseases, one of the most important mechanisms of neutrophil accumulation is a delay in apoptosis as a result of the excessive production of neutrophil survival factors (6); thus, inflammation is prolonged followed by tissue damage, a feature associated with chronic inflammation in various diseases. Intrinsic factors have been shown to play an important role in aberrant myeloid cell behavior (7). Accordingly, the pathogenesis of chronic inflammation in diabetic foot ulcers may be due to intrinsic defects of diabetic-derived neutrophils. To promote diabetic skin wound healing, the mechanism behind such chronic inflammation therefore should be elucidated.

miRNAs are small noncoding RNAs of ∼21–25 nucleotides in length; they regulate posttranscriptional expression through binding to the 3′-untranslated region (3′-UTR) of target mRNAs (8,9). Reports have described that miRNAs have an important function in several diseases (10) and wound healing, and they have been shown to comprehensively regulate a number of important biological processes within the cell (11,12). Specifically, miRNAs play key roles in diseases such as diabetes and cancer and in chronic wounds and are associated with cell migration, proliferation, invasion, and apoptosis. miR-126 overexpression has been shown to rescue the diabetes-induced impairment of phagocytosis of apoptotic cardiomyocytes (13). In addition, Let-7b was revealed to inhibit keratinocyte migration in cutaneous wound healing (14). Moreover, reports have described that the topical application of miR-132 mimic mixed with pluronic F-127 gel in chronic wounds promotes reepithelialization (15) and that miR-27b rescues impaired BM-derived angiogenic cell function and improved wound healing in type 2 diabetic mice (16). miR-191 modulates cellular migration and angiogenesis to delay the tissue repair process (17). We also reported that miR-142 is required for the clearance of Staphylococcus aureus at skin wound sites (18). Against this background, functional analysis of miRNAs in the complex process of wound healing could confer great benefits for manipulation in the clinic, but the exact molecular mechanisms involved in diabetic skin wound healing leading to chronic inflammation remain largely unknown.

We hypothesized that miRNAs are involved in the functional regulation of diabetic-derived neutrophils in chronic inflammation. To clarify the molecular mechanism of inflammatory control in diabetic-derived neutrophils, we used microarrays to screen for changes of miRNA expression. Next, we evaluated the expression of a specific miRNA and its target genes in diabetic-derived neutrophils and/or skin wounds. Finally, we examined the involvement of this miRNA in diabetic skin wound healing.

The animal care committee of Nagasaki University approved the protocol for this study (approval number 1407101159). BKS.Cg-Dock7^m +/+ Lepr^db/J (+Lepr^db/+Lepr^db and Dock7^m+/+Lepr^db) mice (5 weeks old) were purchased from Charles River Laboratories (Yokohama, Japan). They were housed under a 12/12-h light/dark cycle (lights on at 0700 h, lights off at 1900 h) at constant temperature and humidity and allowed free access to food and water. Male mice between 8 and 12 weeks of age were used and were age matched to controls. To eliminate the effect of hormonal action related to sexual maturation on skin wound healing, we used only male mice. Full-thickness excisional dorsal wounds (4 mm) were made by a biopsy punch. Wounds were harvested, including a 2-mm skin margin.

BM was flushed from femurs and tibiae. Neutrophils from the pooled BM of three db or three non-db mice were isolated using a neutrophil isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany), and miRNA was purified from neutrophils using a microRNA isolation kit, Mouse Ago2 (Wako, Osaka, Japan), for microarray analysis.

Microarray analysis was performed on a total of eight pools (four pools of three db BM samples and four pools of three non-db BM samples) using SurePrint G3 Mouse miRNA Microarray (Agilent Technologies, Tokyo, Japan) in accordance with the manufacturer’s instructions. Bioinformatic analyses were performed using GeneSpring 13.0 software (Agilent Technologies). The data discussed in this article have been deposited in the National Center for Biotechnology Information Gene Expression Omnibus and are accessible through GEO Series accession number GSE100577.

Neutrophils from BM of six non-db mice were isolated using a neutrophil isolation kit (Miltenyi Biotec). Macrophages, T cells, and B cells from BM of six non-db mice were isolated with a MicroBead Kit (Miltenyi Biotec) in accordance with the manufacturer’s instructions. Cells were incubated with anti-CD11b antibody (Ab), anti-CD5 Ab, and anti-CD19 Ab to isolate macrophages, T cells, and B cells, respectively.

Neutrophils from BM and skin wounds of six db and six non-db mice were isolated using a neutrophil isolation kit and Anti-Ly-6G MicroBead Kit (Miltenyi Biotec). Wound skin from D2W and D3W was harvested by a biopsy punch (6 mm). It was then dissolved in QIAzol Lysis Reagent (QIAGEN, Germantown, MD). Total RNA, including miRNA, was extracted using miRNeasy Mini Kit and RNeasy Mini Kit (QIAGEN) in accordance with the manufacturer’s instructions. Total RNA was quantified using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA). RNA samples were stored at −80°C until use.

Total RNA (850 ng) was used as a template, and cDNA was synthesized using PrimeScript RT Reagent Kit for mRNA expression analysis (Takara Bio, Kusatsu, Japan) in accordance with the manufacturer’s instructions. Quantitative real-time PCR (qRT-PCR) was performed in a 10-μL reaction system using SYBR Premix Ex Taq (Takara Bio) and a Thermal Cycler Dice Real Time System (Takara Bio). The contents of the amplification mix and thermal cycling conditions were set in accordance with the manufacturer’s instructions. The following primers were purchased from Takara Bio: caspase 6 (Casp6), chemokine (C-C motif) receptor 1-like 1 (Ccr1l1), chemokine (C-C motif) receptor 2 (Ccr2), caspase 8 (Casp8), death effector domain-containing DNA binding protein 2 (Dedd2), Gapdh, CCR2, DEDD2, and GAPDH. TaqMan Gene Expression Assay (Thermo Fisher Scientific) for CASP6 was performed in accordance with the manufacturer’s instructions.

Total RNA (10 ng) was used as a template, and cDNA synthesis and qRT-PCR were performed using miRCURY LNA Universal RT microRNA PCR and LNA PCR primer set for miRNA expression analysis (Exiqon, Vedbæk, Denmark). Primers mmu- (hsa-) miR-129-2-3p and 5S rRNA were purchased from Exiqon. The relative quantification of mRNA transcripts and miRNA was performed using the ∆∆Ct method (19).

Putative target genes of miR-129-2-3p were predicted using GeneSpring software. DNA synthesis of Casp6, Ccr2, and Dedd2 was performed by Hokkaido System Science Co. (Sapporo, Japan). Luciferase reporter plasmids were constructed to confirm the regulation of target genes by miR-129-2-3p. miR-129-2-3p mimic (chemically synthesized double-stranded mature miR-129-2-3p) and mutation as a negative control were chemically synthesized by GeneDesign (Ibaraki, Osaka, Japan).

3T3 cells were cultured for luciferase reporter assay in DMEM (Wako) with high glucose, l-glutamine, 10% FBS, and 1% penicillin-streptomycin. These cells were then harvested, seeded onto a 96-well plate at ∼3.0 × 10⁴ cells per well in DMEM with 10% FBS without 1% penicillin-streptomycin, and cultured for 24 h. Subsequently, these cells were washed with Opti-MEM (Thermo Fisher Scientific), supplemented with 100 μL Opti-MEM in each well, and incubated at 37°C before transfection.

The 3′-UTRs of miR-129-2-3p targets were predicted using TargetScan (http://www.targetscan.org/vert_71) and microT-CDS in DIANA TOOLS (http://diana.imis.athena-innovation.gr/DianaTools/index.php?r=microT_CDS/index). Vectors were constructed with pmirGLO Dual-Luciferase miRNA Target Expression Vector (Promega, Madison, WI) in accordance with the manufacturer’s instructions. Primers consisting of the 3′-UTRs of predicted miR-129-2-3p target sequences and appropriate restriction sites were synthesized, annealed, and cloned downstream of the firefly luciferase reporter (luc2) gene in pmirGLO. Sequences were as follows: Casp6 sense 5′-aaacTGTTGGACGTGGTGGAAGGGCTAt-3′ Casp6 antisense 5′-ctagaTAGCCCTTCCACCACGTCCAACAgttt-3′, Ccr2 sense 5′-aaacAGTGATTAGACTAAAAATAATAAGGGCTt-3′, Ccr2 antisense 5′-ctagaAGCCCTTATTATTTTTAGTCTAATCACTgttt-3′, Dedd2 sense 5′-aaacCTGCCCCACACACTTTAG-CCTAAGGGCTAt-3′, and Dedd2 antisense 5′-ctagaTAGCC-CTTAGGCTAAAGTGTGTGGGGCAGgttt-3′. Uppercase and lowercase letters indicate the 3′-UTR and restriction sites (PmeI and XbaI), respectively.

Sequences of miR-129-2-3p mimic and mutation of seed sequence as a negative control were as follows: miR-129–2-3p mimic 5′-AAGCCCUUACCCCAAAAAGCAU-3′ and miR-129–2-3p mutation 5′-AAUCCCCUACCCCAAAAAGCAU-3′. 3T3 cells (3.0 × 10⁴ cells/100 μL) were cotransfected with the miR-129-2-3p mimic or mutation and a reporter plasmid containing the 3′-UTR of Casp6, Ccr2, and Dedd2. The miR-129-2-3p mimic and mutation were added at a final concentration of 40 nmol/L with Lipofectamine 3000 (Invitrogen). At 48 h after transfection, luciferase activity was assessed using a Dual-Glo Luciferase Assay System (Promega) in accordance with the manufacturer’s instructions.

The HL-60 human promyelocyte cell line (RBRC-RCB0041) was provided by RIKEN BioResource Research Center through the National BioResource Project of the Ministry of Education, Culture, Sports, Science and Technology, Japan. The HL-60 cells were cultured in RPMI medium (Thermo Fisher Scientific) with 10% FBS and 1% penicillin-streptomycin. Neutrophil differentiation was induced by exposing HL-60 cells to 1.3% DMSO for 5 days. At 3 days after exposure to 1.3% DMSO, HL-60 cells were transfected with miR-129-2-3p mimic or mutation at a final concentration of 70 nmol/L using Lipofectamine 3000 in Opti-MEM with 1.3% DMSO. Two days after transfection, cells and conditioned media were harvested and applied to target gene expression analysis.

Harvested tissues (D0W–D3W) were fixed in 4% paraformaldehyde overnight and embedded in paraffin. All specimens were cut into 4-μm-thick sections. Immunohistochemistry (IHC) was performed with anti-Ly-6G (α-Ly-6G, neutrophil marker) as a primary antibody in accordance with the manufacturer’s protocol (Abcam, Cambridge, U.K.). Rat anti-mouse Ly-6G was purchased from Novus Biologicals (Littleton, CO). Samples were incubated with N-Histofine Simple Stain Mouse MAX PO (Rat) (Nichirei Bioscience, Tokyo, Japan) for 1 h at room temperature. The Histofine DAB-3S kit (Nichirei Bioscience) was used as a color developer. Hematoxylin was used as a nuclear counterstain. Observations were made through Aperio AT Turbo (Leica Microsystems, Tokyo, Japan).

Double-label fluorescent IHC (fIHC) was performed with CASP6 (dilution 1:500) (ab185645; Abcam) or CCR2 (dilution 1:100) (ab203128; Abcam) and anti-Ly-6G (α-Ly-6G, neutrophil marker, dilution 1:200) as primary Abs in accordance with IHC and immunofluorescence protocols (Abcam). Goat anti-rabbit IgG (H+L), Alexa Fluor 488–conjugated (α-rabbit 488) and Alexa Fluor 594–­conjugated (α-rat 594) secondary Abs (dilution 1:500) were purchased from Life Technologies. Observations were made through confocal microscopy (C2+ system; Nikon, Tokyo, Japan). NIS-Elements C software version 4.13 (Nikon) and IMARIS 7.6.5 (Bitplane, Zurich, Switzerland) were used for data analysis.

For each tissue image of three to five wounds in the wound area for IHC, binarization was performed. Ratios of the neutrophil-positive area relative to the wound area were calculated.

In situ hybridization (ISH) was performed using miRNA ISH buffer set and miRCURY LNA Detection 3′ and 5′ DIG-labeled probes (QIAGEN) in accordance with the manufacturer’s instructions. In brief, 4% paraformaldehyde perfusion–fixed tissues were embedded in paraffin. Six-micrometer-thick sections were deparaffinized and incubated with proteinase K solution (DAKO, Glostrup, Denmark) for 10 min at 37°C. After washing in PBS, sections were dehydrated. Hybridization was performed using 20 nmol/L miRNA probe in miRNA ISH buffer (QIAGEN) at 50°C for 3 h. Sections were rinsed in 5× saline sodium citrate (SSC) at 50°C for 5 min, twice with 1× SSC at 50°C for 5 min, twice with 0.2× SSC at 50°C for 5 min, and with 0.2× SSC at room temperature for 5 min. Sections were treated with blocking solution (Nacalai Tesque, Kyoto, Japan) for 15 min at room temperature and then incubated with anti-DIG Ab (1:800) (Roche Diagnostics, Mannheim, Germany) in blocking solution overnight at 4°C. Sections were developed using nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl-phosphate (Roche Diagnostics) at 30°C. Observations were made through Aperio AT Turbo and confocal (C2+ system) microscopy. NIS-Elements C software version 4.13 was used for data analysis.

Skin tissue was homogenized using TissueLyser II (QIAGEN). T-PER Tissue Protein Extraction Reagent (Thermo Fisher Scientific), which consists of proteinase and dephosphorylation inhibitor, was then added. Debris was removed from the supernatant using an Ultrafree-MC 0.45-mm filter (Merck Millipore, Darmstadt, Germany). Filtered protein samples were quantified using a Direct Detect Spectrometer (Merck Millipore), separated on 4–12% NuPAGE Novex Bis-Tris Gels (Thermo Fisher Scientific), transferred to polyvinylidene difluoride (PVDF) membranes, and blotted in accordance with standard protocols (Ab details are listed in Table 1). Protein bands were visualized using ImmunoStar LD (Wako), and band intensity was calculated using MultiGauge version 3.1 (Fujifilm, Tokyo, Japan).

For in vivo experiments, miR-129-2-3p mimic or mutation as a negative control (10 μmol/L in 50 μL of 30% pluronic F-127 gel [Sigma-Aldrich, St. Louis, MO]) was applied topically immediately after wounding. Thereafter, the proportion of wound area on each day after wounding relative to the initial wound area was measured using Adobe Photoshop CC.

Data are shown as mean ± SD. The statistical significance of differences between means was assessed by Mann-Whitney U test, one-way ANOVA followed by Tukey multiple comparison test, and two-way ANOVA followed by Bonferroni posttest to compare replicate means (GraphPad Software, San Diego, CA). P < 0.05 was considered significant.

Microarray analysis showed that the expression levels of 22 miRNAs in diabetic-derived neutrophils were more than double those in nondiabetic-derived neutrophils (Supplementary Fig. 1A), whereas those of 80 miRNAs were decreased to less than one-half in fold change analysis (Supplementary Fig. 1B), but statistical analysis was not performed on this.

In turn, we performed a moderated t test (cutoff <0.05) and Storey bootstrapping for microarray data using GeneSpring software. Microarray analysis showed that the expression levels of 10 miRNAs in diabetic-derived neutrophils were significantly decreased compared with those in nondiabetic-derived neutrophils (Fig. 1A and Table 2). We focused here on miR-129-2-3p because the microarray data indicated that the signal values of db in miRNAs were too low, with the exception of miR-129-2-3p (Fig. 1B), and that miR-129-2-3p in diabetic-derived neutrophils was downregulated to less than one-third of the level in nondiabetic-derived neutrophils (Table 2).

qRT-PCR using the SYBR Green I assay showed that the expression of miR-129-2-3p in diabetic-derived neutrophils was significantly decreased (expression level mean ± SD: non-db 6.3 ± 4.4, db 0.70 ± 0.35; P = 0.0079) (Fig. 1C).

To examine the cellular expression of miR-129-2-3p, we isolated neutrophils, macrophages, B cells, and T cells from BM and spleen in non-db using MicroBead Kit and examined the expression of miR-129-2-3p using qRT-PCR; miR-129-2-3p expression was significantly increased in neutrophils compared with that in other cells in BM (Fig. 1D) and was increased in neutrophils and macrophages compared with the levels in B cells and T cells in the spleen (Supplementary Fig. 2). Accordingly, this finding suggests that miR-129-2-3p is related to inflammation, especially early inflammation.

More than 800 mRNAs were predicted to be targets of the miRNAs listed in Table 2 that were significantly differentially expressed in diabetic-derived neutrophils (Supplementary Fig. 3). Thus, we performed gene ontology (GO) and pathway analyses using GeneSpring software to survey them. The results showed that candidate target mRNAs for miR-129-2-3p are associated with many biological processes and pathways, including inflammatory response, apoptosis, chemotaxis, phagocytosis, endocytosis, and chemokine signaling. Accordingly, a number of biological processes may be defective in diabetic-derived neutrophils (Table 3).

GO and pathway analyses showed that Casp6, Ccr1l1, and Ccr2 are associated with inflammatory responses and that Casp8 and Dedd2 also are involved in apoptosis. These genes were expressed at high levels in diabetic-derived neutrophils compared with their levels in nondiabetic-derived neutrophils, as confirmed by qRT-PCR (expression levels mean ± SD: Casp6: non-db 0.76 ± 0.13, db 0.97 ± 0.079 [P = 0.0042]; Ccr1l1: non-db 0.33 ± 0.22, db 1.0 ± 0.23 [P = 0.0023]; Ccr2: non-db 0.37 ± 0.061, db 1.4 ± 0.33 [P = 0.0022]; Casp8: non-db 0.88 ± 0.074, db 1.2 ± 0.16 [P = 0.0077]; Dedd2: non-db 0.60 ± 0.13, db 0.89 ± 0.086 [P = 0.0043]) (Fig. 2). These results support the prediction that these mRNAs are targets of miR-129-2-3p because the expression of these genes was significantly increased, whereas the expression of miR-129-2-3p was significantly decreased in diabetic-derived neutrophils.

To verify that the mRNAs that we identified were targets of miR-129-2-3p, we tested each in a luciferase reporter assay. miR-129-2-3p is predicted to bind with high affinity to Casp6, Ccr2, and Dedd2 3′-UTRs (Fig. 3A). In this assay, a decrease in luciferase activity indicates the binding of the miRNA mimic to the 3′-UTR of the target sequence. Luciferase reporter assays showed that the miR-129-2-3p mimic could effectively inhibit the expression of Casp6 (control 1.0 ± 0.15, mimic 0.61 ± 0.13; P = 0.0022), Ccr2 (control 1.0 ± 0.083, mimic 0.73 ± 0.10; P = 0.029), and Dedd2 (control 1.0 ± 0.13, mimic 0.69 ± 0.089; P = 0.0079) (Fig. 3B–D); thus, we concluded that miR-129-2-3p directly regulates the expression of Casp6, Ccr2, and Dedd2 in vitro.

To determine whether CASP6, CCR2, and DEDD2 can be direct targets of miR-129-2-3p, we used HL-60 cells, which are human neutrophil-like cells. qRT-PCR showed that the expression of CASP6 and DEDD2 in HL-60 cells transfected with miR-129-2-3p mimic was significantly decreased compared with that in those transfected with mutant miR-129-2-3p (expression levels mean ± SD: CASP6: mutant 1.0 ± 0.23, mimic 0.71 ± 0.050 [P = 0.029]; DEDD2: mutant 1.1 ± 0.30, mimic 0.57 ± 0.12 [P = 0.017]) (Fig. 3E and F), although the expression of CCR2 could not be detected in HL-60 cells. These results suggest that these target genes might be directly regulated by miR-129-2-3p in HL-60 cells. Therefore, further investigation of these genes is necessary using human diabetic wound samples to apply the these findings to diabetes in humans.

To investigate the proportion of neutrophils among cells present at the early stage of inflammation (D1W and D2W), we performed IHC at wound sites in db and non-db mice. IHC for neutrophils showed a stronger signal in db D2W compared with non-db (Fig. 4A). Moreover, we calculated ratios of neutrophil-positive area relative to the wound area at D1W and D2W. The results showed that there were significantly more neutrophils present in D2W of db (signal level mean ± SD: non-db 4.5 ± 0.97, db 9.01 ± 1.3; P = 0.029) (Fig. 4B).

To determine which cells express miR-129-2-3p during the early stage of inflammation, we performed ISH in D1W of non-db. ISH showed that miR-129-2-3p was predominantly expressed in wound-infiltrating neutrophils in D1W (Fig. 4C).

To elucidate whether miR-129-2-3p and its target genes are involved in prolonged inflammation and delayed wound healing, we examined the expression of these genes at the skin wound site in D2W.

The expression of miR-129-2-3p did not show a significant difference between non-db and db at D2W (data not shown) upon isolating the neutrophils from non-db and db wound skin at D2W using Anti-Ly-6G MicroBead Kit and examining the expression of miR-129-2-3p in equal numbers of neutrophils from the two groups. The expression of miR-129-2-3p in neutrophils from db D2W was significantly decreased compared with that in non-db (expression level mean ± SD: non-db 7.0 ± 2.1, db 2.2 ± 0.69; P = 0.016) (Fig. 4D). The expression of the miR-129-2-3p target gene Casp6 in db D2W was significantly increased compared with that in non-db (expression level mean ± SD: non-db 2.1 ± 0.24, db 2.6 ± 0.082; P = 0.0029) (Fig. 4E). Similarly, the expression of Ccr2 in db D2W was significantly increased compared with the level in non-db (expression level mean ± SD: non-db 4.4 ± 0.55, db 5.7 ± 0.34; P = 0.0025) (Fig. 4F). In contrast, the expression of Dedd2 in db D2W was significantly decreased compared with the level in non-db (expression level mean ± SD: non-db 0.85 ± 0.11, db 0.39 ± 0.12; P = 0.0022) (Fig. 4G). The data clearly show that in neutrophils, Dedd2 is a target of miR-129-2-3p; however, at this time point during wound healing, some other factors, such as other miRNAs, inhibit Dedd2, and this mechanism is possibly even more effective in diabetic wounds. In addition, the expression of Casp8 in db D2W was significantly increased compared with the level in non-db (Supplementary Fig. 4A).

The expression of cleaved CASP6 in D2W did not show a significant change between non-db and db, although the number of neutrophils at D2W in db was significantly higher than in non-db (Fig. 4H). This result suggests that apoptosis of cells present in the wound site might be delayed.

To determine whether neutrophils express CASP6 and CCR2 during the early stage of inflammation, we performed fIHC in db D2W. fIHC showed that CASP6 and CCR2 were predominantly expressed in neutrophils (Supplementary Fig. 4B).

On the basis of the results of IHC (Fig. 4A and B) and our previous report (5), the expression of miR-129-2-3p showed a significant decrease in db, although the number of neutrophils at D2W in db was significantly higher than in non-db. In addition, Casp6 and Ccr2 tended to be overexpressed in db D2W, and miR-129-2-3p, CASP6, and CCR2 were predominantly expressed in wound-infiltrating neutrophils (Fig. 4C and Supplementary Fig. 4B). These results suggest that miR-129-2-3p is insufficiently activated in diabetic-derived neutrophils at D2W.

We previously reported the usefulness of antisense oligonucleotides using pluronic F-127 gel in skin wounds (20). Therefore, it is useful to use a gel to verify the role of molecules in wound healing.

First, to clarify the pathophysiological role of miR-129-2-3p in skin wound healing, we made a wound in the dorsal skin of db and topically applied miR-129-2-3p mimic or mutant negative control mixed with pluronic F-127 gel immediately after wounding. Wound closure was significantly accelerated in miR-129-2-3p mimic–treated compared with miR-129-2-3p mutant control–treated wounds from D7W to D21W in db (wound area [%] mean ± SD: D7W: mutation 130.5 ± 18.2, mimic 101.4 ± 18.1 [P < 0.05]; D10W: mutation 132.2 ± 38.3, mimic 96.6 ± 16.3 [P < 0.01]; D14W: mutation 115.2 ± 25.9, mimic 70.4 ± 29.0 [P < 0.001]; D21W: mutation 36.8 ± 14.4, mimic 11.0 ± 9.5 [P < 0.05]) (Fig. 5A and B), although there was no apparent effect in non-db mice (wound area [%] mean ± SD: D7W: non-db 38.2 ± 25.7, db 130.5 ± 18.2 [P < 0.001]; D10W: non-db 10.1 ± 1.9, db 132.2 ± 38.3 [P < 0.001]; D14W: non-db 2.0 ± 0.86, db 115.2 ± 25.9 [P < 0.001]).

Next, to investigate the proportion of neutrophils among cells present at D3W, we performed IHC from application of mimic or mutation immediately after wounding of db. IHC for neutrophils showed a positive signal in miR-129-2-3p mutant control–treated wounds (Fig. 5C). Moreover, we examined the relative neutrophil-positive area in D3W and found a reduced signal in miR-129-2-3p mimic–treated compared with that in miR-129-2-3p mutant control–treated wounds (signal level mean ± SD: mutation 8.9 ± 2.8, mimic 5.1 ± 1.1; P = 0.057) (Fig. 5D).

Finally, to confirm the specificity of the miR-129-2-3p mimic, the expression of Casp6 and Ccr2 was examined at D3W from application of mimic or mutant control immediately after wounding in db. The results showed that the expression of Casp6 and Ccr2 was significantly decreased in mimic-treated D3W (expression level mean ± SD: Casp6: mutation 1.1 ± 0.12, mimic 0.85 ± 0.17 [P = 0.016]; Ccr2: mutation 1.3 ± 0.25, mimic 0.96 ± 0.22 [P = 0.029]) (Fig. 5E and F).

Taken together, these results strongly suggest that the miR-129-2-3p mimic was effective at regulating gene expression in diabetic-derived neutrophils and could potentially rescue biological processes such as apoptosis in diabetic-derived neutrophils in the wound healing process. This would result in an improvement in delayed wound healing (Fig. 6) but does not completely rule out the possibility that miR-129-2-3p also may affect the behavior of other cells in vivo, which also may contribute to enhanced healing.

Wound healing is a complex process that comprises inflammatory, proliferative, and remodeling phases. BM-derived cells (BMDCs) migrate to and participate in the homeostasis of skin tissue. After cutaneous injury, a heterogeneous population of BMDCs are recruited to the site of injury and contribute directly to the repair process by differentiating into various types of skin cells, such as fibroblasts, keratinocytes, and endothelial cells (21,22). They can also indirectly modulate repair and regeneration by producing cytokines and growth factors that promote reepithelialization, neovascularization, and wound closure at the site of injury (3). In patients with diabetes and animal models of diabetes, BMDCs, including neutrophils, contribute to an impaired healing/chronic wound environment by prolonging the inflammatory response and/or failing to promote the regenerative phase of wound healing (23–25). Neutrophils are the first immune cells recruited to the injured site in acute wound inflammation; they constitute up to 50% of the cells during the early phase of inflammation (5) and prevent microbe invasion through the process of phagocytosis (26). We previously showed that the recruitment and/or retention kinetics of a heterogeneous population of BMDCs, including neutrophils, in diabetic cutaneous wounds are aberrant, leading to prolonged inflammation (5). These cells constitute the first subset of leukocytes to localize to injured tissue and may influence the entire localized inflammatory response.

miRNAs have been reported to be involved in both innate and adaptive immune responses (10); however, their role and regulation in neutrophils in the diabetic environment have remained unknown. To shed light on their possible role in dysfunction of diabetic-derived neutrophils, we examined miRNA expression and function in diabetic-derived neutrophils. Of note, the results showed that the expression of miRNAs involved in the inflammatory response changed in diabetic-derived neutrophils. Regarding miR-223, the expression of which was increased in diabetic-derived neutrophils in the current study, it was reported to be associated with cell proliferation, apoptosis, migration, and invasion in gastric cancer (27). We further elucidated the function of miR-223 in skin wound healing by analyzing miR-223 knockout mice (28). Similarly, miR-31 was highly expressed during the transition from the inflammatory to the proliferative phase in an in vivo human skin wound healing model, and the overexpression of miR-31 promoted cell proliferation and migration in human primary keratinocytes (29). On the other hand, miR-149 expression was decreased, and its target genes were shown to be involved in cell proliferation and apoptosis in patients with acute injuries of the skin (30). Thus, these miRNAs might be involved in cell proliferation, migration, and apoptosis in diabetic-derived neutrophils. A recent article showed that one of the factors associated with delayed wound healing in type 2 diabetic mice is Dnmt1-dependent dysregulation of hematopoietic stem cell (HSC) differentiation toward macrophages and that the expression of Dnmt1 is regulated by miRNAs (31). In the current study, the expression of miRNAs and mRNAs was significantly altered in diabetic-derived neutrophils, and this was associated with impaired wound healing. Alterations in gene expression in diabetic-derived neutrophils might be predetermined at the level of HSCs as described above, and epigenetic modifications in HSCs may be induced by type 2 diabetes.

In this study, microarray and qRT-PCR showed that the expression of miR-129-2-3p was downregulated in diabetic-derived neutrophils. Other recent studies have shown that miR-129-2 is regulated epigenetically by DNA methylation (32). Analysis of chromatin immunoprecipitation data of the regulatory region in putative intron 1 of the gene (∼4,000 base pairs upstream of the sequence encoding the mature miRNA) showed that this region is bound by many transcription factors, including Pu.1 and Cebp, both of which are underexpressed in diabetic-derived Gr-1⁺CD11b⁺ myeloid cells (33), which include neutrophils (Supplementary Fig. 4A and B). Thus, it is possible that the decrease in these transcription factors in diabetic-derived neutrophils contributes to the decreased expression of miR-129-2-3p. There are also reports that miR-129 family members are associated with proliferation and apoptosis in some types of cancer, such as esophageal carcinoma and breast cancer (34,35). Moreover, Wang et al. (36) reported that the topical administration of miR-129 agomir in diabetic animals promotes diabetic wound healing. GO and pathway analyses in this study also indicated that the predicted target genes of miR-129-2-3p are involved in a number of biological processes, including the inflammatory response, neutrophil chemotaxis, phagocytosis, and the execution phase of apoptosis, and are associated with multiple pathways, such as cell differentiation, Toll-like receptor signaling, chemokine signaling, IL-6 signaling, and the inflammatory response pathway. We thus hypothesized that miR-129-2-3p in particular might be involved in the functional regulation of diabetic-derived neutrophils in chronic inflammatory processes.

Neutrophils are produced constantly in large numbers in BM, and by definition, the same numbers of cells must die or migrate away within a defined time period for homeostasis to be maintained (6). Several studies also have suggested that the caspase family plays an important role in both spontaneous and Fas receptor–mediated apoptosis in neutrophils (37–39). The activation of death receptors with Fas ligand is involved in the activation of Casp8, which actually is a component of the death-induced signaling complex and activates downstream signaling. The activation of Casp8 has been noted in neutrophils, and the inactivation of this protease has been shown to delay apoptosis (37). In the current study, the expression of Casp8 was significantly increased in db skin wound on D2W. In addition, the expression of Casp6, which is a downstream executioner caspase, increased in db skin wound on D2W. These results suggest that apoptosis of neutrophils in db skin wound sites on D2W might be facilitated and/or inhibited by the activation of Casp8 and/or the direct regulation of Casp6 (40). Ccr2 is a chemokine receptor expressed in monocytes and lymphocytes but not in neutrophils. However, its expression changes under acute inflammation or in response to specific inflammatory stimuli in wounds. In mice with severe sepsis, Ccr2 is expressed in neutrophils (41–45), and wound recruitment is controlled by Ccr2 (46,47). Accordingly, Ccr2 expression on diabetic-derived neutrophils may be critical for chronic inflammation. In the current study, the expression of Ccr2 was also significantly increased in D2W. Thus, diabetic-derived neutrophils from BM may be excessively recruited to wound sites on D2W.

Previously, we showed that the numbers of myeloid cells, including neutrophils, in cutaneous wounds are significantly increased on D2W, and the recruitment and/or accumulation kinetics of these cells are altered (5). In the current study, the expression levels of Casp6 and Ccr2 in D2W of db increased compared with those in non-db. Moreover, our results showed that miR-129-2-3p directly regulated Casp6 and Ccr2 translation. Our in vivo analysis showed that skin wound healing in db was significantly accelerated from D7W. These results strongly suggest that the recruitment and accumulation kinetics of diabetic-derived neutrophils might be improved by overexpression of miR-129-2-3p, resulting in improved wound healing.

In conclusion, miRNAs are differentially expressed in diabetic-derived neutrophils compared with their levels in nondiabetic-derived neutrophils, particularly miR-129-2-3p. The results suggest that miR-129-2-3p directly regulates Casp6 and Ccr2 translation and is involved in inflammatory responses, apoptosis, chemotaxis, phagocytosis, and endocytosis. These findings further suggest that the deregulation of miR-129-2-3p contributes to the dysfunction of diabetic-derived neutrophils. The retention kinetics of neutrophils and chronic inflammation may be initiated through miR-129-2-3p-regulated genes, such as Casp6 and Ccr2. Accordingly, we suggest that miR-129-2-3p might be involved in the cellular kinetics and functional regulation of wound-recruited neutrophils and, as such, may prove to be a useful target for manipulation in a clinical context.

Acknowledgments. The authors thank the Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript.

Funding. This work was supported in part by the Japan Society for the Promotion of Science (Grant-in-Aid for Young Scientists B, 15K20314 and 17K17021) and the Cell Science Research Foundation (Osaka, Japan).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. T.U. and R.M. analyzed the results. T.U., R.M., K.A.M., and K.I. conceived the experiments. T.U., T.M., Y.A., and T.Y. conducted the experiments. All authors reviewed the manuscript. T.U. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.