MicroRNA-34a deficiency leads to impaired wound closure by augmented inflammation in mice.

BACKGROUND: Proper inflammation resolution is critical for cutaneous wound healing and disordered inflammation resolution results in chronic nonhealing wounds. However, the cellular and molecular mechanisms for resolution of inflammation during skin wound healing are not well understood. MicroRNA-34a is regarded as one tumor suppressor with complexed immune regulatory effects, yet its role during skin wound repair is still unclear.
METHODS: Circular full thickness excisional wounds were made on the dorsal skin of C57 mice and miR-34a expression pattern was examined by real time RT-PCR and in situ hybridization. The wound healing rates and histologic morphometric analysis were quantified and compared between wounds treated with antagomir-34a and autologous control antagomir-NC wounds, as well as wounds between miR-34a knockout (KO) and wild type (WT) mice. Immunohistochemistry (IHC) for both MPO and F4/80 were performed to examine the infiltrative neutrophils and macrophages in wounds from miR-34a KO and WT mice. Cytokines including IL-1β, IL-6, TNF-α and IL-10, were detected and analyzed by real time RT-PCR during wound healing. IHC for IL-6 and p-STAT3 were quantified, and WB for p-STAT3 and IL-6R were examined in wounds of miR-34a KO and WT mice.
RESULTS: We found miR-34a was significantly downregulated in the inflammatory phase and back to normal levels in the proliferative phase. Both topical knockdown wounds miR-34a levels by antagomir gel and systematic knockout miR-34a using KO mice resulted in impaired wound healing with delayed re-epithelialization and augmented inflammation. IHC results indicated that there were more residual infiltrative inflammatory cells in the proliferative phase. Moreover, over-activated IL-6/STAT3 signal pathway was identified in the wounds of miR-34a KO mice.
CONCLUSIONS: Our findings reveal that miR-34a deficiency augments skin wound inflammation response and leads to impaired wound healing, which suggest that targeted inhibition of miR-34a for tissue repair/regeneration should be with serious consideration. 2020 Annals of Translational Medicine. All rights reserved.

Introduction

Skin wound healing is one fundamental physiological process with a series of complex and well-orchestrated cellular and molecular events (1,2). The whole wound healing process comprises three sequential and overlapping phases: inflammatory phase, proliferative phase and remodeling phase. And appropriate transition from one phase to the next phase is key for proper healing (3-5). In particular, the transition from inflammatory phase to the proliferative phase plays critical role and determinates the outcome of healing (3). Resolution of inflammation in time facilitates the transition of the wound into the proliferative phase, and then promotes wound healing. In contrast, the constant inflammatory state results in failure to enter the proliferative phase, leading to chronic nonhealing wounds characterized by excessive and persistent inflammation (6,7). Thus, after skin injury, resolution of inflammation is one key step for successful healing. Although many factors were identified as key participators regulating wound inflammation resolution, both the cellular and molecular mechanisms of inflammation resolution during skin wound healing are still not fully elucidated.

MicroRNAs are endogenous small non-coding RNAs, which play key roles in diverse physiologic and pathologic processes (8). Recently, the great potential of miRNAs as therapeutic targets for correcting obstacles to avoid healing discords like chronic wounds or keloids has attracted broad interest in investigating their roles in cutaneous wound healing (9,10). Several microRNAs were identified as key regulators of inflammation resolution during healing (11-13). For example, miR-132 was highly upregulated in keratinocytes during the inflammatory phase and peaked in subsequent proliferative phase, which is a critical regulator of skin wound healing facilitating inflammation resolution by targeting HB-EGF (11). To further understanding mechanisms of inflammation resolution during wound healing, more candidate microRNAs are worth exploring.

In this study, miR-34a, the high abundant member of microRNA-34 family in mouse skin, was found significantly downregulated in the inflammatory phase and back to normal levels in the proliferative phase during healing. Both topical knockdown miR-34a levels in wounds by antagomir gel and systematic knockout miR-34a using KO mice resulted in impaired wound healing. Further histopathological investigations showed excessive inflammation in the wounds with miR-34a deficiency. Moreover, over-activated IL-6/STAT3 signal pathway was identified in the wounds of miR-34a KO mice. Our findings indicate that dynamic expression of miR-34a is critical for limiting the inflammatory response and facilitating normal wound healing in mice.

Methods

Animals

All mice related experiments were approved by Institutional Animal Care and Use Committee of Third Military Medical University (TMMU, Chongqing, China). miR-34a knockout mice were obtained from Professor Jun Peng (Department of Hematology, Qilu Hospital, Shandong University, Jinan, China). To minimize potential interference, both sex and age matched wild-type littermates were used as control for miR-34 KO mice. C57BL/6J mice used for detecting miR-34a expression pattern and topical knockdown experiments were obtained from the Institute of Zoology (Chinese Academy of Sciences, Beijing, China). All the mice were bred and maintained at the Center of Experimental Animal, TMMU.

Wound healing studies

To examine the expression pattern of miR-34a during wound healing, the 8 to 12-week-old C57BL/6J mice were anesthetized with 1% pentobarbital sodium (30 mg/kg) and two 6.0 mm-diameter circular full-thickness excisional wounds were made in the shaved dorsal skin. At 0, 3, 7 and 14 days after wounding mice were euthanized, the wound tissues were harvested with a 8-mm biopsy punch for either RNA isolation or in situ hybridization assays.

To knockdown miR-34a in the topical wound sites, antagomir-34a in Pluronic F-127 gel with a concentration of 5 um were used immediately following wound creation as described previously (14). To evaluate the effects of miR-34a knockout on wound healing, the wounds were made on the back skin of both the miR-34a KO and wild-type mice. And then, the wound healing process was digitally photographed at different time points. Wound area measurement was performed as described previously (15,16).

Quantitative real-time PCR

For detecting microRNAs, Bulge-Loop miRNA qRT-PCR Primer Set (Ribobio Co.) was used according to manufacturer’s instructions. For regular quantitative real-time PCR, the method was performed as described previously. Primer pairs used are listed as following: IL-1β F: TCTCGCAGCAGCACATCA; IL-1β R: CACACACCAGCAGGTTAT; IL-6 F: TGGGAAATCGTGGAAATGAG; IL-6 R: CTCTGAAGGACTCTGGCTTTG; TNF-α F: CCCGGGCTCAGCCTCTTCTCATTC; TNF-α R: GGATCCGGTGGTTTGCTACGACGT; IL-10 F: CAACATACTGCTAACCGACTCCT; IL-10 R: TGAGGGTCTTCAGCTTCTCAC; TBP F: AAGGGAGAATCATGGACCAG; TBP R: CCGTAAGGCATCATTGGACT.

In situ hybridization

In situ hybridizations of paraffin-embedded skin sections to detect miR-34a were performed according to previous report (14).

Histology and morphometric analysis

H&E sections of the central portion of the wounds were made for histology and morphometric analyses as previously described (15,17). In brief, epidermal thickness was measured at five different positions per mouse and averaged to one data set. Wound width was determined as the distance between the wound margins, which were defined by the last hair follicles. The percentage of re-epithelialization was calculated as distance covered by epithelium dividing the wound width.

Immunohistochemistry

IHC staining in skin tissues were performed as previously described (18). Briefly, tissue sections were stained with rabbit anti-MPO (1:200, Thermo), mouse anti-F4/80 (1:200, BioLegend), rabbit anti-IL-6 (1:600, Proteintech), and mouse anti-p-STAT3 (1:200, CST). Then, DAB chromogenic system was used for final chromogen. Images of areas of interest were collected by Olympus IX73-A21PH microscope (Olympus, Japan). Epidermal p-STAT3 positive cells were quantitified as the number of positive cells per mm migrating epidermis. Quantifications for MPO, F4/80, IL-6, and stromal p-STAT3 positive cells were performed by using Image J by calculating 6-8 40× high power field photos per mouse.

Western blots

Western blots were performed according to our previous publication (19). In brief, protein lysates from wound tissues were extracted using Complete Lysis-M buffer containing both protease inhibitors and phosphatase inhibitors. The protein samples were resolved by SDS-PAGE, transferred onto PVDF membrane, blocked with 5% non-fat milk and incubated with primary antibodies at 4 °C overnight. Primary antibodies used include rabbit anti-p65 (1:1,000, CST), rabbit anti-p-p65 (1:1,000, CST), rabbit anti-STAT3 (1:1,000, CST), mouse anti-p-STAT3 (1:1,000, CST), rabbit anti-IL-6R (1:1,000, Proteintech), and mouse anti-actin (1:1,000, Beyotime, Hangzhou, China). Membranes were incubated with relevant IRDye conjugated secondary antibodies for one hour at room temperature (1:5000, Rockland Immunochemicals Inc., Gilbertsville, PA). The infrared fluorescence image was obtained using Odyssey infrared imaging system (Odyssey CLx, Li-Cor Bioscience, Lincoln, NE).

Statistical analysis

Experimental data were analyzed using GraphPad Prism 5.0 software. Data were presented as mean ± SD. Statistical analysis was performed by Student’s t tests (the U-test was performed to test differences between independent groups at different time points during wound healing). A P value <0.05 was considered to be statistically significant.

Results

Characterization of miR-34a expression during skin wound healing

Recent studies have shown that the miR-34 family plays important roles as one tumor suppressor (20,21). It has been reported that tumor suppressors may have special meaning for tissue wound repair and regeneration (22,23). To investigate the possible contributions of miR-34 family members during skin wound healing, their basic expression levels in intact mouse skin tissue samples were examined using real-time RT-PCR. As shown in , miR-34a was strongly expressed in the intact mouse skin samples, whereas miR-34c was expressed at lower levels, and miR-34b was expressed at lowest levels. Then, we examined the dynamic expression of the three miR-34 family members during skin wound healing process. Our results indicated that the three members had diverse dynamic expression during healing ( and ). Although both miR-34c and miR-34b had significantly upregulated in certain time points after injury, their expression levels were still far below miR-34a. So, our subsequent studies focused on miR-34a. Results of real time RT-PCR showed miR-34a expression level was significantly downregulated in day 3 after wounding, then recovered in day 7 and day 14 (). Further examination by in situ hybridization confirmed the results of PCR, which indicated the signal intensity of ISH fluorescence of wounds from day 3 was relatively weaker ().

Figure 1

Expression pattern of miR-34a during mice skin wound healing. (A) Expression levels of miR-34 family members were measured using real-time RT-PCR in intact full-thick skin tissues of mice (n=4–6). miRNA samples were normalized to U6 snRNA expression levels. (B) miR-34a levels were measured in wounds of day 0, 3, 7,14 time points during wound healing using real-time RT-PCR (n=4–6). (C) In situ hybridization for miR-34a performed in mouse skin wound sections of day 0, 3, and 7 after wounding. **P<0.01 versus day 0 time-point.

Figure S1

Dynamic expression of miR-34b and miR-34c during skin wound healing in mice. miR-34b (A) and miR-34c (B) levels were measured in wounds of day 0, 3, 7,14 time points during wound healing using real-time RT-PCR (n=4–6). *P<0.05, and **P<0.01 versus day 0.

Topical inhibition of miR-34a by antagomir resulted in delayed skin wound healing

As a well-known pleiotropic molecule, there are extensively reported that miR-34a has tumor suppressive functions like inhibiting cell stemness, proliferation, and migration, promoting apoptosis, and participating in immunological regulation (24-30). So, we speculated that the abundant miR-34a may have certain role in modulating skin wound healing. To investigate the in vivo effects of miR-34a on wound healing, endogenous miR-34 a was downregulated by administrating Pluronic gel with antagomir-34a to the wounds. The availability of antagomir gel for local inhibition of target miRNA in wounds have been reported by others and us previously (14,31). In this experiment, topical antagomir-34a treatment inhibited the miR-34a expression levels effectively examined by real time RT-PCR (). The antagomir-34a treated wounds showed slower healing for day 1, day 5 and day 7 time points compared with the autologous antagomir-NC treated control wounds, although all the wounds healed before day 11 (). Further morphometric analyses of the histopathological H&E sections confirmed the gross observation. Local knockdown of miR-34a delayed wound re-epithelialization (), showed larger wound widths and wound areas (), although without impacts on epidermal thickness ().

Figure S2

Quantitative real time RT-PCR analysis showed significantly inhibited miR-34a expression in wounds after using antagomir-34a at day 7 after wounding. *P<0.05 versus antagomir-NC.

Figure 2

Topical inhibition of miR-34a by antagomir Pluronic gel resulted in delayed skin wound healing. (A) Macroscopic appearance of wound closure in antagomir-NC and antagomir-34a treated mice at day 0, 3, and 7 after wounding. (B) Quantification of wound area in antagomir-NC and antagomir-34a treated autologous wounds in mice (n=5–8). (C) Representative H&E wound sections from antagomir-NC and antagomir-34a treated autologous wounds on day 7 after wounding. Quantification and calculation percentages of wound re-epithelialization (D), remaining wound widths (E), remaining wound areas (F), and epidermal thickness (G), on H&E sections at day 7 after wounding from antagomir-NC and antagomir-34a treated wounds. *P<0.05, **P<0.01, and ***P<0.001 versus antagomir-NC group at certain time-point.

miR-34a knockout mice showed impaired wound closure

Further experiments on miR-34a KO mice were implemented to confirm the effects of miR-34a on skin wound healing. miR-34a was undetectable in the skin samples of miR-34a KO mice (Figure S3A). And miR-34a knockout had no effects on the expression levels of both miR-34b and miR-34c in skin samples (Figure S3B,C). In line with the results of topical knockdown of miR-34a on wound healing, wounds of miR-34a KO mice showed significantly impaired healing from the day 1 to day 11 after wounding (). Histologically, miR-34a KO wounds had shorter epithelial migrating tongues by day 3 and day7, and thicker epidermis by day 7, compared to WT controls (). Meanwhile, both wound width and wound area were significantly larger in miR-34a KO wounds ().

Figure 3

miR-34a KO mice showed impaired wound closure. (A) Macroscopic appearance of wound closure in miR-34a KO and WT mice at day 0, 3, and 7 after wounding. (B) Quantification of wound area in wounds of miR-34a KO and WT mice (n=5–8). (C) Representative H&E wound sections from wounds of miR-34a and WT mice on day 7 after wounding. Quantification and calculation percentages of wound re-epithelialization (D), epidermal thickness (E), remaining wound widths (F), and remaining wound areas (G) on H&E sections at day 3 and 7 after wounding from wounds of miR-34a KO and WT mice. *P<0.05, **P<0.01, and ***P<0.001 versus WT mice at certain time-point.

Loss of miR-34a enhanced wound inflammation

miR-34a downregulated in day 3 after injury and then restored to the normal levels (), which suggested miR-34a was specifically reduced in the inflammatory phase during wound healing. And further local knockdown miR-34a using antagomir or systematic knockout miR-34a in mice resulted in impaired healing with larger residual crusta (), which was a feature of excessive inflammatory response. So, we examined the infiltration of neutrophils and macrophages after wounding, which are critical inflammatory cells for proper healing. As shown in , the accumulations of both MPO-positive neutrophils and F4/80-positive macrophages were significantly increased in miR-34a KO mice compared to WT mice in wounds of day 7 after wounding. Further WB for examining both p65 and p-p65 expression levels corroborated the enhanced inflammatory response in wounds of miR-34a KO mice compared to WT mice (). To explore the possible mechanism of enhanced inflammatory response in wounds of miR-34a KO mice, real-time RT-PCR for classical cytokines expression levels were implemented (). During wound healing process, miR-34a knockout mice had little effect on IL-1β expression levels and showed increased TNF-α expression in day 3 wounds but without significant difference. However, IL-6 expression was downregulated in day 7 wounds of WT mice, while sustained in those of KO mice. In addition, IL-10 expression showed a sharp reduction in day 7 wounds of KO mice.

Figure 4

Effect of miR-34a knockout on inflammatory response during wound healing. (A) Representative images of day 7 wounds IHC (up panel) and quantification of MPO positive neutrophils of day 3 and 7 wounds of miR-34a KO and WT mice (down panel) (n=5–8). (B) Representative images of day 7 wounds IHC (up panel) and quantification of F4/80 positive macrophages of day 3 and 7 wounds of miR-34a KO and WT mice (down panel) (n=5–8). (C) Detection of p65 and p-p65 protein levels in wounds between miR-34a KO and WT mice at different time points during wound healing by WB. Relative densitometric quantifications of p-p65 protein normalized to p65 and p65 protein normalized to beta-actin are indicated. (D) Examination of cytokines including IL-1β, IL-6, TNF-α, and IL-10 mRNAs levels in wounds between miR-34a KO and WT mice at different time points during wound healing by real time RT-PCR (n=5–8). *P<0.05, and **P<0.01 versus WT mice at certain time-point.

Over-activated IL-6/STAT3 signaling in wounds of miR-34a KO mice

IL-6/STAT3 signaling pathway has showed extensive interactions with miR-34a (32-35), so we focused on this pathway in further investigations. To verify the constant elevated IL-6 expression detected by real-time PCR, immunohistochemical staining was used for semi-quantitative analysis of IL-6-positive cells in granulations, which confirmed the persistent high expression of IL-6 in day 7 wounds of miR-34a KO mice (). As STAT3 is one critical component of IL-6 downstream signal pathway, we examined the phosphorylated STAT3 in the wounds by immunohistochemistry. Quantization of p-STAT3-positive epidermal cells indicated more numbers of positive cells in KO mice at both day 3 and day 7 time points, although there was no significant difference in day 7 between the two groups (). Moreover, the p-STAT3 positive cells in dermal granulation in wounds of miR-34a KO mice were also significantly higher than those of WT mice in day 7 (). In line with the IHC results, WB for the phosphorylated STAT3 showed similar trend that miR-34a KO in wounds promoted STAT3 signaling activation (). As one key miR-34a functional target, IL-6R was also examined. The result indicated elevated sIL-6R (soluble IL-6R) level in day 0 and 3 wounds of KO mice, while less affected mIL-6R (membrane-bound IL-6R) level ().

Figure 5

Quantification of IL-6 positive cells in granulations. (A) Representative images of IL-6 staining in wounds of miR-34a KO and WT mice at indicated time points. (B) Quantification of IL-6 positive cells in wounds of miR-34a KO and WT mice at indicated time points (n=5–8). *P<0.05 versus WT mice at certain time-point.

Figure 6

Over-activated IL-6/STAT3 signaling in wounds of miR-34a KO mice. (A) Representative images of p-STAT3 staining in epidermis of wounds from miR-34a KO and WT mice at day 3 after wounding. (B) The number of epidermal p-STAT3 positive cells per mm is graphed (n=5–8). (C) Representative images of p-STAT3 staining in stromal granulations of miR-34a KO and WT mice at day 7 after wounding. (D) The number of stromal granulation p-STAT3 positive cells percentage is graphed (n=5–8). (E) Detection of IL-6R, STAT3 and p-STAT3 protein levels in wounds between miR-34a KO and WT mice at different time points during wound healing by WB. Relative densitometric quantifications of p-STAT3 level normalized to STAT3, m-IL-6R and s-IL-6R normalized to beta-actin are indicated. *P<0.05, and **P<0.01 versus WT mice at certain time-point.

Discussion

In the current study, we found that the cutaneous abundant miR-34a deficiency resulted in impaired wound healing, which was caused by excessive inflammation response through over-activated IL-6/STAT3 signaling pathway in mice.

Previous studies showed different expression patterns of three miR-34 family members in skin tissues (25,36). Our results indicated that miR-34a was the abundant expression member of the family in skin, which expressed higher than miR-34b and miR-34c in the order of magnitude. This expression feature from intact full-thickness mouse skin samples is in accord with the results from primary mouse keratinocytes and human skin tissues (25,36). Further experiments were performed to examine the dynamic expressions of miR-34 family members during skin wound healing of mice. Although both miR-34b and miR-34c expressed variably, miR-34a was still the abundant member of miR-34 family during healing. Thus, we focused on role of miR-34a in wound healing for next investigations. We found that miR-34a was downregulated specifically in inflammatory phase of wound repair (day 3 after wounding) detected by both real time RT-PCR and ISH. As one tumor suppressor microRNA, the early reduction of miR-34a may facilitate wound healing by promoting keratinocytes proliferation and migration, just like inhibition of p53 and RB at the initial stages of regeneration (22,23). Also, miR-34a plays important roles in immune response, including impairing neutrophils migration (37), inhibiting macrophages efferocytosis (38), modulating dendritic cells differentiation (39), regulating cytokines productions (29), and so on. Thus, the downregulation of miR-34a in inflammatory phase may be beneficial to neutrophils chemotaxis and cytokines producing in this repair stage.

To determine the roles of miR-34a during skin wound healing, both topical inhibition of miR-34a by antagomir gel in wounds and systematic knockout of miR-34a in mice models were employed. Unexpectedly, both models showed miR-34a deficiency resulted in impaired wound closure. Recent study reported that elevated miR-34a in wounds by local intradermal injecting mimics also led to delayed healing (36). Those conflictive results displayed the complexity of miR-34a function in skin wound healing. Theoretically, microRNAs serve as one moderate gene regulatory model, which have target genes with differential sensitivity under different cellular contexts (40,41). Studies have shown transgenic expression and deletion of one same miRNA regulate largely distinct sets of target genes (40). So, loss-of-function and gain-of-function could impact different target genes and signaling pathways, then present different phenotypes. That is, the appropriate expression levels of miR-34a, neither high or low, is critical for proper wound healing. Experimentally, well-designed in vivo loss-of-function studies are more convincing than the gain-of-function research. The microRNA mimics should be used with caution for introducing the supraphysiological levels of mature miRNAs and artifactual RNA species led to non-specific changes in gene expression (42). We note that the wounds treated with miR-34a mimics achieved far more than the physiological conditions (36). In addition, indiscriminate upregulating miR-34a levels in all the repairing cells would confuse the actual role of miR-34a during wound healing (36), for miR-34a showed diverse or even opposite effects in different cell types (43,44).

There are reports that miR-34a in keratinocytes can inhibit proliferation, migration, and promote differentiation (24,25,36). Also, reports indicated miR-34a can inhibit angiogenesis (45,46). So, the effect of miR-34a deficiency on delayed healing in mice is a bit surprising to us. In the other hand, miR-34a also have extensive and complexed regulatory roles in inflammatory response (29,37-39,43,44). And persistent inflammatory response with resolution dysregulating could led to chronic non-healing wounds (3,6,47,48). Further investigations indicated the phenotypes of miR-34a-deficiency-related impaired wounds resembled the chronic wounds, which showed excessive inflammatory response and delayed re-epithelialization. Therefore, we speculated that augmented inflammation caused by miR-34a loss-of-function overcame the possible advantages from miR-34a deficiency like promoting keratinocytes proliferation and migration, improving angiogenesis, and so on.

To interpret the augmented inflammation induced by miR-34a deficiency, we identified IL-6/STAT3 signaling pathway was over-activated in KO wounds, which suggested sustaining activation of this pathway may block resolution of inflammation and then resulted in delayed wound closure. It has been reported that IL-6R/STAT3/miR-34a feedback loop promotes EMT-mediated colorectal cancer invasion and metastasis, which considers that STAT3 activation can inhibit miR-34a transcription and in turn miR-34a can suppress IL-6R to control the IL-6/STAT3 pathway activation (32). If so, it partly explains the downregulated miR-34a in the early inflammatory phase, which stage showed stronger STAT3 pathway activation. However, the IL-6R protein levels did not show significant increasement in KO wounds in the mIL-6R form except the slight upregulated sIL-6R form in early KO wounds. Although sIL-6R can activate the proinflammatory IL-6 trans-signaling even on cells without IL-6R expression (49), its contribution in the delayed wound closure of miR-34a KO mice needs further investigation with caution.

Conclusions

Taken together, our study revealed the dynamic expression of miR-34a is critical for resolution of inflammation during wound healing in mice. Our findings indicate that miR-34a deficiency augments skin wound inflammation response through over-activated IL-6/STAT3 signaling pathway, and then leads to impaired wound closure. Those results suggest targeted inhibition of miR-34a for tissue repair and/or regeneration should require serious consideration.

Effect of miR-34 knockout on the expression levels of the three miR-34 family members. Quantitative real time RT-PCR analysis showed the expression levels of miR-34a (A), miR-34b (B), and miR-34c (C) in skin samples between miR-34a KO and WT mice. ***P<0.001 versus WT mice.

The article’s supplementary files as