G-CSF Administration Accelerates Cutaneous Wound Healing Accompanied With Increased Pro-Hyp Production In db / db Mice

Shiro Jimi1*, Kenji Sato2, Masahiko kimura3, Junji Suzumiya4, Shuuji Hara3, Francesco De Francesco5, Hiroyuki Ohjimi6 1Central Laboratory for Pathology and Morphology, Faculty of Medicine, Fukuoka University, Fukuoka, Japan. 2Marine Biological Function, Division of Applied Biosciences, Kyoto University, Kyoto, Japan. 3Department of Drug Informatics, Faculty of Pharmaceutical Sciences, Fukuoka University, Fukuoka, Japan. 4Department of Oncology/Hematology, Shimane University Hospital, Shimane, Japan. 5Multidisciplinary Department of Medical-Surgical and Dental Specialties, Second University of Naples, Italy. 6Departments of Plastic, Reconstructive and Aesthetic Surgery, Faculty of Medicine, Fukuoka University, Japan Clinical Research in Dermatology: Open Access Open Access Research Article


Introduction
With the increasing number of diabetic patients, treatment for delayed healing of wounds, including foot diseases and ulcerations, has become a serious medical concern worldwide [1].Many approaches to treatment have been developed; Negative-pressure wound therapy is one option to remove the harmful excess fluid that results from delayed wound healing and improve blood circulation, leading to more favorable outcomes [2].It has also been shown that growth factor therapy using basic fibroblast growth factor promotes wound healing through its positive effects on fibroblast and keratinocyte growth and neovascularization [3].However, the available strategies to address non-healing wounds that arise from various pathologies are still limited.Therefore, it is necessary to explore the novel tissue factors and/or mechanisms involved in delayed wound healing.
An inflammatory cascade takes place after wounding, and phagocytic cells, including neutrophils, are recruited from the bloodstream [4][5][6].The density of inflammatory cells and/or the duration of inflammation are important factors that influence wound healing.Prolonged production of proinflammatory cytokines, reactive oxygen species, and protein-digestive enzymes can give rise to uncontrolled leukocyte accumulation.This pathological state can be observed in the chronic refractory wounds that accompany peripheral vascular disease and diabetes [7].Wound healing is also delayed in a neutropenic state [8].Thus, there is a delicate balance between healing and inflammation in healing wounds.

G-CSF Administration Accelerates Cutaneous Wound Healing Accompanied With Increased Pro-Hyp Production In db/db Mice
Copyright: © 2017 Jimi, et al.
IL-8 and leukotriene from the wounded tissue, recognize selectin expressed on endothelial cells and start rolling, and then attach to intercellular adhesion molecule-1.Neutrophils then penetrate the endothelial monolayer through recognition of platelet/ endothelial cell adhesion molecule-1 and transmigrate into the tissue matrix.Finally, the neutrophils reach the wound via the gradient of migration factors.
Neutrophils are produced in the bone marrow following stimulation with proinflammatory cytokines, stem cell factor (SCF), and granulocyte colony-stimulating factor (G-CSF), which induce their differentiation, and mature neutrophils are mobilized into the bloodstream [9].G-CSF is used clinically to treat neutropenic patients and to harvest hematopoietic stem cells.However, it is also known that G-CSF possesses a variety of functions, such as, endothelial-progenitor cell mobilization [10], stimulation of endothelial proliferation [11], mesenchymal stem cell mobilization [12], and augmentation of antibacterial function [13].Since all these activities favor wound healing, many clinical studies of G-CSF have been conducted on refractory wounds in neutropenic patients [14], diabetic foot infections [15], and acute myocardial infarctions [16].
The sequential steps in wound healing are inflammation, granulation tissue formation, wound contraction, and epithelial regeneration [6,17,18].These steps proceed in a serial and overlapping manner.Of these steps, the formation of granulation tissue is critical because it acts as a scaffold for epidermal regeneration and an axis for the blood supply needed to maintain cellular activities [19].The basic structure of granulation tissue consists of fibroblasts and extra cellular matrix (ECM) components, including collagens, which act as a scaffold for endothelial cells to form neovascular networks [20].In the final step, a wound is covered by regenerative epithelial cells, and the granulation tissue regresses.
An important function of neutrophils is ECM degradation, which leads to the remodeling of healing tissue [18].In studies on mice that overexpress the collagen catabolic enzyme matrixmetalloproteinase-8 (MMP-8) [21] and MMP-8 knockout mice [21], wound healing is delayed.It has also been shown that metabolites of collagen possess biological functions.Ingested collagens are supposed to be digested to single amino acids.However, collagen dipeptides, including proline-hydroxyproline (Pro-Hyp) and hydroxyproline-glycine (Hyp-Gly), appear for a certain time in the blood [22].It has been shown that these dipeptides exert biological effects on skin barrier function [23] and fibroblast growth [24].
In the present study, with an aim to clarify the effect of G-CSF administration on refractory diabetic wounds, we examined the relationship between neutrophil infiltration and collagenderived peptides in wounds using genetically diabetic db/db mice.This is the first report to clearly show that the collagenderived dipeptide Pro-Hyp is produced in inflamed wound tissue, which may play a role in wound healing.
All procedures were conducted under aseptic conditions, using autoclaves, ethylene oxide gas, 70% ethanol, and povidoneiodine.Mice were anesthetized with pentobarbital (Somnopentyl; KYORITU SEIYAKU, Tokyo, Japan).After removing the dorsal hair with a commercial depilatory, a circular tattoo (1 cm in diameter) was made at the center of the lumbar area, and this section of skin was completely excised with scissors.Finally, the wound was covered with a polyurethane film dressing (Tegaderm; SUMITOMO 3M, Tokyo, Japan).

Measurement of G-CSF
G-CSF was measured in blood and tissue samples.Blood samples were obtained from the orbital sinus 3-96 hours after wounding.Tissue samples were also obtained from the wounded and normal subcutaneous tissue on the back of the same mice 3-24 hours after wounding.After weighing the tissue, it was kept in ice-cold protein extraction buffer (RIPA Lysis Buffer; MERCCK MILLIPORE Co., Darmstadt, Germany).The tissue was then homogenized with a Biomasher (Nippi Inc., Tokyo, Japan).After centrifugation at 9000 g for 15 min at 4°C, the supernatant was used for measurement.The G-CSF concentration in blood and tissue was determined with the mouse G-CSF immunoassay kit (R&D Systems Inc., MN, USA), and the absorbance was measured with a microplate reader (SUNRISE Rainbow RC-R; TECAN Japan, Kawasaki, Japan) at 450 nm/540 nm.

Measurement of collagen peptides
Wounded and normal subcutaneous tissue samples (1 cm in diameter) were collected at 1, 3, 5, and 7 days after skin excision and weighed.Samples were immersed in 75% ethanol and stored in a deep freezer until analysis.Tissue was homogenized in a Biomasher and centrifuged at 9,000 g for 15 min at 4°C.The supernatant was used to measure Hyp-Gly and Pro-Hyp levels.Peptides in the samples were derivatized with 6-aminoquinolyl-N-hydroxy succinimidyl carbamate using the AccQ Tag kit (Waters, Milford, MA, USA).The derivatives were resolved by reversed phase HPLC using an Inertsil ODS-3 column (2 mm × 250 mm; GL Sciences, Tokyo, Japan) with a binary gradient elution consisting of 0.1% formic acid and 80% acetonitrile containing 0.1% formic acid and detected by MS/MS using a 3200 QTRAP (SCIEX, Framingham, MA, USA) equipped with an electron spray

G-CSF Administration Accelerates Cutaneous Wound Healing Accompanied With Increased Pro-Hyp Production In db/db Mice
Copyright: © 2017 Jimi, et al.
ionization probe in multi-reaction monitoring (MRM) mode.Optimization of the MRM conditions and quantification of the peptides were performed using installed software, Analyst and MultuQuant (AB CIEX, MA, USA), respectively.

Administration of recombinant human G-CSF
After skin excision, the subcutaneous splint model [25] was used to mimic human wound healing.The wound was sealed with a polyurethane film dressing, and 25 µg/100 µL recombinant human G-CSF (rhG-CSF: filgrastim; Kyowa Hakko Kirin Co., Ltd, Tokyo, Japan) or vehicle alone was injected onto the wound surface.On days 0, 3, and 5, the wounds were photographed along with a scale.The wound area was measured using a computer-assisted morphometric analyzer (VH Analyzer VH-H1A5; KEYENCE Co., Osaka, Japan).On days 3 and 5, the mice were sacrificed.Wounded tissue, including the splint, was dissected, and used for histological examination.

Histological examination
The wound tissue was fixed in 10% buffered formaldehyde (pH 7.4) for several days.Two cross-cut tissue samples from each wound were excised and embedded in paraffin blocks using a tissue processor (Tissue-Tec VIP Premiere; Sakura Seiki Co. Ltd., Nagano, Japan).Then, 4-µm-thick tissue sections were cut with a microtome (RM2235; Leica Biosystems, Nußloch, Germany).For frozen specimens, another tissue sample was fixed in 10% buffered formaldehyde (pH 7.4) containing 10% sucrose for 24 hours, and then washed with 15% and 20% sucrose-containing buffer solutions (pH 7.4) for 24 hours.The tissue was immersed in tissue compound (OCT; Sakura Seiki Co. Ltd.), vacuumed for 30 min, and frozen in liquid nitrogen.The frozen tissue block was cut with a cryostat (CM3050S; Leica Biosystems) into 4-µm-thick sections.

Evaluation of neutrophil infiltration and MMP-8 expression
Semi-quantitative evaluations of NIMP-R14-positive neutrophil infiltration and MMP-8 expression were performed using reference positive-stained photographs (0: absent, 1: several, 2: mild, 3: moderate, 4: severe, and 5: very severe).During the evaluations, the image assessment was blinded to the photographs that were used as evaluation criteria.Mean values were used.

Statistical analysis
Values were expressed as mean ± standard error.The two groups were compared using Student's t-test.P values less than 0.05 were considered significant.

G-CSF production after skin excision
G-CSF production was evaluated in C57BL/6 mice.Basal levels of G-CSF in the blood and normal subcutaneous tissue without injury remained at minimum levels (Fig. 1A).However, 3 hours after skin excision, G-CSF levels in wounded tissue rapidly reached a maximum, and gradually decreased thereafter over the 24 hours of the study.However, no such increase was observed in the uninjured subcutaneous tissue on the same mice.Blood levels of G-CSF also increased and reached a maximum value 6 hours after skin excision, and then decreased over time.G-CSFexpressing cells were not detected in the exposed muscle fascia 1 hour after skin excision, but spindle-to oval-shaped cells were detected after 3 hours (Fig. 1B).

Physiological changes in db/db and wild-type mice after skin excision
Blood glucose and body weight of db/db mice were quite high, whereas those of wild-type mice were in the normal range.After skin excision/wounding, there was no significant difference in body weight, WBC, RBC, or PLT between db/db and wild-type mice, and no significant changes were detected during the 7 days of the study (Table 1).

Plasma G-CSF concentration in db/db and wild-type mice after skin excision
Plasma G-CSF levels were determined 6 hours after skin excision (Fig. 2).Prior to the start of the study, plasma G-CSF levels were significantly lower in the db/db mice than in the wildtype mice (p < 0.01).After skin excision, G-CSF levels increased significantly in both mice; however, the levels in the db/db mice were significantly lower than those in the wild-type mice (p < 0.01).

Neutrophil infiltration and collagenase expression
Acute phase neutrophil infiltration and MMP-8 expression after skin excision were examined in serial sections.

G-CSF Administration Accelerates Cutaneous Wound Healing Accompanied With Increased Pro-Hyp Production In db/db Mice
Copyright: © 2017 Jimi, et al.
Representative photographs on days 3 and 5 are shown in Figure 3A.Infiltrated NIMP-R14-positive neutrophils and MMP-8-positive cells in both mice increased with time.On day 3, intracellular MMP-8 expression was detected, and it was also detected in the ECM on day 5, especially in the wild-type mice.Semi-quantitative analysis of neutrophil infiltration and MMP-8 expressing cells (Fig. 3B) showed no significant difference in the mice on days 3 and 5.However, both of them were significantly greater in wild-type mice than in db/db mice.
Colocalization of MMP-8 and NIMP-R14-positive neutrophils in wound tissue from a wild-type mouse was analyzed by double fluorescent immunostaining (Fig. 3C).Many of the NIMP-R14positive neutrophils also stained positive for MMP-8.MMP-8-positive but NIMP-R14-negative cells were also detected.Based on their structure, these cells may be other types, such as macrophages or endothelial cells.

Collagen peptide production in wound tissue
The granulation tissue that develops after wounding is remodeled by inflammatory cell infiltration.In this process, collagen, a major ECM component, is degraded by MMPs, including MMP-8.Here, we analyzed the changes in collagen dipeptide content in wounded and normal tissues in db/db and wild-type mice (Fig. 4).The levels of Hyp-Gly and Pro-Hyp in the wounded tissue were significantly lower in the db/db mice than in the wild-type mice.Pro-Hyp levels in wounded tissue gradually increased in both mice during the 7-day study; however, the levels were significantly lower in the db/db mice than in the wild-type mice.In contrast, in normal tissue obtained from the same mice, Pro-Hyp levels were unchanged during the study.In addition, no changes were observed in the Hyp-Gly levels in wildtype and db/db mice during the study.

Effect of topical rhG-CSF administration treatment on wound healing
The effects of topical rhG-CSF administration (25 µg/100 µL) on wound healing were examined.After administration of rhG-CSF, no significant changes in WBC, RBC, and PLT were observed (Fig. 5A).Wound healing, as shown by the percent wound closure 5 days after wounding, was more delayed in db/db mice than in wild-type mice (Fig. 5B).However, wound closure, measured as percent wound closure, was accelerated by rhG-CSF administration in db/db mice (p < 0.05), but was decreased (although not significantly) in wild-type mice.

G-CSF Administration Accelerates Cutaneous Wound Healing Accompanied With Increased Pro-Hyp Production In db/db Mice
Copyright: © 2017 Jimi, et al.
days after wounding were compared (Fig. 5C).Granulation tissue in all treatment groups consisted of spindle-shaped fibroblasts, round-shaped inflammatory cells, and MT-positive collagen fibers.In wild-type mice, no apparent differences were detected between rhG-CSF-treated and untreated mice.However, in the db/db mice, thickened collagen fibers were deposited in the matrix of rhG-CSF-treated mice.Such changes were detected in 4 out of 5 db/db mice (80%).

Neutrophil infiltration and MMP-8 expression in granulation tissue after rhG-CSF administration
Neutrophil infiltration and MMP-8 expression were quantified and assessed as 1 of 6 grades.Neutrophils and MMP-8 expressing cells were scarce in unwounded subcutaneous tissue.After skin excision, the extent of neutrophil infiltration into granulation tissue was lower in db/db mice than in wild-type mice (Fig. 6A).The number of neutrophils in the granulation tissue was analyzed 3 and 5 days after topical administration of rhG-SCF on the wounds.The number of neutrophils did not change or decreased slightly in the wild-type mice, but tended to increase in the db/db mice on day 5. MMP-8 expression in the granulation tissue also exhibited a similar tendency as found in neutrophil infiltration (data not shown).

Collagen peptide production after rhG-CSF administration
After topical administration of rhG-CSF on the wounds, Hyp-Gly and Pro-Hyp contents in granulation tissue was measured on day 5 (Fig. 6B).No differences in Hyp-Gly content in the db/db and wild-type mice with/without rhG-CSF administration were detected.In contrast, Pro-Hyp content in the db/db mice was significantly increased by rhG-CSF administration (p < 0.05); no such change was detected in the wild-type mice.

Discussion
It has been reported that plasma levels of G-CSF increase in response to severe inflammation in acute appendicitis [26].It has also been reported that inflammation-related products, such as IL-1 and lipopolysaccharide, stimulate G-CSF production by fibroblasts originating from the skin [27].In this study, we showed that G-CSF levels in wounded tissue reached a plateau as early as 3 hours after total skin excision, and spindle-to ovalshaped mesenchymal cells (probably fibroblasts) in the wounded tissue express G-CSF.The time gap between peak expression in the tissue and blood (6 hours) may be due to the time required to reach the blood stream.Thereafter, G-CSF levels in tissue and blood decreased rapidly.Such a transient but sensitive response in G-CSF production in the peripheral tissue suggests that G-CSF is an important cytokine connecting tissue injury to the organs that produce/store neutrophils.G-CSF can strongly mobilize not only granulocytes but also hematopoietic stem cells and progenitor cells from the bone marrow, and it is clinically used for the treatment of leukocytopenia and hematopoietic stem cell collection [28].It has also been shown that bone marrow cells have therapeutic potential for neovascularization and wound healing [29].These

G-CSF Administration Accelerates Cutaneous Wound Healing Accompanied With Increased Pro-Hyp Production In db/db Mice
Copyright: © 2017 Jimi, et al.

G-CSF Administration Accelerates Cutaneous Wound Healing Accompanied With Increased Pro-Hyp Production In db/db Mice
Copyright: © 2017 Jimi, et al.
previous findings suggested the possible utility of additional G-CSF supplementation.Therefore, many interventional trials and double-blinded trials using G-CSF had been performed, especially for acute myocardial infarction and diabetic foot ulcers [15,16].However, its effects are still controversial.It is known that capacity of neutrophils from diabetic wounds to produce reactive oxygen species is deteriorated [30], and that in diabetic patients with foot infections, reactive oxygen species production was restored by G-CSF administration [31].In addition, delayed wound healing in neutropenic patients was also improved by administration of G-CSF [14].This evidence suggests that G-CSF has beneficial effects on wound healing, which may involve augmentation of neutrophil numbers and/or function.
Plasma levels of G-CSF were significantly lower in the db/db mice than in wild-type mice.There was no difference in the initial number of WBCs between these mice, and the numbers did not change after skin excision.However, the number of infiltrated neutrophils in the granulation tissue of db/db mice was lower than that in the tissue of wild-type mice.These results suggest that infiltration capacity may be impaired in the diabetic mice.Moreover, in contrast to the wound healing in wild-type mice, wound healing in db/db mice was accelerated by topical rhG-CSF administration, which was accompanied by an increase in neutrophil infiltration.These findings indicate that neutrophil infiltration after wounding is an essential initial event for subsequent healing.However, why neutrophil infiltration was decreased or unchanged after rhG-CSF administration in the wildtype mice is unclear, although a negative feedback mechanism may be involved, since nondiabetic mice secrete enough G-CSF to induce neutrophil infiltration.
The roles of neutrophils in wounded tissue are not only killing bacteria and phagocytizing unfavorable materials but also degradation of the ECM.The ECM is simultaneously remodeled through the corporative action of neutrophils and fibroblasts [18].It has been reported that the degradation products of orally ingested collagen appear in the blood as dipeptides, such as Pro-Hyp and Hyp-Gly [22], and oral intake induced a physiologically favorable effect on the skin [23].In our study, Pro-Hyp levels in granulation tissue gradually increased during the 7 days after wounding, whereas Hyp-Gly levels did not.Different mechanisms may exist in the digestive tract and wound tissue in terms of the inherent metabolism in different organs.The specific production of Pro-Hyp in wounded tissue is first addressed in this study.We hypothesized that Pro-Hyp may be produced by collagen digestive enzymes in wounded tissue.Therefore, we examined MMPs expression.In analysis using serial sections, its expression pattern was quite similar to that of NIMP-R14-positive neutrophils in wound healing process with/without G-CSF administration.Infiltrated neutrophils primarily expressed MMP-8, but other types of cells, such as endothelial cells, also expressed MMPs.Pro-Hyp production was lower in the db/db mice, which was accompanied by a decrease in neutrophil infiltration and MMP-8 expression.Furthermore, Pro-Hyp production in the db/db mice was significantly increased by rhG-CSF administration, which was also accompanied by an increase in neutrophil infiltration and MMP-8 expression.These results suggest the involvement of Pro-Hyp in the wound healing process.
The granulation tissue that developed in normal and db/db mice after skin excision was quite different; normal granulation tissue was observed in wild-type mice, which has an organized fibrous structure.However, in db/db mice, many lipid-filled adipocytes with cholesterol clefts in the small amount of ECM were observed.Under diabetic conditions, the defect in inflammatory cell infiltration in the wounded tissue may be due to a decrease in not only their attachment to endothelial cells and migratory capacity [28] but also ECM quality, including glycation of collagens, which may acquire resistance to MMPs [32].
It has been shown that the growth of cultured dermal fibroblasts is stimulated by Pro-Hyp [24,33], and that keratin gene expression in keratinocytes cocultured with fibroblasts

G-CSF Administration Accelerates Cutaneous Wound Healing Accompanied With Increased Pro-Hyp Production In db/db Mice
Copyright: © 2017 Jimi, et al.
is also increased by the addition of Pro-Hyp [34].In addition, osteoblastic differentiation of MC3T3-E1, accompanied by an increase in Runx2 and Col1a1 gene expression, is induced by Pro-Hyp [35].These in vitro studies may also support our in vivo findings.

Conclusion
The present study shows that delayed wound healing in diabetic mice may be related to the decreases in G-CSF secretion, inflammatory cell infiltration, and Pro-Hyp production after wounding.Supplementation of diabetic wounds with G-CSF accelerated wound healing and Pro-Hyp production.Therefore, Pro-Hyp produced in wound tissue may play a central role in healing.The more precise mechanisms underlying G-CSF-induced wound healing should be clarified.Clarification of the role of Pro-Hyp in diabetic wound healing is also an important challenge, and one which we are currently pursuing for our next study.

Figure 1 :
Figure 1: G-CSF production in tissue and blood in mice following skin wounding.A: Time course of G-CSF production in wounded and normal subcutaneous tissues (upper panel) and blood (lower panel).Data are mean ± SEM (n = 3).B: Immunohistochemical detection of mouse G-CSF in the exposed muscle fascia 1 and 3 hours after total skin excision.G-CSF expression was detected in spindle-to oval-shaped cells at 3 hours post wounding.Bar = 50 µm.

Figure 4 :
Figure 4: Changes in collagen dipeptide content in wound tissue after skin excision.After skin excision, wounded tissue (filled columns) and normal subcutaneous tissue (open columns) was obtained from the same mice.The levels of the collagen dipeptides Hyp-Gly and Pro-Hyp in wild-type and db/db mice was measured on days 1, 3, 5, and 7 after skin excision.Normal tissue vs. wounded tissue; *: p < 0.05, **: p < 0.01.Data are mean ± SEM (n = 3).

Figure 5 :
Figure 5: The favorable effect of rhG-CSF administration on wound healing in db/db mice.After skin excision, rhG-CSF (filled columns) or vehicle alone (open column) was topically administered on the wound of wild-type and db/db mice.A: Hematological data, B: percent wound closure 7 days after wounding, and C: developed granulation tissue stained by MT staining.Data are mean ± SEM (n = 5).Bar = 20 µm.

Figure 6 :
Figure 6: The effect of rhG-CSF administration on neutrophil infiltration and collagen peptide production.After skin excision, rhG-CSF (filled) or vehicle alone (open) was administered once, and db/db (circles) and wild-type (squares) mice were sacrificed on days 3 and 5. A: The number of neutrophils infiltrated in the granulation tissue on days 3 and 5 was graded (see Materials and Methods).Data are mean ± SEM (n = 5).Representative photographs are also shown.Bar =100 µm.B: After skin excision, wounded (filled columns) and normal subcutaneous (open columns) tissue samples were obtained from the same mice.Hyp-Gly and Pro-Hyp content was measured on day 5. Normal tissue vs. wounded tissue; *: p < 0.05, **: p < 0.01.Data are mean ± SEM (n = 3).

Table 1 :
Physiological data after skin excision.