Our research group previously described wound healing of the palatal mucosa in Goto­Kakizaki (GK) rats

INTRODUCTION

Our research group previously described wound healing of the palatal mucosa in Goto­Kakizaki (GK) rats

that spontaneously developed type 2 dia­

betes when an antibiotic was not administered.

In this study, we investigated whether antibiotic ad­

ministration improves spontaneous wound healing in GK rats wounded in the same manner.

MATERIALS AND METHODS Experimental animals

Thirty­six experimental animals were used, com­

prising a control group of 18, 8­week­old male Wistar rats and a diabetes mellitus (DM) group of 18, 8­week­old male GK rats purchased from Shimizu Laboratory Supplies (Kyoto, Japan). This experiment was approved by the Osaka Dental Uni­

versity Animal Research Committee (Approval no : 15­02005) and complied with the guidelines for ani­

mal experiments.

Experimental methods Creation of the defect

Each of the experimental animals was weighed, and a cylindrical defect 1.5 mm in diameter was created in the maxillary right first molar region of the palatal mucosa using a biopsy punch (Dispos­

able Biopsy Punch

; Kai Industries, Gifu, Japan) (Fig. 1). All of the soft tissue, including the pe­

riosteum, was removed with the animal under inha­

lation anesthesia with isoflurane (Forane

; Abbott Japan, Tokyo, Japan). Determination of the location of the defect was described in Fig. 2. The defect was left open, and immediately after this surgery, 1.5 mg/kg of the antibiotic lincomycin hydrochloride hydrate (Lincocin

Injection ; Pfizer Japan, Tokyo, Japan) was administered intraperitoneally. Three animals in each group were sacrificed for histologi­

cal analysis using the methods described below at Hiroichi Orihara, Mamoru Uemura and Akimichi Takemura

Department of Anatomy, Osaka Dental University, 8-1, Kuzuhahanazono-cho, Hirakata-shi, Osaka 573-1121, Japan

We investigated the effect of antibiotic administration on spontaneous healing in Goto­

Kakizaki (GK) rats with type 2 diabetes and an experimentally created defect in the pala­

tal mucosa, comparing a control group of 18 male 8­week­old Wistar rats and a diabetes mellitus (DM) group of 18 male 8­week­old GK rats. In both groups, a cylindrical defect 1.5 mm in diameter was created in the maxillary right first molar region of the palatal mu­

cosa using a biopsy punch to remove all the soft tissue, including the periosteum. Imme­

diately after this surgery, an antibiotic was administered intraperitoneally. Three animals per group were sacrificed for histological testing at healing periods of 1, 3, 5, 7, 14 and 28 days after the surgery. It took longer for the wound to be covered with skin in the DM group than in the control group, and the appearance of granulation tissue and collagen fibers in the lamina propria was delayed. No sequestrum formation occurred in either group. The results of our comparative histological observations showed that spontane­

ous healing was delayed in GK rats compared with the controls. Our results suggest that if an antibiotic is administered, sequestrum formation may not occur. (J Osaka Dent Univ 2019 ; 53 : 15­23)

Key words : Wound healing ; Diabetes mellitus ; Rats ; Wistar ; Antibacterial drug

days 1, 3, 5, 7, 14 and 28 after the surgery.

Specimen preparation

Each of the experimental animals was weighed af­

ter fasting for 24 hours, and then heparin sodium (Novo­Heparin Injection 1000

; Mochida Pharma­

ceutical, Tokyo, Japan) was injected intraperito­

neally with the animal under isoflurane inhalation anesthesia. After 30 minutes, the animals were euthanized under isoflurane inhalation anesthesia with an intraperitoneal overdose of pentobarbital sodium (Nembutal

; Dainippon Sumitomo Pharma, Osaka, Japan). The thoracic cavity was opened, and blood was sampled from the left cardiac ventri­

cle. Next, a cannula was inserted from the left ven­

tricle into the ascending aorta, and the descending aorta was ligated. The animal was perfused with physiological saline and exsanguinated via the right cardiac atrium. After perfusion fixation with a 4%

formaldehyde solution via the ascending aorta, the palate was removed en bloc and fixed by immer­

sion in the same solution for 24 hours at 4° C.

The specimens were then decalcified with 10%

ethylenediaminetetraacetic acid (Ethylenediamin­

etetraacetic Acid Disodium Salt, 2­Hydrate

; Kishi­

da Chemical, Tokyo, Japan) in a microwave rapid sample processor (ML­77 ; Azumaya, Tokyo, Ja­

pan) for 10 days. After decalcification, 20 μm frozen frontal tissue slices of the palate were prepared dis­

tally from the mesial side of the maxillary right first molar using a cryostat (HM 500­OM ; Carl Zeiss Japan, Tokyo, Japan). Hematoxylin­eosin staining was performed using standard methods. The pre­

pared specimens were examined under an optical microscope fitted with a digital camera (NIS­

Elements BR

; Nikon, Tokyo, Japan), and digital photographs were obtained. When recording the state of wound healing, the cylindrical defect was described as a test cavity, and for convenience, it was divided into three equal parts described in Fig.

3.

Measurements of fasting blood glucose and he- moglobin A1c

The blood sampled from the left cardiac ventricle was centrifuged, and its serum was collected. Fast­

ing blood glucose was measured using assay re­

Fig. 1 Biopsy punch with 1.5 mm diameter.

Fig. 2 Determination of the site of the defect. The palatal midline is identified, and a straight line is drawn parallel to the midline and half way to the palatal side of the first mo­

lar (line a). Then a straight line (line b) is drawn joining the centers of the first molars, intersecting line a perpen­

dicularly. The intersection of lines a and b is designated as point c, which is used as the center of the defect.

Fig. 3 Regions of the experimental cavity. The cylin­

drical defect is described as a test cavity, which for convenience it is divided into three equal parts starting from the palatal surface comprising the top, middle and bottom of the cavity. The shallow depression made in the maxillary bone is termed the furrow.

group. Hemoglobin A1c (HbA1c) in blood samples was measured using assay reagents with the latex agglutination method.

statistical analysis, with the level of significance set at 1%.

RESULTS

Fasting blood glucose and HbA1c

Mean fasting blood glucose levels were 115.72±

14.44 mg/dL in the control group and 186.27±

17.10 mg/dL in the DM group, which represented a significant difference (Fig. 4). Mean HbA1c level was 4% in the control and DM groups.

Wound healing Day 1

In both the controls and DM group, the entire test cavity was filled with fibrin and clotted blood, and there was severe infiltration of inflammatory cells, principally neutrophils, throughout the test cavity (Fig. 5 A1, A2, B1 and B2).

Fig. 5 Histologic specimens at day 1 after surgery (A1) in the controls, (A2) higher magnifi­

cation of the square, (B1) in the diabetic group, and (B2) higher magnification of the square.

〇Colt,▲Fibrin,→Inflammatory cell.

Fig. 4 Comparison of fasting blood glucose levels in the type 2 spontaneous diabetes mellitus (DM) and controls (*p＜0.01).

Day 3 controls

The top of the cavity was covered by mucosal epi- thelium. No stratum corneum was observed. The middle of the cavity beneath the mucosal epithe- lium contained clotted blood composed of a dense fibrous network stained strongly by eosin. The clot- ted blood contained inflammatory cells (neutrophils, histiocytes, and lymphocytes) (Fig. 6 A1 and A2).

Day 3 DM group

There was no mucosal epithelium at the top of the cavity. As in the controls, clotted blood was com- posed of a dense fibrous network stained strongly by eosin. The clotted blood contained inflammatory cells (neutrophils, histiocytes, and lymphocytes) (Fig. 6 B1 and B2).

Day 5 controls

The mucosal epithelium was present at the top of the cavity, as well as an extremely thin stratum cor- neum. The mucosal epithelium in the center of the

top was slightly depressed. Granulation tissue con- taining numerous fibroblasts and a small amount of neovascularization was visible in the middle of the cavity immediately below the mucosal epithelium, as well as in the bottom of the cavity and the sul- cus. This granulation tissue contained a small num- ber of neutrophils and histiocytes, and an even a smaller number of lymphocytes. New small-dia- meter blood vessels were present throughout the cavity (Fig. 7 A1 and A2).

Day 5 DM group

The top of the cavity was not completely covered by the mucosal epithelium. There was severe in- flammatory cell infiltration of the middle of the cav- ity, including neutrophils and histiocytes. Irregular neovascularization was evident throughout the cav- ity (Fig. 7 B1 and B2).

Day 7 controls

The mucosal epithelium was continuous and promi-

Fig. 6 Histologic specimens at day 3 after surgery (A1) in the controls, (A2) higher magnifi- cation of the square, (B1) in the diabetic group, and (B2) higher magnification of the square.

nent from the top of the cavity, and a thin stratum corneum was evident in its superficial layer. There were no epithelial feet within the test cavity. Numer- ous fibroblasts and collagen fibers were present be- neath the mucosal epithelium throughout the cavity, and the granulation tissue was thicker on day 7 than on day 5. A few new small-diameter blood vessels were present in the sulcus (Fig. 8 A1 and B2).

Day 7 DM group

The mucosal epithelium was continuous and de- pressed in the top of the cavity. Compared to the control group, the DM group had a much thinner stratum corneum in the superficial layer of the mu- cosal epithelium. There were no epithelial feet within the test cavity. Granulation tissue containing numerous fibroblasts was present beneath the mu- cosal epithelium throughout the cavity. Mild inflam- matory cell infiltration caused by a small number of neutrophils and histiocytes and very few lympho-

cytes was present beneath the mucosal epithelium in the middle and bottom of the cavity. A few new small-diameter blood vessels were present in the sulcus (Fig. 8 B1 and B2).

Day 14 controls

The mucosal epithelium projected well beyond the top of the cavity across its entire width, and its thickness increased on day 14 compared with day 7. The superficial layer of the mucosal epithelium had numerous irregularities, and the stratum cor- neum was thicker on day 14 than on day 7. Evenly distributed rod-shaped epithelial feet were present on the mucosal epithelium ; they formed long, thin extensions into the lamina propria, reaching the middle of the cavity. In the lamina propria in the middle of the cavity, connective tissue papillae pro- jected to fit the irregularities in the mucosal epithe- lium. The lamina propria in the middle and bottom of the cavity and the sulcus contained numerous fi- broblasts and collagen fibers, as well as a large

Fig. 7 Histologic specimens at day 5 after surgery (A1) in the controls, (A2) higher magnifi- cation of the square, (B1) in the diabetic group, and (B2) higher magnification of the square.

〇Remains of clot,△Fibroblastic cell,→Inflammatory cell, v : Newborn blood vessel.

number of small-diameter blood vessels. There were almost no inflammatory cells within the lamina propria, and no sequestrum formation occurred in any animal (Fig. 9 A1 and A2).

Day 14 DM group

The mucosal epithelium reached the top of the cav- ity, and its superficial layer contained a stratum cor- neum that was thicker on day 14 than on day 7.

The mucosal epithelium had irregularly distributed epithelial feet that were shaped like cones, with their peaks pointing to the bottom of the cavity.

There were fewer epithelial feet in the DM group than in the control group, and they reached the middle of the cavity. Granulation tissue containing numerous fibroblasts and collagen fibers was pre- sent beneath the mucosal epithelium in the middle of the cavity. The bottom of the cavity and the sul- cus contained granulation tissue comprising numer- ous fibroblasts and a small number of collagen fi-

bers, with numerous small-diameter blood vessels.

Severe inflammatory cell infiltration was present, chiefly comprising neutrophils and histiocytes with very few lymphocytes, and this area was sur- rounded by granulation tissue. Some clotted blood was still present. No sequestrum formation oc- curred in any animal (Fig. 9 B1 and B2).

Day 28 controls

The mucosal epithelium projected onto the mucosal epithelial side. The mucosal epithelium and the stratum corneum were thinner on day 28 than on day 14. Evenly distributed rod-shaped epithelial feet were present on the mucosal epithelium ; these were greater in number on day 28 than on day 14, and they had become much shorter in length. The lamina propria was thick in the middle of the cavity and reached the upper margin of the cavity top, projecting onto the mucosal epithelium side. The lamina propria, which projected into the cavity top

Fig. 8 Histologic specimens at day 7 after surgery (A1) in the controls, (A2) higher magnifi- cation of the square, (B1) in the diabetic group, and (B2) higher magnification of the square.

→Collagen fiber.

and in the middle of the cavity, contained a small number of fibroblasts, numerous collagen fibers, and a large number of small-diameter blood ves- sels. The bottom of the cavity and the sulcus con- tained numerous fibroblasts and collagen fibers, and there were a few small-diameter blood vessels at the bottom of the cavity (Fig. 10 A1 and A2).

Day 28 DM group

The mucosal epithelium in the middle of the cavity projected well beyond the top of the cavity on the mucosal epithelial side. The mucosal epithelium and the stratum corneum had both thickened. Ir- regularly distributed rod-shaped epithelial feet of varying sizes were present on the mucosal epithe- lium. The lamina propria of the epithelial feet in the middle of the cavity projected well beyond the top of the cavity on the mucosal epithelial side. The lamina propria, which projected into the top, middle, and bottom of the cavity, contained numerous fibro-

blasts and collagen fibers (Fig. 10 B1 and B2).

DISCUSSION

Comparison of wound healing

Numerous comparative morphological studies of the organs of the oral cavity in GK and normal rats have been published.

Reports focusing on the oral mucosa in particular include a study by Yasuda et al.

on the palatal gingival mucosa in the maxil- lary first molar region and one by Akai et al.

on the palatal mucosa. Both found that the stratum cor- neum and the granular layer of the mucosal epithe- lium are thickened by diabetes. The connective tis- sue beneath the mucosal epithelium exhibited diabetes-induced atrophic changes, and diabetic microangiopathy of the capillary vessels was pre- sent. The current study addressed wound healing in the mucosal epithelium and lamina propria of diabetic rats compared with controls.

Fig. 9 Histologic specimens at day 14 after surgery (A1) in the controls, (A2) higher magni- fication of the square, (B1) in the diabetic group, and (B2) higher magnification of the square.

→Granulation tissue, A : Artery, N : Neuron-like substance.

Wound healing in the mucosal epithelium Herein, binding of the mucosal epithelium was evi- dent on day 3 in the controls, but not until day 7 in the DM group. This indicated that wound healing of the mucosal epithelium was delayed in the DM group compared with the controls. Epithelial cell mi- gration is necessary for the mucosal epithelium to cover a wound. In an in vitro system, high glucose concentrations severely inhibited oral epithelial cell migration compared with low glucose concentra- tions.

The expression of integrin, which is required for binding with the extracellular matrix, is en- hanced by high glucose concentrations, inhibiting cell migration.

This finding suggests that the cell migration required for wound coverage may be in- hibited when glucose concentration is high, delay- ing healing of the mucosal epithelium.

Wound healing in the lamina propria

In this study, granulation tissue appeared on day 5 in the control group and on day 7 in the DM group.

At high glucose concentrations, decreased capacity of glycosaminoglycan production by fibroblasts re- sults in inadequate granulation tissue formation.

This may have delayed the appearance of granula- tion tissue in the DM group compared with the con- trols.

Collagen fibers appeared on day 7 in the controls and on day 14 in the DM group. Fibroblast prolif- eration reportedly diminishes at high glucose con- centrations, and other studies have reported de- creased collagen production and elevated collage- nase activity in the gingival tissue of rats treated with streptozotocin.

This factor may have de- layed the appearance of collagen fibers in the DM group compared with the controls. At high glucose concentrations, fibroblast proliferation and collagen fiber productivity are thus reduced, and this may have delayed wound healing in the lamina propria.

Sequestrum formation

In a study by Ehara et al., sequestrum formation

Fig. 10 Histologic specimens at day 28 after surgery (A1) in the controls, (A2) higher magni- fication of the square, (B1) in the diabetic group, and (B2) higher magnification of the square.

Herein, the administration of an antibiotic immedi- ately after the surgery reduced the number of in- flammatory cells in the control group. The presence of inflammatory cell infiltration, chiefly by neutro- phils, in the DM group, which was more susceptible to infection, suggested that purulent inflammation persisted for longer in this group than in the con- trols. However, compared with the results of Ehara et al., these inflammatory cells disappeared more rapidly when an antibiotic was not administered.

The results of our comparative histological observa- tions showed that spontaneous healing was de- layed in GK rats compared with the controls after antibiotic administration. There is a high incidence of sequestrum formation in GK rats when antibiotics are not administered ; however, antibiotic admini- stration prevented sequestrum formation. Thus, our results suggest that when performing surgical den- tal procedures involving palatal mucosal defects in diabetic patients, an antibiotic should be adminis- tered, and wound healing should be carefully moni- tored.

This study was presented at the 66th Annual Meeting of the Japanese Association for Dental Research held on No- vember 17-18, 2018, in Sapporo, Hokkaido. We thank the staff of the Institute of Dental Research of Osaka Dental University for caring for the animals and processing the im- ages. We also express our appreciation for the valuable help and cooperation of the staff of the Department of Anatomy. The authors have no conflicts of interest to de- clare with respect to this paper.

REFERENCES

1. Goto Y, Kakizaki M. The spontaneous-diabetes rat : a model of noninsulin dependent diabetes mellitus. Proceed- ings of the Japan Academy, Series B1981; 57: 381-384.

2. Ehara D, Takemura A, Suwa F. Wound healing on palatine mucosa in the type 2 diabetes mellitus model rats.Journal of Osaka Dental University2011; 45: 37-53.

3. Slein M. D-glucose determination with hexokinase and glucose-6-phoshate dehydrogenase. In : Bergmeyer H, ed.

Pharmaceutical Science1995; 34: 125-136. (Japanese) 5. Hata Y, Suwa F, Imai H. Histological study of gingival epi-

thelium in a type II diabetes mellitus rat model. Journal of Osaka Dental University2004; 38: 23-29.

6. Kanemura N, Suwa F, Ueda M. Morphological study of gin- gival connective tissue papillae in type II diabetes model rats.Journal of Osaka Dental University2004; 41: 7-14.

7. Uemura M, Tamada Y, Suwa F. Morphological study of the connective tissue papillae and the capillary loops on the lin- gual dorsum in the type 2 diabetes mellitus model rats.

Okajimas Folia Anat Jpn2009; 85: 139-149.

8. Yasuda K, Uemura M, Suwa F. Morphological study of the palatal gingiva of the maxillary first molar in the type 2 dia- betes mellitus model rat. Okajimas Folia Anat Jpn 2011; 88: 65-74.

9. Akai K, Uemura M, Suwa F. Morphological study of the palatine mucosa in the type 2 diabetes mellitus model rat.

Journal of Osaka Dental University2013; 47: 185-194.

10. Morishita A, Uemura M, Suwa F. Morphological study of the submandibular gland in the type 2 diabetes mellitus model rat.Journal of Osaka Dental University2014; 48: 1-8.

11. Uemura M, Toda I, Kawashima W, Yoshimoto G, Fang YR, Xu YJ, Liu Y, Zhang L, Takemura A. Morphological study of the articular disc and capillary of the retrodiscal tissue in a type 2 spontaneous diabetes mellitus rat model. Okajimas Folia Anat Jpn2016; 92: 53-59.

12. Tazawa K. Effects of high glucose concentration on cell pro- liferation,in vitrowound healing and IL-1αproduction by the EGF-stimulated oral epithelial cell line, Ca 9-22.Nihon Univ Dent J 2002; 76: 25-30. (Japanese)

13. Nishimura F, Terranova V, Foo H, Kurihara M, Kurihara H, Murayama Y. Glucose-mediated alteration of cellular func- tion in human periodontal ligament cells.J Dent Res 1996; 75: 1664-1671.

14. Nishimura F, Takahashi K, Kurihara M, Takashiba S, Mu- rayama Y. Periodontal disease as a complication of diabe- tes mellitus.Ann Periodontol 1998; 3: 20-29.

15. Hehenberger K, Hansson A, Heilborn JD, Abdel-Halim SM, Ostensson CG, Brismar K. Impaired proliferation and in- creased L-lactate production of dermal fibroblasts in the GK- rat, a spontaneous model of non-insulin dependent diabetes mellitus.Wound Repair Regen 1999; 7: 65-71.

16. Hehenberger K, Hansson A. High glucose-induced growth factor resistance in human fibroblasts can be reversed by antioxidants and protein kinase C-inhibitors. Cell Biochem Funct1997; 15: 197-201.

17. Hehenberger K, Kratz G, Hansson A, Brismar K. Fibroblasts derived from human chronic diabetic wounds have a de- creased proliferation rate, which is recovered by the addition of heparin.J Dermatol Sci 1998; 16: 144-151.

18. Golub LM, Schneir M, Ramamurthy NS. Enhanced collage- nase activity in diabetic rat gingiva : in vitroandin vivo evi- dence.J Dent Res1978; 57: 520-525.