The procedures and protocol design described here were completely in accordance with the guidelines and regulations of the Animal Research Center of Mashhad Medical Science. Six healthy 3-year-old male beagle dogs, weighing between 18 and 20 kg were used in this study. They had intact teeth and the clinical and radiographic survey showed no pervious periodontal disease (Figure 1). Pre-study preparation included scaling and daily plaque control to obtain periodontal health.

Ghee obtained from the refined sheep butterfat, was heated to 70˚C and then mixed with the powdered rhizomes of Curcuma longa (Zingiberaceae-ginger

F. curcuma longa L. as authenticated by Dr. Davari Nejad, department of Botany, Ferdowsi University) at the same temperature under special circumstance. After cooling to 30˚C - 40˚C, the ointment was homogenized using a silverson homogenizer at 1500 rpm for 40 minutes. This ointment was applied with 2 different ratio including material A (powdered rhizomes of Curcuma longa 25% and ghee 75%) and material B (the same ratio of powdered rhizomes of Curcuma longa and ghee). Besides, hyaluronic acid (Fibrous Fleece of hyallofill, 9.5 cm × 9.5 cm, Convatec, US) was applied that makes a soft and sticky gel near the wound exudates due to the surgical incision.

All surgical procedures were performed under general anesthesia in sterile conditions. Medetomidine hydrochloride (0.05 mg/kg, Dormitor, Orion Corporation, Espoo, Finland) was administered intramuscularly as a premedication. General anesthesia was achieved using intravenous sodium thiopental injection (25 mg/kg, Ravonal, Tanabe. Osaka, Japan), and spontaneous breathing was monitored. Routine dental infiltration anesthesia (lidocaine hydrochloride 2%, 36 mg in each cartridge, with epinephrine 1:80,000, Astra Pharmaceuticals, Westborough, MA) was used at the surgical sites. Experimental periodontitis was induced by a ligature (diameter 0.25 mm, tomy International, Tokyo, Japan) tied around the cemento-enamel junction of maxillary and mandibular I₂, P₁, P₃ and M₁. Subsequently, no plaque control measures were performed for three months to allow accumulation of dental plaque on the wire. Acute inflammation was reduced after the ligature wire was removed, and a control plaque regimen, which consisted of a topical application of chlorhexidine (chlorhexidine gluconate 20%, ICI Pharmaceutical Group, Wilmington, DE; 40 ml of a 2% solution) three times a week for 2 weeks, was instituted. Pre-operative radiographs were exposed along with clinical assessment to confirm the presence of experimental periodontitis (Figure 2).

Following sulcular incision and elevation of buccal and lingual mucoperoistal flaps, all the granulation tissue was removed and the root surface was thoroughly scaled and planed using Gracy curettes and a chisel (Hu- Friedy co., Chicago, IL). On the basis of a simple randomized allocation, three materials including hyaluronic acid, material A and material B were applied topically in each two areas of six test regions and then the flaps were repositioned and sutured (Silk 0 - 3, 20 mm, Reverse cutting, Gore & Associated Inc., Flagstaff, AZ). Subsequently, all regions including untreated wounds in the two control areas were covered with periodontal pack (Coe-PACK, GC, Japan). In order to eliminate any effect of the materials on each others, a minimum space of one tooth was considered between the above mentioned areas. An aluminum foil with 0.2 mm thickness (Toos Corporation, Mashhad, Iran) was placed on the pack and was sutured so as to prevent the removal of pack from the region of surgery.

A soft diet (Pedigree, Kal Kan Foods Inc., Vernon, CA) was chosen to reduce potential mechanical interference with healing during food intake. All the dogs received antibiotics (Penicillin G, 200,000 U) and an analgesic (butrophanol tartrate injection) daily for two days. Besides, plaque control was maintained daily during healing phase by topical application with chlorhexidine.

Three dogs at day 7 and the three remaining ones two weeks post surgery were euthanized using an intravenous injection of concentrated of thiopental sodium. Tissue blocks including the experimental and control sites were dissected with the surrounding soft and hard tissues. The blocks were fixed in 10% buffered formalin for 3 to 5 days, decalcified in 5% formic acid for 8 to 10 weeks, trimmed, dehydrated, and embedded in paraffin (Paraffin tissue processor HMP 110, Microm International, Heidelberg, Germany). Sections of 4 µm thickness were cut in a mesio-distal plan and then were stained with hematoxylin and eosin and evaluated under light microscopy (Eclipse, E800, Nikon, Japan).

The parameters used for evaluation were inflammatory cells count, necrosis, acute hemorrhage, chronic hemorr- hage and edema as the inflammatory parameters and neovascularization, fibroblast proliferation and collagen density as the repair parameters of the healing process [20] .

Each of the parameters was scored from 0 to 3, with 0 = normal, 1 = mild increase, 2 = moderate increase, 3 = marked increase. Sparsely scattered neutrophils, plasma cells, macrophages, eosinophils and mast cells arranged in a random fashion were considered normal. Localization of 3 - 10, 11 - 30, or 31 or more cells in the wound tissue per 400× magnification field was considered a mild, moderate, and marked increased, respectively.

Sparsely scattered necrotic cell debris, extravasated erythrocytes (indicative of acute hemorrhage), and haemosiderin-laden macrophages (indicative of chronic hemorrhage) were considered a mild increase for each of these parameters. Focal dense accumulations of these components in the wound tissue were considered a moderate increase. Extensive tissue necrosis, massive hemorrhage involving the surrounding tissue, and the presence of greater than 30 haemosiderin-laden macrophages per 400× magnification fields were considered a marked increase for each of these parameters, respectively.

Slight separation of cells and collagen from one another in the wound tissue by non-staining or poorly staining acellular material was considered a mild edema. Separation of 30 - 50 µm by this acellular material was considered a moderate edema. While separation of greater than 50 µm was considered a marked edema.

Fibroblasts and collagen organized in a random pattern, typical of the normal connective tissue of gingiva, were considered normal. Localization of fibroblasts and capillary buds in the wound tissue in number of 3 - 10, 11 - 30, and greater than 31 per 400× magnification field were considered a mild, moderate, and a marked increase, respectively, for fibroblast and neovascularization. Scant collagen bundles slightly separating fibroblast in the wound tissues were considered a mild increase in collagen density. Somewhat dense accumulation of collagens between fibroblasts in the wound tissues was considered a moderate increase in collagen density. Extensive separation of fibroblast by abundant collagen bundles was considered a marked increase in collagen density.

All samples were evaluated for one parameter before proceeding to the next which allowed for more accurate comparison of individual parameters. Mean scores were used for objective evaluation of histological changes. The data was statistically analyzed by Kruskal-Wallis test using commercial software (GraphPad Prism Version 4, GraphPad Software, San Diego, CA). Differences were considered to be staisticaly significant at p < 0.05.

The result of this study showed that the local application of this novel ointment promote the healing process remarkably in comparison with hyaluronic acid and control group. There was a significant difference between material B and three remaining groups in terms of inflammatory cell count and edema (p = 0.005), acute and chronic hemorrhage and necrosis (p = 0.006) at day 7. However, on comparing materials A and B with hyaluronic acid, we observed considerable difference between fibroblast proliferation (p = 0.38) and collagen density (p = 0.004) one week post-operatively (Table 1).

Scores of all inflammatory and repair parameters, suggested remarkable difference at day 14 between the control group and the regions in which this three materials were applied (Table 2). Besides, there was meaningful difference between “CG” ointment and hyaluronic acid in terms of inflammatory cell count (p = 0.025), fibroblast proliferation (p = 0.27) and collagen density (p = 0.001) suggested the persisting inflammatory response led to relatively significant delay in the repair stage of healing two weeks after surgery in hyaluronic acid treated wounds.

Histological analysis of untreated wounds of control group at day 7 revealed abundant infiltration of inflammatory cells in the connective tissue supporting gingival epithelium, limited amount of vessels buds, extensive edema and poor repair parameters (Figure 3). However, we observed mild inflammatory response with scant chronic inflammatory cells among the gingival epithelial rete processes and deep in connective tissue, abundant new collagen synthesis with irregular orientation, moderate neovascularization which generally presented a relatively improved healing one week post operatively for wounds treated with material B (Figure 4).

Number in parentheses = standard error, Cell coun = Cell Count, Edema = Edema, Nec = Necrosis, Fibro = Fibroblast, Neovas = Neovascularization, Coll Den = Collagen Density.

Number in parentheses = standard error, Cell coun = Cell Count, Edema = Edema, Nec = Necrosis, Fibro = Fibroblast, Neovas = Neovascularization, Coll Den = Collagen Density.

On the other hand, the amount of inflammatory parameters particularly inflammatory cell count was significantly higher (p = 0.001) compared with material B which probably results from the higher percentage of ghee containing prostaglandin precursors in material A. In contrast, abundant of new vessels bud and collagen synthesis directly below the gingival epithelium was apparent. Wounds treated with hyluronic acid showed moderate to severe inflammatory response, tissue disruption in periodontal ligament space, cementum resorption which consequently presented better healing compared with control group one week post operatively (Figure 5).

In the histological survey at day 14, in the control group, we observed moderate to severe chronic inflammatory cells accumulation particularly in lamina properia directly below the gingival epithelium and in the periodontal ligament space with extensive tissue disruption as well as edema, weak scaffold of collagen fibers which generally indicated relatively poor healing that seemed insignificant compared with day 7. Wounds treated with hyaluronic acid presentedrelatively moderate persisting inflammatory response with highly vascularized connective tissue with extensive edema adjacent to periodontal pocket, week attachment of junctional epithelium to the root surface and moderate repair parameters (Figure 6). However, wounds treated with material A showed mild inflammatory response with sparsely scattered chronic inflammatory cells infiltration which appeared obviously less significant compared with day 7, dense accumulation of irregular-aligned new collagen fibers and proper attachment of connective tissue fibers throughout the root surface which indicated relatively complete healing of periodontium.

In the material B sites, we observed no signs of inflammatory response in the connective tissue and periodontal ligament, dense accumulation of relatively well-aligned new collagen fibers which attached throughout the root surface properly that in general terms has all the hallmarks of complete healing two weeks post operatively (Figure 7).