Actinomycosis is a chronic granulomatous, suppurative and fibrosing disease, which can be isolated from various sites, including the oral mucosa, dental plaque, deep dental cavities, and periodontal pockets. Actinomycosis usually occurs in the cervicofacial region, the abdominal, thoracic and cerebral cavities, and in the periapical region of the teeth. As periapical actinomycosis is thought to be a sequel to dental caries, infected root canals are the primary portal of entry for the micro-organisms into the periapical tissues. Thus, A. israelii is involved in the pathogenesis of periapical infection.

Bacteria sometimes remain in the dentinal tubules, cementum and/or periapical region after conventional endodontic therapy, and consequently cause chronic inflammatory reactions in the periapical tissues. If an experimental animal model of this periapical disease were to be established, we could understand more clearly its pathogenesis. In general, however, it is difficult to induce experimental chronic inflammatory reactions such as those in actinomycosis. There are only a few published reports on experimental actinomycosis.

On the other hand, animal models have been often used to understand periapical pathosis. In these, the periapical tissues and/or apical pulpal tissue are exposed to the mouth, and the invading bacteria easily eliminated by acute inflammatory responses. Thus, the bacterial irritation was not persistent and the tissue responses in these models differ from those in clinical chronic periapical lesions. Recently, we reported an animal model of chronic actinomycotic infection in the mouse peritoneum and demonstrated that the use of an entrapping alginate gel is effective for the induction of chronic actinomycotic lesions. However, the peritoneal lesion is not satisfactory for an exact understanding of the clinical periapical chronic lesion, because the periapical tissue is surrounded by calcified bone, very different from the peritoneum. If actinomycosis could be produced more efficiently in tissues such as collagenous connective tissue and calcified bone, it would no doubt help to clarify the process of actinomycotic periapical pathosis. Furthermore, the development of chronic actinomycotic lesions has yet to be fully elucidated at the level of electron microscopy.

Our purpose now was to clarify the cellular responses to experimental actinomycotic lesions in the mouse cranium using light and transmission electron microscopy.

A. israelii (ATCC 10048) was grown in brain–heart infusion broth at 37°C for 5 days in an anaerobic globe box (Type AZ-Hard; Hirasawa, Tokyo, Japan) with an atmosphere of 10% CO₂ and 10% H₂ in nitrogen. The bacteria were harvested by centrifugation, washed twice with sterile saline and suspended in 1% solution of sodium alginate (Wako, Osaka, Japan). The suspension was agitated and then dropped into 0.07 M CaCl₂ solution to make gel particles (0.8–1 mm dia.) containing bacteria. The particles in the solution were stirred with a magnetic stirrer for 5 min to prevent their aggregation during curing. The particles were then collected in a stainless-steel wire net of 0.5 mm dia. mesh to eliminate smaller particles, washed twice with sterile saline and suspended in 5 ml of sterile saline. The particles contained 7.5×10⁶ bacteria/ml. Twenty-eight BALB/c mice, 30 days old, were used in this study. A volume (200 ?l) of alginate gel–bacterial suspension was injected into the subcutaneous tissue over the periosteum of the cranium. As a control, gel particles without bacteria were also prepared and injected into mice in the same manner.

Materials were collected at 1, 3, 7, 14, 21, 30, 60, and 120 days after the injection. At each stage, the chosen animals were anaesthetized by intraperitoneal injection of chloral hydrate (400 mg/kg) and perfused through the ascending aorta with physiological saline, followed by 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). The crania with induced lesions along with the adjacent tissue were removed and immersed in the same fixative overnight before demineralization in 5% EDTA at 4°C for 4 weeks. The demineralized tissues were dehydrated through a graded series of ethanol and embedded in paraffin. Paraffin sections, 6 ?m in thickness, were stained with haematoxylin–eosin, azan or Goodpasture Gram stain. Some specimens were osmificated in 1% OsO₄ in 0.1 M phosphate buffer, dehydrated and embedded in epoxy resin (Epon 812; TAAB, Aldermaston, U.K.). Semithin sections (1 ?m) were stained with toluidine blue. Ultrathin sections, double stained with uranyl acetate and lead citrate, were examined in a transmission electron microscope (H-7000; Hitachi, Tokyo, Japan).

The induced lesions were removed aseptically, homogenized in 1 ml of phosphate-buffered saline (pH 7.0) with a sterile glass homogenizer, and the dispersed samples diluted 10-fold (10¹–10⁶) with the same buffered saline. Portions of 0.1 ml of the dispersed original solution and each dilution were cultured on brain–heart infusion blood agar plates under anaerobic conditions in an anaerobic glove box at 37°C for 7 days. Bacterial genera and species were identified according to Bergey’s manual.

Serum IgG against A. israelii in the experimental mice was determined by ELISA, as described by Kettering et al. (1991). The reagents and the concentrations used were as follows: bacterial sonicate, 10 ?g protein/ml; mouse serum, 1:500; peroxidase conjugated antimouse IgG (Organon Teknika Corp., West Chester, PA), 1:3000 in Tris buffer; and 2,2?-azino-bis (3-ethylbenzothiazoline-6-sulphonic acid) diammonium salt (Wako) for visualization. The visualized colour was determined at 415 nm and 630 nm with an ELISA plate reader (MTP-100; Corona Electric, Katsuta, Japan).

Results obtained from ELISA were analysed by student t-test.

Actinomycotic lesions were induced in all the mice (n=21) by injecting A. israelii entrapped in alginate gel. The skin over the lesions was firm, elevated, and had lost hair, producing a red appearance. The general health of the mice was normal. In control mice (n=7), no actinomycotic lesions were recognizable. More than 10⁵ bacteria were recovered from the lesions and they were identified as A. israelii, indicating that the lesions were monoinfections. Titres of serum IgG against A. israelii were significantly raised in mice with the induced actinomycotic lesions, suggesting that humoral immunity was also induced by chronic infection with A. israelii.

After 1 day, in the core of the lesion, there was a mass of unstained amorphous material in which bacteria were scattered, corresponding to the injected alginate gel containing A. israelii. Numerous neutrophils surrounded the core structure. Fibrous structures intensely stained with haematoxylin were recognized in the areas infiltrated by neutrophils. The Gram-stained sections revealed that numerous Gram-positive bacteria were exclusively present in the alginate gel and partly in the masses of neutrophils. In semithin sections, masses of neutrophils were seen engulfing the bacteria. In the control groups, although numerous neutrophils surrounded the alginate gel, their numbers and aggregation were less than in the experimental group. Almost all the alginate gel disappeared during the preparation of the specimens.

Fig. 1. One day after injection. (a) Sagittal section of the experimental group, stained with haematoxylin and eosin (H-E). The alginate gel (*) is seen between the muscle layer (M) and periosteum. Magnification ×10; scale bar, 500 ?m. (b) Higher magnification of Fig. 1(a). The marked aggregation of neutrophils (*) is observed among the alginate gel enclosing haematoxylin stained fibrous structures (arrowheads). Magnification ×300; scale bar, 50 ?m. (c) Experimental group (Goodpasture Gram staining). Gram-positive bacteria (arrowheads) are observed in the alginate gel and the mass of neutrophils. Magnification ×250; scale bar, 20 ?m. (d) Semithin section from the experimental group (toluidine blue). Several neutrophils gather to form one mass enclosing bacteria (arrowheads). Magnification ×250; scale bar, 20 ?m. (e) Control (H-E staining). Although numerous neutrophils surround the alginate gel (*), their number and aggregation is less than in the experimental group. Note the haematoxylin-stained fibrous structures (arrowheads). Magnification ×300; scale bar, 50 ?m. CC, cranial cavity.