5780447305

154 B. Rihn et al. IToxicology 109 (1996) 147-156

with, however very large individual variations. Although nose-only inhaled fibers are also cleared by the ciliated airways into the gut, the fiber distribution into the respiratory tract is

morę equal using such technology. Therefore,

•

intratracheal instillation does not allow the stan-dardization of the lung burdens and this route cannot yet be recommended for risk prediction to humans (Anonymous, 1992). Intraperitoneal and intrapleural, routes are freąuently used in rat intoxication studies (Stanton et al., 1981; Winkler and Ruttner, 1982; Brody et al., 1989; Pott et al., 1989; Szymaniec et al., 1989; Morgan et al., 1993). These studies were useful for the characterization of the "solid-state” carcinogenesis, especially in Stanton’s work (1981) relating the appearance of pleural sarcomas to the fiber size distribution of the inhaled particles.

Szymaniec et al. (1989) evinced the importance of the immunologie and the inflammatory re-sponses in crocidolite toxicity. In this latter study, the bronchoalveolar macrophages de-creased while the polymorphonucleated cells in-creased. Indeed 4 days after an injection of 2.5 mg of crocidolite, the macrophage percentage de-creased from 93.6% (control rats) to 64.3% and the polymorphonucleated cells inereased from 0.5% (control) to 33.3%. This may correspond to a primary unspecific inflammatory response con-trasting with our study which demonstrates per-sistent inflammatory response with an inereased presence of giant macrophages and lymphocytes. In this context, the TNF-a values observed are morę in accordance with a chronię than with an acute inflammation.

Although many strains of mice develop pul-monary tumors spontaneously or after asbestos inhalation (Reeves et al., 1974; Davis and Donal-son, 1993) few studies have been performed with mice. In such studies assessing the fiber carcino-genic potential, results obtained were inconclus-ive due to genetic susceptibilities of the various mice strains. This, however, should not be the case for short-term studies with genetically modi-fied mice like Big Blue™ (Stratagene™) or Muta™ Mouse (Hazleton™), transgenic, respect-ively for LacI and LacZ genes. They should allow the evaluations of the mutation ratę of a reporter gene after, for example, fiber inhalation.

In addition the biopersistence of the crocidolite fibers is easily studied with a mouse model. Although few crocidolite intoxications have been designed with mice, it has already been shown that this model gives pertinent results in fiber toxicity evaluation. Namely, anterior studies have allowed by scanning or transmission elec-tron microscopy, the study of (i) the internaliz-ation of fibers in the peritoneal macrophages (Winkler and Ruttner, 1982), (ii) the pattern of crocidolite deposition in mice lungs (Brody and Roe, 1983; Brody, 1993), (iii) the intracellular distribution of fibers (Johnson and Davies, 1981; McDonald and Kane, 1986). This last work had shown the cellular partial engulfment of the fibers even by the nucleus. We did not observe this phenomenon but as shown in Fig. 3B and C, we observed long fiber internalization in giant poly-nucleated macrophages. The different behaviors may be explained by the difTerent origins of macrophages (pulmonary and peritoneal) and therefore may limit the relevance of the peritoneal model.

Moalli et al. (1987) have developed an intraperitoneal assay in order to study the acute injury and regeneration of diaphragmatic me-sothelium in mice. This mechanistic study has shown the changes of mesothelial cells induced by long crocidolite fibers: celi deaths, loss of microvilli and appearance of bloated membranes. The regeneration of the epithelium was measured by [³H]thymidine in vivo incorporation into DNA and was seen 1-3 days after the injury to inerease by up to 4-fold above the control radio-activity following a crocidolite (Moalli et al., 1987) or a chrysotile intoxication (Brody et al., 1989). In conclusion, intraperitoneal routes in mice are tests that could be used for mechanistic studies but not as models of human carcinogenic-ity (Anonymous, 1992).

In order to extrapolate from animal studies to human studies, it is essential to use properly conducted particulate inhalations in rodents. In the past they were done in “walk-in"*chambers (Lynch et al., 1957; Reeves et al., 1974; Brody and

155