Objective. Contradictory results have emerged from recent studies describing low-level radiofrequency radiation (RFR) as a hazardous factor for the central nervous system while other studies have described such type of exposure as totally safe. In the brain, heat shock proteins (Hsps) are often induced under harmful conditions such as ischemia, traumatic injury, epilepsy, hyperthermia, drug administration, and neuro-degenerative diseases. Under those conditions, activation of the microglial cell population is often observed. In this work we studied the effect of two types of mobile phone signals, GSM-1800 and UMTS, on the expression of two major Hsps which are induced in the brain under harmful conditions, Hsp 70 and Hsp25. We also studied microglial activation in young adult (8 weeks) and elderly (17 months) Wistar rats. Groups of 8 animals were exposed or sham-exposed. Methods. Exposures were performed using a brain-averaged SAR level of 2 W/kg using a loop antenna, following two types of protocols: an acute exposure, with exposure lasting only two hours, and a subchronic exposure in which the animals were exposed for two hours per day, five days per week, during four weeks. In all cases, rats were progressively habituated to the exposure setup (rockets) over two weeks to avoid stress, and a sham group was exposed for each condition. Positive controls were performed by induction of a status epilepticus using a subcutaneous injection of kainic acid (10 mg/kg). At the end of exposure, rats were anesthetized with isofuran and perfused via the heart with 4%paraformaldehyde in phosphate buffer (0.1 M). Brains were kept in the fixative solution overnight at 4 C and then cryo-preserved in 20 % sucrose in phosphate buffer for 48 h at 4 C. The brains were removed and placed in fixative solution for 24 hours prior to be frozen in isopentane. Sections (10 μm thick) were prepared on slides in 3 separated zones taken approximately at -0,8, -3,8 and -8 mm posterior to bregma (zone 1, 2 and 3 respectively). Brain samples were coded and the analysis was performed in a blind manner. The sections were immunohistochemically stained with antibodies raised in rabbits against Hsp25 and against the inducible form of Hsp70. The whole microglial cell population was detected by its common cell surface glycoconjugates, which bind the plant Griffonia simplicifolia I isolectin (Iso 4). Activated microglia was specifically detected by immunostaining with a polyclonal antibody raised against a type-3 complement receptor (CD11b). The qualitative analyses were done on the cerebral cortex of the three zones, in the corpus callosus in zones 1 and 2, in the hippocampus in zone 2 and in the substantia nigra in zone 3. The significance of differences between groups was evaluated using the Kruskal-Wallis test for independent samples. Results. We have already completed the analyses of the effect of GSM-1800 sub-chronic exposure on hippocampus subfields, i.e. CA1, CA2, CA3, CA4 and Gyrus Dente of the young-adult and elderly rats. Microglial content and activation: • Compared to sham-exposed animals, IsoB4 labeling, which is related to the whole microglial population is not significantly changed in none any of the hippocampus subfields of the GSM-1800-exposed young-adult or elderly rats. Nevertheless, it is to be noted that kainate-induced microglial colonization of the Hippocampus is significantly decreased in the CA4 of the old animals compared to the young ones (P 0.05) Figure 1. Immunohistochemical detection of Hsp70 in the hippocampus of elderly rats • Compared to sham-exposed animals, CD11b detection related to the activated microglial subpopulation is not significantly changed in none of the subfields of the GSM-1800-exposed young-adult or elderly animals. Nevertheless, kainate-induced microglial activation is significantly decreased in the CA4 of the old animals compared to the young ones (P 0.05). Moreover, CD11b labeling was significantly higher in the CA2 and CA3 of the old animals control group, compared to the young ones (P ) 0.05. Hsp70 and Hsp25 expression: • Compared to sham-exposed rats, Hsp70 labeling was not significantly changed in any of the hippocampus subfields of the GSM-1800-exposed elderly animals (Fig.1). In young adults (Fig. 2) a slight but not significant increase was observed in the CA4 (P ), while other subfields remained unchanged. Nevertheless, kainate-induced expression of Hsp70 was significantly decreased in the CA4 and Dentate Gyrus of the old animals compared to the young ones with P 0.05 and 0.025 respectively. • Compared to sham-exposed rats, Hsp25 labeling was not significantly changed in any of the hippocampus subfields of the GSM-1800-exposed elderly animals (Fig. 3). In young adults (Fig. 4), a significant increase was observed in the CA1 (P 0.075), and CA4 (P 0.05) and GSM 1800 induced non significant increases in Hsp25 in the others subfields. Moreover, kainate-induced expression of Hsp25 was significantly decreased in the CA4 (P 0.05) of the old animals compared to the young ones. Conclusion. Our data show a clear increase in Hsp-25 protein in the CA4 and a less pronounced increase in the CA1 subfield of the hippocampus of the young-adult group after Figure 2. Immunohistochemical detection of Hsp70 in the hippocampus of young-adult rats sub-chronic exposures to GSM-1800. Such a phenomenon was not observed in the elderly population suggesting a higher resistance in that group. Compared to Hsp70, which is mostly secreted by neurons under stress conditions, the small size Hsp25 protein is principally secreted by microglia. Nevertheless in forebrain regions like hippocampus and corpus callosum, microglia was also reported to secrete Hsp70. The decline of immunological functions, which is always described in ageing, may contribute to the non-response to exposure to GSM-1800 we observed in elderly animals. In elderly, the relative resistance to kainateinduced colonization/activation of microglia associated with an absence of Hsps increase in the CA4 subfield of the hippocampus, may proceed from the same mechanisms. No definitive conclusion can be drawn before other parts of the brain will be analyzed. Comparison with UMTS exposed ratss and acutely-exposed animals will also be presented at the meeting