0. Fractionation experiments
were controlled by malate dehydrogenase activity measurement, which is found only in the soluble fractions (Cox et al., 2005). For each sample, 15 μg of cytoplasmic/periplasmic and membrane fractions were loaded onto 12% SDS-PAGE gels. Immunoblotting was carried out as described previously by Guzzo et al. (1998). Transformed E. coli cells were used to determine the amount of denaturated E. coli soluble proteins, subjected to heat treatments at 55 °C lasting for 30 min, according to Yeh et al. (1997). Briefly, cytoplasmic and periplasmic proteins were quantified using a BioRad protein assay method with bovine serum albumin as a standard and diluted at 2 mg mL−1 in 20 mM Tris-HCl buffer, pH 8.0. Protein samples were heated at 55 °C for 30 min, Tanespimycin cost ABT 263 and the denaturated proteins were pelleted by centrifugation at 16 000 g for 10 min. The amount of proteins in the pellet and supernatant fractions was determined. The amount of aggregation in the soluble protein fraction of transformed E. coli cells was determined over a period of 1 h according to Leroux et al. (1997) and Yeh et al. (1997), with modifications. Cellular extracts at a concentration of 2 mg mL−1 were analysed by light scattering at 340 nm in a UV spectrophotometer (Uvikon
XS, Secomam) thermostated at 55 °C. All experiments were performed in 20 mM Tris-HCl buffer, pH 8.0, in a total volume of 2 mL. The control reaction was performed at 37 °C. The aggregation speed was determined for each analysis and its percentage of reduction was calculated using
E. coli cells transformed with the vector alone as a calibrator. Cellular extracts were treated with formaldehyde, to a final concentration of 1% (w/w), as described previously by Derouiche et al. (1995). Protein kinase N1 Cross-linking experiments were performed as described by Delmas et al. (2001). The membrane fluidity variations of transformed E. coli cells were measured according to Beney et al. (2004) after a heat shock treatment at 50 °C for 30 min. A one-way anova was performed using sigmastat® v. 3.0.1 software (SPSS Inc.), using the Holm–Sidak test (n=3, P<0.05) to locate significant differences. We generated three Lo18 proteins with amino acid substitutions, based on previous information relating to point mutations reported by Lentze et al. (2003) on the Bradyrhizobium japonicum HspH. Various amino acids in the α-crystallin domain were substituted (Fig. 1). The Y107A, V113A and A123S substitutions of Lo18 corresponded, respectively, to the F94A/D, L100A and A109S of HspH in B. japonicum (Lentze et al., 2003). We focused on these three amino acids because they presented different characteristics in HspH. F94A/D was unable to form dimers and resulted in a significant decrease in chaperone activity.