An analysis of the absorption spectra of natural yellow sapphires shows that the absorption is due to single Fe3+ ions and pairs of ions Fe3+-O2--Fe3+. Assignments of all levels from the 4G, 4P, and 4D have been made as well as four simultaneous electronic excitations of a pair of ions. The temperature dependence of the intensity of the pair absorption shows that one pair is mainly involved, probably the fourth- nearest neighbour pair. The value of the Heisenberg exchange parameter (J Sa. Sb) lies in the range 30-40 K. Estimates of this parameter for various excited states have been made. Synthetic yellow sapphires have spectra which duplicate the natural specimens. Blue and green natural sapphires have, in addition to the bands present in the spectra of yellow sapphires, spectra with bands at 17800 (┴C), 14200 (//c), 11500 (┴C), and 10000 om-1 (//c). The first two can be linked to Fe,Ti pairs and the evidence favours the nearest neighbour pair Ti4+-O2?Fe2+ for the 17800 cm-1 band and possibly the first neighbour pair for the 14200 cm-1 band. The second two can be produced in synthetic crystals by growth from fluoride-containing flux and the evidence supports an explanation involving second-nearest neighbour pairs Fe2+- O2--Fe3+ for the 11500 cm-1 absorption and first neighbours for the 10000 om-1 absorption.

Shiga toxin is a bacterial protein composed of one A and five B subunits. Its A chain possesses a protease sensitive loop (Cys-242-Cys-261) that is cleaved to produce an enzymatically active A1 domain and an A2 fragment associated with its B subunit pentamer. The proposed mode of action of the toxin is linked to its retrograde transport to the ER lumen followed by the translocation of its catalytic A1 chain to the cytoplasmic side of the ER membrane. A signal sequence-like domain (residues 220-246) which constitutes the C-terminus of the A1 chain precedes a region within the protease sensitive loop (residues 247-258) that contains known and putative cleavage sites. Two peptides corresponding to this C-terminus (residues 220-246) were chemically synthesized to investigate if this signal sequence-like domain can interact with membranes. Such a property may provide a clue to the mechanism of translocation of the A1 domain across the ER membrane. The first peptide represented the native sequence, which includes a naturally occurring cysteine at position 242 and provided a thiol moiety for the attachment of a spinlabel. A second peptide was designed to contain a single tryptophan residue (Ile232Trp) located within the hydrophobic core of the sequence which served as an intrinsic fluorescence probe. The interactions of both peptides with lipid vesicles were analyzed by circular dichroism, fluorescence, and EPR spectroscopy. The peptides lack structure in aqueous buffers and adopted an alpha-helical geometry when bound to negatively charged lipid vesicles. The addition of lipid vesicles to a solution of the tryptophan-containing peptide results in a blue shift in the wavelength of its fluorescence maxima as well as an increase in fluorescence intensity at 335 nm, suggesting that the hydrophobic core of this A1 peptide relocated to a nonpolar environment. EPR measurements of a proxyl-labeled analog of the peptide (introduced at Cys-242) indicated a decreased mobility of a fraction of the proxyl probe in the presence of lipid vesicles. At pH 7, the membrane-bound probe was completely reduced by ascorbate trapped inside vesicles but only partially reduced by ascorbate added outside the vesicles, suggesting that the C-terminal region of the peptide traversed the membrane bilayer or relocated close to the surface of its inner lipid leaflet. Finally, the peptide was shown to insert into lipid vesicles, causing the release of calcein at a high peptide:lipid ratio. These results suggest that the C-terminal tail of the A1 chain may anchor this domain into the ER membrane.

No abstract

No abstract

No abstract