following correction should be noted. In Fig. 1b, the designed sequence of amino acids S 36 contains a typographical error in that only 35 of its 36 monomers appear. The missing amino acid is of type R and should be located between amino acids E (sixth) and G (seventh). The correct sequence is:Although this typographical error does not change the results presented in the paper because they were obtained with the correct number and sequence of amino acids, it constitutes a nuisance for anybody interested in reproducing our results, and we apologize for the inconvenience. We want to thank Prof. H. S. Chan, of the University of Toronto, for calling our attention to this error.Cell Biology. In the article "Quantitative analysis of biological membrane lipids at the low picomole level by nanoelectrospray ionization tandem mass spectrometry" by B. Brügger, G. Erben, R. Sandhoff, F. T. Wieland, and W. D.Lehmann, which appeared in number 6, March 18, 1997, of Proc. Natl. Acad. Sci. USA (94,(2339)(2340)(2341)(2342)(2343)(2344), Table 1

Heterogenous nucleation on small molecule crystals causes a monoclinic crystal form of bacteriorhodopsin (BR) in which trimers of this membrane protein pack differently than in native purple membranes. Analysis of single crystals by nano-electrospray ionization-mass spectrometry demonstrated a preservation of the purple membrane lipid composition in these BR crystals. The 2.9-Å x-ray structure shows a lipid-mediated stabilization of BR trimers where the glycolipid S-TGA-1 binds into the central compartment of BR trimers. The BR trimer͞lipid complex provides an example of local membrane thinning as the lipid head-group boundary of the central lipid patch is shifted by 5 Å toward the membrane center. Nonbiased electron density maps reveal structural differences to previously reported BR structures, especially for the cytosolic EF loop and the proton exit pathway. The terminal proton release complex now comprises an E194-E204 dyad as a diffuse proton buffer.

Proteins were digested in normal and highly 18O-enriched water using proteases commonly employed for protein sequencing. The extent of 18O incorporation into the resulting peptide fragments was characterized by electrospray ionization (ESI) and matrix-assisted laser desorption/ionization (MALDI) mass spectrometry (MS). The endoproteinases trypsin, Lys-C and Glu-C incorporate two atoms of 18O, resulting in a mass shift of +4 D for the peptide fragments. This indicates that, following proteolytic cleavage, peptide products continue to interact with these proteases and undergo repeated binding/hydrolysis cycles, resulting in complete equilibration of both oxygens in the carboxy terminus of the fragments with oxygen from solvent water. In contrast, chymotrypsin and Asp-N incorporate only one atom of 18O, resulting in a mass shift of +2 D, indicating that after the cleavage step these proteases do not accept the peptides as substrates. In addition, it was found that the proteases trypsin, Glu-C, and Lys-C exhibit minor or nontypical sequence specificities, resulting in unexpected peptide fragments. These fragments incorporate only one 18O atom, indicating that they do not undergo further binding/hydrolysis cycles with the enzyme. Thus, it is possible to discriminate between enzyme-typical peptide fragments with mass shifts of +4 D and nontypical fragments with mass shifts of only +2 D. Based on these observations, protein digest strategies are described for the generation of 1:1 ion doublets spaced either by 2 or 4 D. In addition, the C-terminus of a protein can be identified by the absence of an ion doublet in the corresponding peptide fragment. In protein sequencing by mass spectrometry, digest protocols generating ion doublets provide the most clear-cut analytical results for the recognition of ion series in ESI-MS/MS and MALDI post-source decay (PSD) product ion spectra. Only the mass spectrometric fragment ions of a C-terminal series show ion doublets spaced either by 2 or 4 D, whereas the fragment ions belonging to an N-terminal series remain unshifted. This assignment unequivocally reveals the direction of the identified sequence.

Quantitative analysis of time-resolved data in primary erythroid progenitor cells reveals that a dual negative transcriptional feedback mechanism underlies the ability of STAT5 to respond to the broad spectrum of physiologically relevant Epo concentrations.