The proteome of growing cells of Bacillus subtilis was analyzed in order to provide the basis for its application in microbial physiology. DNA arrays were used to calculate the number of genes transcribed in growing cells. From the 4100 B. subtilis genes, 2515 were actively transcribed in cells grown under standard conditions. From these genes 1544 proteins should be covered by our standard gel system pI 4-7. Using this standard gel system and supplementary zoom gels (pI 5.5-6.7, 5-6, 4.5-5.5, and 4-5) 693 proteins which are expressed in growing cells were detected that cover more than 40% of the vegetative proteome predicted for this region. Particularly broad coverage and thus comprehensive monitoring will be possible for central carbohydrate metabolism (glycolysis, pentose phosphate shunt, and citric acid cycle), amino acid synthesis pathways, purine and pyrimidine metabolism, fatty acid metabolism, and main cellular functions like replication, transcription, translation, and cell wall synthesis. Comparing the theoretical pI and Mr values with those experimentally determined a reasonable correlation was found for the majority of protein spots. By a color code outliers with dramatic deviations in charge or mass were visualized that may indicate post-translational modifications. In addition to the cytosolic neutral and alkaline proteins, 130 membrane proteins were found relying on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) separation in combination with electrospray ionization-tandem mass spectrometry (ESI-MS/MS) techniques. The vegetative proteome containing 876 proteins in total is now ready for physiological applications. Two main proteome fractions (pI 4-7 and zoom gel pI 4.5-5.5) should be sufficient for such high-throughput physiological proteomics.

H2‐forming methylenetetrahydromethanopterin dehydrogenase (Hmd) is an unusual hydrogenase present in many methanogenic archaea. The homodimeric enzyme dubbed ‘metal‐free’ hydrogenase does not contain iron–sulfur clusters or nickel and thus differs from [Ni‐Fe] and [Fe‐Fe] hydrogenases, which are all iron–sulfur proteins. Hmd preparations were found to contain up to 1 mol iron per 40 kDa subunit, but the iron was considered to be a contaminant as none of the catalytic and spectroscopic properties of the enzyme indicated that it was an essential component. Hmd does, however, harbour a low molecular mass cofactor of yet unknown structure. We report here that the iron found in Hmd is most probably functional after all. Further investigation was initiated by the discovery that Hmd is inactivated upon exposure to UV‐A (320–400 nm) or blue‐light (400–500 nm). Enzyme purified in the dark exhibited an absorption spectrum with a maximum at approximately 360 nm and which mirrored its sensitivity towards light. In UV‐A/blue‐light the enzyme was bleached. The cofactor extracted from active Hmd was also light sensitive. It showed an UV/visible spectrum similar to that of the active enzyme and was bleached upon exposure to light. Photobleached cofactor no longer had the ability to reconstitute active Hmd from the apoenzyme. When purified in the dark, Hmd consistently contained per monomer about one Fe, which was tightly bound to the cofactor. The iron was released from the enzyme and from the cofactor upon light inactivation. Hmd activity was inhibited by high concentrations of CO and CO protected the enzyme from light inactivation indicating that the iron in Hmd is of functional importance. Therefore, reference to Hmd as ‘metal‐free’ hydrogenase is no longer appropriate.

We use microfluidic chips to detect the biologically important cytokine tumor necrosis factor alpha (TNF- alpha) with picomolar sensitivity using sub-microliter volumes of samples and reagents. The chips comprise a number of independent capillary systems (CSs), each of which is composed of a filling port, an appended microchannel, and a capillary pump. Each CS fills spontaneously by capillary forces and includes a self-regulating mechanism that prevents adventitious drainage of the microchannels. Thus, interactive control of the flow in each CS is easily achieved via collective control of the evaporation in all CSs by means of two Peltier elements that can independently heat and cool. Long incubation times are crucial for high sensitivity assays and can be conveniently obtained by adjusting the evaporation rate to have low flow rates of approximately 30 nL min(-1). The assay is a sandwich fluorescence immunoassay and takes place on the surface of a poly(dimethylsiloxane)(PDMS) slab placed across the microchannels. We precoat PDMS with capture antibodies (Abs), localize the capture of analyte molecules using a chip, then bind the captured analyte molecules with fluorescently-tagged detection Abs using a second chip. The assay results in a mosaic of fluorescence signals on the PDMS surface which are measured using a fluorescence scanner. We show that PDMS is a compatible material for high sensitivity fluorescence assays, provided that detection antibodies with long excitation wavelength fluorophores ( > or =580 nm) are employed. The chip design, long incubation times, proper choice of fluorophores, and optimization of the detection Ab concentration all combine to achieve high-sensitivity assays. This is exemplified by an experiment with 170 assay sites, occupying an area of approximately 0.6 mm(2) on PDMS to detect TNF-alpha in 600 nL of a dendritic cell (DC) culture medium with a sensitivity of approximately 20 pg mL(-1)(1.14 pM).

The hmd gene, which encodes the metal-free hydrogenase in methanogenic archaea, was heterologously expressed in Escherichia coli. The overproduced enzyme was completely inactive. High activity could, however, be induced by the addition of ultrafiltrate from active enzyme denatured in 8 M urea. The active fraction in the ultrafiltrate was heat-labile and migrated on gel filtration columns with an apparent molecular mass well below 1000 Da. ß 2000 Federation of European Biochemical Societies. Published by Elsevier Science B.V. All rights reserved.

Four different dehydrogenases are known that catalyse the reversible dehydrogenation of N5,N10-methylenetetrahydromethanopterin (methylene-H4MPT) or N5,N10-methylenetetrahydrofolate (methylene-H4F) to the respective N5,N10-methenyl compounds. Sequence comparison indicates that the four enzymes are phylogenetically unrelated. They all catalyse the Re-face-stereospecific removal of the pro-R hydrogen atom of the coenzyme's methylene group. The Re-face stereospecificity is in contrast to the finding that in solution the pro-S hydrogen atom of methylene-H4MPT and of methylene-H4F is more reactive to heterolytic cleavage. For a better understanding we determined the conformations of methylene-H4MPT in solution and when enzyme-bound by using NMR spectroscopy and semiempirical quantum mechanical calculations. For the conformation free in solution we find an envelope conformation for the imidazolidine ring, with the flap at N10. The methylene pro-S C-H bond is anticlinal and the methylene pro-R C-H bond is synclinal to the lone electron pair of N10. Semiempirical quantum mechanical calculations of heats of formation of methylene-H4MPT and methylene-H4F indicate that changing this conformation into an activated one in which the pro-S C-H bond is antiperiplanar, resulting in the preformation of the leaving hydride, would require a deltadeltaH(f) of +53 kJ mol-1 for methylene-H4MPT and of +51 kJ mol-1 for methylene-H4F. This is almost twice the energy required to force the imidazolidine ring in the enzyme-bound conformation of methylene-H4MPT (+29 kJ mol-1) or of methylene-H4F (+35 kJ mol-1) into an activated conformation in which the pro-R hydrogen atom is antiperiplanar to the lone electron pair of N10. The much lower energy for pro-R hydrogen activation thus probably predetermines the Re-face stereospecificity of the four dehydrogenases. Results are also presented explaining why the chemical reduction of methenyl-H4MPT+ and methenyl-H4F+ with NaBD4 proceeds Si-face-specific, in contrast to the enzyme-catalysed reaction.