A new bioreactor system has been developed for in vivo NMR spectroscopy of microorganisms under defined physiological conditions. This cyclone reactor with an integrated NMR flow cell is continuously operated in the magnet of a 400‐MHz wide‐bore NMR spectrometer system. The residence times of medium and cells are decoupled by a circulation‐integrated cross‐flow microfiltration module to achieve higher cell densities as compared to continuous fermentations without cell retention (increase in cell density up to a factor of 10 in steady state). Volumetric mass transfer coefficients kLa of more than 1.0 s−1 are possible in the membrane cyclone reactor, ensuring adequate oxygen supply [oxygen transfer rate >15,000 mg O2 ·(L h)−1] of high cell densities. With the aid of the membrane cyclone reactor we were able to show, using continuous in vivo 31P NMR spectroscopy of anaerobic glucose fermentation by Zymomonas mobilis, that the NMR signal intensity was directly proportional to the cell concentration in the reactor. The concentration profiles of intracellular inorganic phosphate, NAD(H), NDP, NTP, UDP‐sugar, a cyclic pyrophosphate, two sugar phosphate pools, and extracellular inorganic phosphate were recorded after a shift from one steady state to another. The intracellular cyclic pyrophosphate had not been detected before in in vitro measurements of Zymomonas mobilis extracts due to the high instability of this compound. Using continuous in vivo 13C NMR spectroscopy of aerobic glucose utilization by Corynebacterium glutamicum at a density of 25 gcell dry weight · L−1, the membrane cyclone reactor served to measure the different dynamics of labeling in the carbon atoms of L‐lactate, L‐glutamate, succinate, and L‐lysine with a time resolution of 10 min after impressing a [1‐13C]‐glucose pulse.

A beta-glucosidase (E.C. 3.2.1.21) was isolated from the culture filtrate of fungus Trichoderma reesei QM 9414 grown in continuous culture with biomass retention. The crude extracellular enzyme preparation was fractionated by a three-step purification procedure [chromatography on Fractogel HW-55 (S) and Bio-Gel A 0.5 plus final preparative isoelectric focusing] to yield three beta-glucosidases with isoelectric points at pH 8.4, 8.0, and 7.4. Only one enzyme (pi 8.4) met the stringent criterion of being homogeneous according to titration curve analysis. This enzyme was then characterized not to be a glycoprotein, although the native protein contained 35% carbohydrate (as glucose). It was found to have an apparent molar mass of 7 x 10(4) g/mol (SDS-PAGE), exhibited its optimum activity towards cellobiose at pH 4.5 and 70 degrees C (30 min test), and lost less than 3% activity at 50 degrees C over a period of 7 h. The K(M) values towards cellobiose and p-nitrophenyl-beta-D-glucopyranoside were determined to be 0.5mM and 0.3mM, respectively. The enzyme hydrolyzed cellodextrins (cellotriose to cellooctaose) by sequentially splitting off glucose units from the nonreducing end of the oligomers. The extent of the observed transfer reactions varied with the initial substrate concentration. No enzyme activity towards microcrystalline cellulose or carboxymethylcellulose could be detected. The classification of the enzyme as beta-glucosidase or exo-beta-1,4-glucan glucohydrolase is discussed with respect to the exhibited hydrolytic activities.

No abstract

No abstract

A racemic mixture of a-hydroxy acids can be transformed to the desired optically active L-amino acid by means of a reaction route that goes through the intermediate formation of the corresponding a-keto acid. The continuous realization of this process is possible in a multienzyme membrane reactor, which has been described earlier as to its technical characteristics.'The feasibility of the process is demonstrated by the transformation of (LD)-hCtate via pyruvate to L-alanine, as first suggested by Mosbach ( See FIG. I). ' In the process, a form of NAD covalently bound to water-soluble polyethylene glycol' (PEG-20000-NAD) together with the three enzymes L-LDH, D-LDH and ALADH is retained by an ultrafiltration membrane. A necessary condition for the steady-state continuous realization of the process is that the concentration of the intermediate (pyruvate) in the reactor has to be held at some level above zero by continuous feeding. If this is not done, the soluble intermediate will be washed out from the reactor and the reaction path will be shifted entirely to the formation of the reduced coenzyme form, NADH, that is, the reaction will be stopped. On the other hand, it is conceivable that large intermediate concentrations will favor the reverse reaction to lactate and theoxidized coenzyme form, NAD'; thus, an optimal feed concentration of the intermediate will be expected.The dependence of the rates of the four individual reaction steps on substrate and product concentrations has been investigated by measuring initial rates and subsequently verified by measuring in the entire range of conversion. The data-fitting has been accomplished by a nonlinear optimization p r~c e d u r e .~The theoretical analysis below is based on the initial rate kinetic data.On the basis of the assumption that the sorption capacities of the enzymatic complexes are practically too low to have any measurable dynamic effect, the following model differential equations can be established for the dynamics of the reaction system under consideration: -d(lac) lac, -lac --dt 7 R , + Rz d(a1a) ala, -ala + R , -R4 --dt T 91(3)