Effects of charged water-soluble polymers on the stability and activity of yeast alcohol dehydrogenase and subtilisin Carlsberg.

Remarkable increases in enzyme catalytic stability resulting from addition of charged water-soluble polymers have recently been reported, suggesting that use of these polymers may be an attractive general strategy for enzyme stabilization. To test the proposed hypothesis that coulombic forces between water-soluble polymers and enzymes are primarily responsible for enzyme stabilization, we examined the catalytic stability and activity of two enzymes in the presence of polymers differing in net charge. All polymers tested increased enzyme lifetimes, regardless of their net charge, suggesting that stabilization of these enzymes by water-soluble polymers is not solely dependent on simple electrostatic interactions between the polymers and enzymes.

In vivo screening of haloalkane dehalogenase mutants.

Haloalkane dehalogenase (Dh1A) from Xanthobacter autotrophicus GJ10 catalyzes the dehalogenation of short chain primary alkyl halides. Due to the high Km and low turnover, wild type Dh1A is not optimal for applications in bioremediation. We have developed an in vivo screen, based on a colorimetric pH indicator, to identify Dh1A mutant with improved catalytic activity. After screening 50,000 colonies, we identified a Dh1A mutant with a lower pH optimum. Sequence analysis of the mutant revealed a single substitution, alanine 149 to threonine, which is located close to the active site of Dh1A. Replacement of alanine 149 via site-directed mutagenesis with threonine, serine or cysteine retained the mutant phenotype. Other substitutions at position 149 show little or no activity.

Haloalkane dehalogenases: steady-state kinetics and halide inhibition.

The substrate specificities and product inhibition patterns of haloalkane dehalogenases from Xanthobacter autotrophicus GJ10 (XaDHL) and Rhodococcus rhodochrous (RrDHL) have been compared using a pH-indicator dye assay. In contrast to XaDHL, RrDHL is efficient toward secondary alkyl halides. Using steady-state kinetics, we have shown that halides are uncompetitive inhibitors of XaDHL with 1, 2-dichloroethane as the varied substrate at pH 8.2 (Cl-, Kii = 19 +/- 0.91; Br-, Kii = 2.5 +/- 0.19 mM; I-, Kii = 4.1 +/- 0.43 mM). Because they are uncompetitive with the substrate, halide ions do not bind to the free form of the enzyme; therefore, halide ions cannot be the last product released from the enzyme. The Kii for chloride was pH dependent and decreased more than 20-fold from 61 mM at pH 8.9 to 2.9 mM at pH 6.5. The pH dependence of 1/Kii showed simple titration behavior that fit to a pKa of approximately 7.5. The kcat was maximal at pH 8.2 and decreased at lower pH. A titration of kcat versus pH also fits to a pKa of approximately 7.5. Taken together, these data suggest that chloride binding and kcat are affected by the same ionizable group, likely the imidazole of a histidyl residue. In contrast, halides do not inhibit RrDHL. The Rhodococcus enzyme does not contain a tryptophan corresponding to W175 of XaDHL, which has been implicated in halide ion binding. The site-directed mutants W175F and W175Y of XaDHL were prepared and tested for halide ion inhibition. Halides do not inhibit either W175F or W175Y XaDHL.

Biodegradation of paint stripper solvents in a modified gas lift loop bioreactor.

Paint stripping wastes generated during the decontamination and decommissioning of former nuclear facilities contain paint stripping organics (dichloromethane, 2-propanol, and methanol) and bulk materials containing paint pigments. It is desirable to degrade the organic residues as part of an integrated chemical-biological treatment system. We have developed a modified gas lift loop bioreactor employing a defined consortium of Rhodococcus rhodochrous strain OFS and Hyphomicrobium sp. DM-2 that degrades paint stripper organics. Mass transfer coefficients and kinetic constants for biodegradation in the system were determined. It was found that transfer of organic substrates from surrogate waste into the air and further into the liquid medium in the bioreactor were rapid processes, occurring within minutes. Monod kinetics was employed to model the biodegradation of paint stripping organics. Analysis of the bioreactor process was accomplished with BIOLAB, a mathematical code that simulates coupled mass transfer and biodegradation processes. This code was used to fit experimental data to Monod kinetics and to determine kinetic parameters. The BIOLAB code was also employed to compare activities in the bioreactor of individual microbial cultures to the activities of combined cultures in the bioreactor. This code is of benefit for further optimization and scale-up of the bioreactor for treatment of paint stripping and other volatile organic wastes in bulk materials.

Magnetic resonance imaging study of the rat cerebral ventricular system utilizing intracerebrally administered contrast agents.

Magnetic resonance imaging (MRI) was employed to study the rat brain in conjunction with intracerebral (ic) injection of three contrast agents: GdHAM, GdDPTA, and MnCl2. The results demonstrate several advantages of ic administration of MRI contrast agents over the other routes of injection in examining CSF dynamics and brain ventricular structure. Apparent affinity of the luminal ventricular wall of the brain for positively charged GdHAM and Mn2+ ions is observed, presumably reflecting the presence of negatively charged wall components. Respiratory distress caused by (intravenous) injection of GdHAM was found to be minimized in the case of ic injections. Time-dependent changes in observed contrast indicate that diffusive processes rather than flow of CSF play a dominant role in distributing the contrast agents. Possible applications of this approach in brain research are discussed.

Ordered synthesis and mobilization of glycogen in the perfused heart.

The molecular order of synthesis and mobilization of glycogen in the perfused heart was studied by 13C NMR. By varying the glucose isotopomer ([1-13C]glucose or [2-13C]glucose) supplied to the heart, glycogen synthesized at different times during the perfusion was labeled at different carbon sites. Subsequently, the in situ mobilization of glycogen during ischemia was observed by detection of labeled lactate derived from glycolysis of the glucosyl monomers. When [1-13C]glucose was given initially in the perfusion and [2-13C]glucose was given second, [2-13C]lactate was detected first during ischemia and [3-13C]lactate second. This result, and the equivalent result when the glucose labels were given in the reverse order, demonstrates that glycogen synthesis and mobilization are ordered in the heart, where glycogen is found morphologically only as beta particles. Previous studies of glycogen synthesis and mobilization in liver and adipocytes [Devos, P., & Hers, H.-G. (1979) Eur. J. Biochem. 99, 161-167; Devos, P., & Hers, H.-G. (1980) Biochem. Biophys. Res. Commun. 95, 1031-1036] have suggested that the organization of beta particles into alpha particles was partially responsible for ordered synthesis and mobilization. The observations reported here for cardiac glycogen suggest that another mechanism is responsible. In addition to examining the ordered synthesis and mobilization of cardiac glycogen, we have selectively monitored the NMR properties of 13C-labeled glycogen synthesized early in the perfusion during further glycogen synthesis from a second, differently labeled substrate. During synthesis from the second labeled glucose monomer, the glycogen resonance from the first label decreased in integrated intensity and increased in line width.(ABSTRACT TRUNCATED AT 250 WORDS)

13C nuclear magnetic resonance evidence for gamma-aminobutyric acid formation via pyruvate carboxylase in rat brain: a metabolic basis for compartmentation.

The compartmentation of amino acid metabolism is an active and important area of brain research. 13C labeling and 13C nuclear magnetic resonance (NMR) are powerful tools for studying metabolic pathways, because information about the metabolic histories of metabolites can be determined from the appearance and position of the label in products. We have used 13C labeling and 13C NMR in order to investigate the metabolic history of gamma-aminobutyric acid (GABA) and glutamate in rat brain. [1-13C]Glucose was infused into anesthetized rats and the 13C labeling patterns in GABA and glutamate examined in brain tissue extracts obtained at various times after infusion of the label. Five minutes after infusion, most of the 13C label in glutamate appeared at the C4 position; at later times, label was also present at C2 and C3. This 13C labeling pattern occurs when [1-13C]glucose is metabolized to pyruvate by glycolysis and enters the pool of tricarboxylic acid (TCA) intermediates via pyruvate dehydrogenase. The label exchanges into glutamate from the TCA cycle pool through glutamate transaminases or dehydrogenase. After 30 min of infusion, approximately 10% of the total 13C in brain extracts appeared in GABA, primarily (greater than 80%) at the amino carbon (C4), indicating that the GABA detected is labeled through pyruvate carboxylase. The different labeling patterns observed for glutamate and GABA show that the large detectable glutamate pool does not serve as the precursor to GABA. Our NMR data support previous experiments suggesting compartmentation of metabolism in brain, and further demonstrate that GABA is formed from a pool of TCA cycle intermediates derived from an anaplerotic pathway involving pyruvate carboxylase.

Use of multiple 13C-labeling strategies and 13C NMR to detect low levels of exogenous metabolites in the presence of large endogenous pools: measurement of glucose turnover in a human subject.

A significant problem which may be encountered in 13C NMR studies of metabolism is the contribution that background levels of 13C may make to the observed spectra when low or tracer levels of the 13C label are used. We propose that the introduction of two or more labeled sites in the same tracer molecule is an effective strategy for eliminating or reducing this difficulty and demonstrate its feasibility in an isotope dilution study of glucose turnover in a human volunteer. This approach has two significant advantages over the more common use of a singly enriched labeling strategy: (i) as a consequence of the scalar coupling interactions, multiple-labeled metabolites will yield spectra distinct from those containing natural abundance 13C, and (ii) at a 99% level of enrichment for the precursor, concentration levels which are approximately 1% of the endogenous pools can be detected with approximately equal sensitivity. As a demonstration of this strategy, glucose production in a human subject was determined by continuous infusion of tracer levels of [U-13C6]glucose over a 4-h period and subsequent analysis of plasma levels of the tracer in vitro by NMR. Mass spectroscopy was used on the same samples to provide a basis for comparison of the precision and accuracy of the NMR technique. The results demonstrate the feasibility of the multiply labeled approach for detection by NMR of tracer amounts of label in the presence of a much larger endogenous pool of glucose. The NMR and mass spectrometric data gave quantitatively identical results for the glucose production rate demonstrating that equivalent data may be obtained by both methods.(ABSTRACT TRUNCATED AT 250 WORDS)

Rates of glycolysis and glycogenolysis during ischemia in glucose-insulin-potassium-treated perfused hearts: A 13C, 31P nuclear magnetic resonance study.

The effects of 11.7 mM glucose, insulin, and potassium (GIK) on metabolism during ischemia were investigated in the perfused guinea pig heart using magnetic resonance spectroscopy. Intracellular metabolites, primarily glycogen and glutamate, were labeled with 13C by addition of [1-13C]glucose to the perfusate during a normoxic, preischemic period. 13C and 31P NMR spectroscopy was used to observe the metabolism of 13C-labeled metabolites simultaneously with high-energy phosphorus metabolites and pH. The extent of acidosis and the rate and amount of labeled lactate accumulation during ischemia were the same in control (3 mM glucose + insulin) and GIK-treated hearts. In contrast, the rate of labeled glycogen mobilization during ischemia in GIK-treated hearts was one third the rate observed in control hearts. These observations suggest that GIK decreased the rate of glycogenolysis during ischemia without affecting the rate of glycolysis. We propose that glucose contributed as a glycolytic substrate to a greater extent during ischemia in GIK-treated hearts than in hearts perfused with 3 mM glucose and insulin. The glycogen-sparing effect of GIK demonstrated in these studies could delay the onset of ischemic damage in a clinical setting by prolonging the availability of glycolytic substrate necessary for production of high-energy phosphate.

Metabolic consequences of anoxia in the isolated, perfused guinea pig heart: anaerobic metabolism of endogenous amino acids.

The metabolic consequences of anoxia in the isolated, perfused guinea pig heart were examined by 13C NMR spectroscopy of 13C-labeled metabolites in situ. Upon addition of [3-13C]pyruvate to the perfusate during normoxic conditions, label is detected in several metabolites, including alanine (C3), glutamate (C2, C3, and C4), and aspartate (C2 and C3), reaching steady state levels 10-15 min after the labeled precursor reaches the heart. During anoxia, the label in glutamate and aspartate decreases and label appears in C2(3) of succinate. This real-time observation demonstrates that in the isolated intact heart, anaerobic metabolism of the amino acids aspartate and glutamate to succinate occurs. These pathways, which were first noted to occur in skeletal muscle of diving mammals, may provide a mechanism supplemental to glycolysis for the production of nucleoside triphosphates during periods of anoxia.

13C NMR studies of the thermal properties of a model high density lipoprotein. Apolipoprotein A-I-dimyristoylphosphatidylcholine complex.

The most abundant lipid and protein components of human plasma high density lipoproteins are phosphatidylcholine and apolipoprotein A-I (A-I). Under appropriate conditions, A-I spontaneously associates with dimyristoylphosphatidylcholine (DMPC) to quantitatively form a lipid-protein complex with a DMPC/A-I molar ratio of 100:1. Differential scanning calorimetry of this complex reveals two broad thermal transitions centered at approximately 27 and 72 degrees C. 13C NMR spectra of the complex have been obtained above, at, and below the lower transition temperature. The 13C resonance arising from the 3' carbon of the fatty acyl chains is a doublet, split by approximately 0.2 ppm, suggesting that the 3' carbon nuclei occupy two magnetically inequivalent sites. By replacing the sn-2 fatty acyl chain with myristate selectively 13C-enriched at carbon 3', we have shown that the splitting is, in fact, a result of magnetic inequivalence of the two sites and have assigned the lower field resonance to the 3' carbon nucleus of the sn-2 chain. The temperature dependence of the NMR relaxation rates indicates that the endothermic transition at 27 degrees C is associated with increased motional freedom for the phospholipids within this complex. The temperature dependence of the fatty acyl chain methylene 13C chemical shifts suggests that the population of gauche conformers increases above the transition temperature. These dynamic and conformational changes are characteristic of gel----liquid crystalline phase transitions observed in pure phospholipid systems. For the DMPC-A-I complex at 37 degrees C, the chemical shifts of the fatty acyl C 4'- 11' methylene envelope and of the C 7' and C 13' resonances occur significantly downfield from the corresponding chemical shifts for the DMPC vesicle. These results suggest that the apoprotein rigidifies the acyl chains by increasing their number of trans conformers.

Carbon-13 nuclear magnetic resonance studies of cholesterol-egg yolk phosphatidylcholine vesicles.

Proton-decoupled natural-abundance (13)C nuclear magnetic resonance spectra at 63 KG were obtained for isolated single-bilayer egg yolk phosphatidylcholine-cholesterol vesicles containing a variable phospholipid/cholesterol ratio. Numerous well-resolved singly carbon resonances of phospholipid and cholesterol carbons were observed. Carbon resonances from different parts of the phospholipid show markedly different behavior as a function of cholesterol content of the vesicles. The line widths of resonances for carbon atoms in the head-group region and the sn-3 carbon of the phospholipid glycerol backbone are relatively independent of cholesterol content. In contrast, resonances from the sn-1 and sn-2 carbon atoms of the glycerol backbone and the envelopes containing the olefinic and aliphatic carbon resonances of the fatty acyl chains of the phospholipids broaden markedly with increasing content of cholesterol. The most prominent cholesterol ring resonance is that for C6. This is, in part, due to its location in a clear window of the spectrum, where it is unobscured by interfering phospholipid resonances. However, resonances for cholesterol ring carbons C9 and C14, 17, which should also appear in clear regions of the spectrum, are not observable. It is suggested that these resonances are broadened by dipolar interactions with neighboring protons. Anisotropic rotation of the cholesterol molecule about its long axis is suggested to be the major mechanism responsible for decreased dipolar interactions for the C6 carbon, while retaining the large dipolar coupling of the C9 and C14, 17 carbons with their neighboring protons. The temperature dependence of spectra of single-bilayer phosphatidylcholine vesicles containing epicholesterol (alpha-cholesterol) is different from that of corresponding cholesterol-containing vesicles. Resonances from the C9 and C14, 17 carbons of epicholesterol in EYPC vesicles are detectable at 35 degrees C whereas the corresponding resonances from beta-cholesterol are not. These data suggest that in vesicles the rotation of cholesterol is more anisotropic than that of epicholesterol and that the stereochemistry of the C3 hydroxyl group of cholesterol is at least partly responsible for the highly anisotropic rotation of the steroid ring within the bilayer.

Theory for nuclear magnetic relaxation of probes in anisotropic systems: application of cholesterol in phospholipid vesicles.

The nuclear magnetic relaxation of a nucleus in a cylindrical probe embedded in a bilayer vesicle is considered. The probe is assumed to diffuse freely about its unique (C infinity) symmetry axis with an effective correlation time tau parallel, and the C infinity axis moves in a potential which is azimuthally symmetric about a director, with an effective correlation time tau perpendicular. The overall isotropic rotational correlation time of the membrane is tau M. Dipolar relaxation and, in the special case that the relevant tensors are axially symmetric, quadrupolar and chemical shift anisotropy relaxation are treated. An expression for the appropriate correlation function is derived which depends on the above effective correlation times, on the order parameter of the C infinity axis of the probe, and on the angle (beta) which in the case of dipolar relaxation of a protonated 13 C nucleus is between the 13 C-H vector and the C infinity axis of the probe. A significant feature of this formulation of the dynamics is that no assumptions need be made concerning the relative order of magnitudes of the effective correlation times and the Larmor frequencies (e.g., such as the extreme narrowing limit). The model not only can be used to describe the dynamics of probes such as cholesterol in membranes but also is applicable to certain anisotropic internal motions in proteins. In addition, by allowing the angle beta to fluctuate, a simple model for segmental motion in lipid molecules can be obtained. As an application, 13 C nuclear magnetic relaxation experiments on cholesterol in sonicated egg yolk phosphatidylcholine vesicles are interpreted within the framework of the model. The remarkable observation that the protonated C6 carbon has a line width which is much narrower than those of other methine carbons and is in fact comparable to the line width of the nonprotonated C5 carbon is shown to be the consequence of (1) the anisotropic nature of the cholesterol motion in the bilayer and (2) the fact that angle between the 13 C6-H vector and the long axis of cholesterol is very close to the "magic" value of 54.7 degrees.

Magnetic resonance studies of apolipoprotein C-I nitroxide labeled or [13C]methyl enriched at methionine-38.

One of the three proposed lipid-binding regions of the human apolipoprotein C-I (apo-C-I) is an amphipathic helix which extends from residue 33 to residue 53 and includes a single methionine at sequence position 38. The involvement of the sequence around methionine-38 in phospholipid binding has been evaluated with paramagnetic and nuclear reported groups attached to the thiomethyl moiety. This moiety has been spin-labeled with N-(2,2,6,6-tetramethylpiperidinyl-1-oxy)bromoacetamide or 13C enriched with 13CH3I. As determined from its EPR spectrum, the nitroxide at Met-38 of apoC-I had a rotational correlation time (tau C) of 0.22 ns. When dimyristoylphosphatidylcholine (DMPC) was bound to the spin-labeled apoprotein, tau c increased to 0.35 ns, indicating decreased motion for the methionyl side chain. The line width (nu 1/2) and spin--lattice relaxation time (T1) for the thiomethyl resonance of 13C-enriched apoC-I in 10 mM phosphate buffer was 6.0 Hz and 320 ms, respectively. When the protein solution was made 1.6 M in Gdn-HCl, these values changed to 2.6 Hz and 970 ms, respectively. Upon addition of DMPC multilamellar liposomes to [13C]apoC-I in 1.6 M Gdn-HCl, the line width increased to 4.7 Hz and the T1 decreased to 380 ms. These results strongly suggest that methionine-38 of apoC-I resides in a region of the apoprotein which undergoes significant secondary and/or tertiary structural change upon disaggregation/unfolding in Gdn-HCl and upon interaction with phospholipid.

Lipoprotein-X: carbon-13 nuclear magnetic resonance studies on native, reconstituted, and model systems.

Lipoprotein-X (LP-X), a lipoprotein isolated from human cholestatic plasma by ethanol--acetate precipitation and zonal ultracentrifugation, has been studied by 13C NMR at 67.9 MHz. Spectra of LP-X and its three subfractions are markedly different from those of normal human high-density lipoprotein3 (HDL3) or low-density lipoprotein (LDL). Spectra of LP-X are characterized by the presence of unusually broad resonance lines, especially those attributable to C6 of unesterified cholesterol (160--260 Hz) and to C beta of phospholipid glyceride (240--290 Hz). In contrast, the CH2O, CH2N, and N(CH3)3 choline resonances have line widths comparable to those of normal LDL and HDL3. For the subfraction LP-X1, spin--lattice relaxation times (T1) of the fatty acyl olefin resonances at 129.8 and 128.0 ppm and of the unesterified cholesterol C6 at 120.1 ppm were measured to be 675, 766, and 162 ms, respectively. These times are comparable to those measured for the corresponding resonances in single bilayer vesicles whose lipid composition approximates that of LP-X. The three LP-X subfractions isolated by zonal ultracentrifugation gave spectra which are identical, within experimental error, as judged qualitatively from their appearance and quantitatively from the line widths of selected resonances. In addition, 13C NMR spectra of sonicated total LP-X lipids are similar to spectra of the intact native lipoprotein. This study suggests (a) that motions of lipids in LP-X as probed by 13C NMR are similar to the motions of lipids found in model vesicular systems, (b) that the motions of the cholesterol rings and phospholipid fatty acyl chains are significantly more restricted in LP-X than in HDL3 and LDL, and (c) that the motions of the phosphoryl moieties in all three systems are similar.

Lipoprotein-X: proton and phosphorus-31 nuclear magnetic resonance studies on native, reconstituted, and model systems.

LP-X, a lipoprotein present in the low-density range (d 1.006--1.063 g/mL) of cholestatic human plasma, has been studied with its normal counterpart (LDL) by 1H and 31P nuclear magnetic resonance. The 220-MHz 1H spectrum of LP-X contains four major lines: the choline CH2N and N+(CH3)3 resonances and the cholesteryl--acyl CH2 and CH3 envelopes. The widths of these four lines at 37 degrees C are approximately 24, 10, 124, and 48 Hz, respectively. The latter two line widths are much greater than the corresponding ones of LDL (28 and 20 Hz), suggesting the much more restricted motion of acyl chains and/or cholesteryl rings in LP-X. This difference persists over the temperature range 15--52.5 degrees C. The microscopic fluidity of LP-X and LDL was compared by titration with 2,2,6,6-tetramethylpiperidinyl-1-oxy (Tempo), a paramagnetic amphiphile which distributes between the bulk aqueous phase and the fluid lipid phase of lipoproteins. Tempo is much less effective in broadening the 1H resonances of LP-X than of LDL, indicating the lower permeability/fluidity of the former. The 40.5-MHz 31P spectrum of LP-X consists of a single resonance whose line width is approximately 20 Hz and whose spin--lattice relaxation time is 2.23 +/- 0.15 s. Titration of LP-X with Pr3+ ions splits this resonance into two lines, one remaining at the chemical shift of the original resonance and the other paramagnetically shifting downfield. The ratio of integrated areas for these two lines was 1:1.72. Titration of phosphatidylcholine--cholesterol vesicles alone, vesicles containing apolipoprotein-C and albumin, or vesicles containing apolipoprotein-X gave results similar to those obtained with native LP-X, suggesting the presence of a single bilayer structure in all of these systems.

Transfer of phosphatidylcholine facilitated by a component of human plasma.

A constituent of lipoprotein-free (p greater than 1.21) human plasma from normolipemic donors facilitates the transfer of diacyl phosphatidylcholine from unilamellar egg yolk phosphatidylcholine liposomes to liver mitochondria. The active component is heat labile, has a hydrated density greater than 1.25 and an apparent molecular weight of more than 100 000. The presence of this protein in plasma may facilitate movement of diacylphospholipids between the surfaces of lipid-containing particles such as lipoproteins and erythrocytes. Knowledge of the properties and behavior of this protein are important in designing methods of drug therapy based on encapsulation in biodegradable lipid vesicles.