5$\alpha$-Cholest-8(14)-en-3$\beta$-ol-15-one is a potent inhibitor of cholesterol biosynthesis which has been found to have significant hypocholesterolemic action upon oral administration to rodents and nonhuman primates. The metabolism of (2,4-$\sp3$H) 5$\alpha$-cholest-8(14)-3n-3$\beta$-ol-15-one was studied in Chinese hamster ovary (CHO-K1) cells. The incorporation of the labeled 15-ketosterol into the cells was linear with respect to sterol concentration in the medium over the range of concentrations studied and was higher than the uptake of cholesterol. The results of detailed analyses of the lipids recovered from the cells after 6 hours of incubation with the (2,4-$\sp3$H) -15-ketosterol indicated that most of the $\sp3$H was associated with the free 15-ketosterol. Considerably smaller amounts of $\sp3$H were associated with esters of the 15-ketosterol. No conversion of the 15-ketosterol to cholesterol or other C$\sb{27}$ monohydroxysterols was observed. The labeled material with the chromatographic behavior of esters of the 15-ketosterol gave, after mild saponification, the free 15-ketosterol which was characterized by cocrystallization and chromatographic studies. The metabolism of the 15-ketosterol was also studied in male baboons (Papio cynocephalus) treated with the 15-ketosterol. After oral administration of a mixture of (2,4-$\sp3$H) 5$\alpha$-cholest-8(14)-en-3$\beta$-ol-15-one and (4-$\sp{14}$C) cholesterol, blood samples were obtained at various times. Marked differences in the time courses of the levels of $\sp3$H and $\sp{14}$C in plasma were observed. $\sp3$H showed maximum levels at 4 to 8 h, while maximum values for the levels of $\sp{14}$C were observed much later. Total lipid extraction of plasma showed that essentially all of the $\sp{14}$C of plasma was recovered in the lipid extract. In contrast, while most of the $\sp3$H was recovered in the lipid extract, a significant amount of $\sp3$H remained in the aqueous phase after lipid extraction. Most of the $\sp3$H in plasma was found as metabolites of the 15-ketosterol, i.e., esters of the 15-ketosterol, cholesterol and cholesteryl esters. The plasma levels of the 15-ketosterol and of each of these metabolites showed different changes with time. The labeled cholesterol (and the cholesterol of cholesteryl esters), formed from the 15-ketosterol, was thoroughly characterized by chromatography and by purification by way of it dibromide derivative. Detailed studies were also made of the effects of the 15-ketosterol and of two other oxygenated sterols on the distribution of $\sp3$H in nonsaponifiable lipids after incubation of ($\sp3$H) acetate with the 10,000 x g supernatant fraction of rat liver homogenates. Whereas 14$\alpha$-ethyl-5$\alpha$-cholest-7-ene-3$\beta$,15$\alpha$-diol (1 $\mu$M) caused a marked accumulation of labeled lanosterol and 24,25-dihydrolanosterol (characterized by chromatographic and cocrystallization experiments), 5$\alpha$-cholest-8(14)-en-3$\beta$-ol-15-one (100 $\mu$M) and 25-hydroxycholesterol (6.25 $\mu$M and 12.5 $\mu$M) had no demonstrable effect on the distribution of $\sp3$H in the nonsaponifiable lipids.

Ligand binding to several naturally occurring and engineered myoglobins and hemoglobins was examined. Overall and germinate rate constants and quantum yields were determined at pH 7, 20$\sp\circ$C for O$\sb2$, NO, CO, and a set of n-series and $\alpha$-substituted alkyl isocyanide complexes of sperm whale myoglobin, isolated $\alpha$ and $\beta$ hemoglobin chains, soybean leghemoglobin, and monomeric hemoglobin component II from Glycera dibranchiata using a 30 ns dye laser pulse and a 0.5 ms xenon flash. Overall rates and quantum yields for several sperm whale myoglobin complexes were compared to parameters obtained for site directed mutants of this protein, in which His-E7 was replaced by one of several different amino acids. Results were analyzed using a linear three step reaction scheme, HbX$\rightleftharpoons$B$\rightleftharpoons$C$\rightleftharpoons$H $+$ X, where X is ligand free in solution, H and HX are unliganded and liganded heme protein respectively, B represents a geminate state in which the ligand is in the distal pocket but not covalently bound to the iron atom, and C is a state in which the ligand is still embedded in the protein but further away from the heme group. The results indicate that low overall quantum yields are due to rapid rates of geminate recombination relative to rates of ligand escape to solvent rather than a low intrinsic photophysical yield. For all CO complexes, Q is large and the rate limiting step for association is iron-ligand bond formation and dissociation is rate limited by thermal bond disruption. For O$\sb2$, NO, and most of the isonitrile complexes, Q is small ($\leq$0.2) indicating the association rate is approximately equal to the rate of ligand migration to state B, and dissociation is limited by escape from the protein. Studies of the site directed mutants of Mb show that the E7 residue acts as a steric barrier for isonitriles. In contrast, the polarity of the E7 side chain is more important in determining the rate and equilibrium constants for O$\sb2$ and CO binding. The dominant factor for the stabilization of the native MbO$\sb2$ complex is the hydrogen bond between bound O$\sb2$ and the N$\varepsilon$ proton of His-E7. The largest kinetic barrier to O$\sb2$ and CO binding appears to be displacement of water found in the distal pocket of native deoxymyoglobin.

Two NADH dependent butanol(ethanol) dehydrogenases from Clostridium acetobutylicum have been purified to near homogeneity. Either of these two dehydrogenases could be part of the pathway for the production of butanola and are thus important for an understanding of the metabolism involved in butanol production. Another butanol dehydrogenase was detected that was largely dependent on NADPH. Studies were also carried out on the effects of rifampicin and chloramphenicol on the metabolism of C. acetobutylicum. Butanol dehydrogenases I and II are composed of two subunits of about 42 kD in weight, with a native molecular weight of 80 kD. Both enzymes were purified with Zn$\sp{2+}$ in the buffer to improve the recovery of enzyme activity. The enzymes were more active at around pH 6.0, within the physiological pH range of 5.6-6.8 as reported for C. acetobutylicum. Both enzymes were substantially more active in the normal metabolic direction (butyraldehyde to butanol), and the kinetics of this reaction were studied in this direction. Both enzymes were determined to have ordered bi bi kinetic binding mechanisms. This mechanism was determined through substrate inhibition analysis with butyraldehyde for BDH I and through competitive and product inhibition studies for BDH II. The reactivity of the enzymes towards other substrates was also determined; the rate of the reduction of aldehydes appears to increase with increasing chain length for both butanol dehydrogenase I and II, none of the ketones studied acted as substrates. The most important difference is that BDH I was substantially more reactive towards acetaldehyde than BDH II. The effect of various metabolites, such as ATP and CoA, was also determined and the results were similar for each enzyme. The amino acid sequences and pI's were determined for both BDH I and II and slight differences were found. The effect of the addition of either rifampicin or chloramphenicol to growing cultures was studied for the products: butyrate, acetate, butanol, and acetone, and for the enzymes: phosphotransbutyrylase, butyrate kinase, coenzyme A transferase, butyraldehyde dehydrogenase, and butanol dehydrogenase (both NADH and NADPH dependent). The results demonstrated that phosphotransbutyrylase and butyrate kinase are constitutive and stable in vivo. CoA transferase and the butanol dehydrogenases were all induced/derepressed and were also stable in vivo. Butyraldehyde dehydrogenase was induced/derepressed but was highly unstable in vivo. The instability of butyraldehyde dehydrogenase coupled with the location of butyraldehyde dehydrogenase as a branch point in the butanol pathway, suggests that butyraldehyde dehydrogenase is highly important in regulating the flow of butyryl-CoA and therefore in regulating the switch from butyrate to butanol production.

Type II dihydrofolate reductases (DHFRs) encoded by the R67 and R388 plasmids are sequence and structurally different from known chromosomal DHFRs, and are responsible for conferring trimethoprim resistance to the host bacterial strain. An overproduced derivative of R388 DHFR, RBG200 DHFR, was purified to apparent homogeneity. The pH versus activity profile of RBG200 DHFR was found to be similar to that previously reported for R388 DHFR, but different from that of chromosomal DHFRs. Gel filtration and equilibrium ultracentrifugation experiments suggested that at pH 6, where the enzyme is most active, RBG200 DHFR is in equilibrium between tetramer and other protein species. Gel filtration studies suggested that the loss of enzyme activity between pH 6 and pH 5 is not due to tetramer dissociation. Gel filtration experiments with R67 DHFR produced similar results. RBG200 DHFR was found to catalyze the transfer of the pro-R hydrogen of NADPH to dihydrofolate, making it a type-A dehydrogenase, along with the chromosomal DHFRs. Addition of NADP$\sp+$ to RBG200 DHFR results in the formation of an initial binary complex which slowly interconverts to a second binary complex, with an apparent first order rate constant of 1.0 $\times$ 10$\sp{-4}$ s$\sp{-1}$, at 25$\sp\circ$C. The binding of NADP$\sp+$ to RBG200 DHFR in the second binary complex was found to be weak, K$\sb{\rm D}$ = 1.9 $\pm$ 0.4 mM. The stoichiometry for coenzyme binding was found to be approximately one coenzyme per tetramer of RBG200 DHFR. Transferred NOEs were used to estimate internuclear distances, and it was found that NADP$\sp+$ binds to RBG200 DHFR in the initial and final RBG200 DHFR$\cdot$NADP$\sp+$ binary complexes, as well as in the ternary complex with folate, with a syn conformation about the nicotinamide-ribose glycosidic bond and an anti conformation about the adenine-ribose glycosidic bond. From the stereochemistry of hydride transfer and the conformation of the enzyme-bound cofactor, a model was proposed for the orientation of coenzyme and substrate as the active site of RBG200 DHFR, which differs from that observed at the active site of chromosomal DHFRs, and may be responsible for the resistance of type II DHFRs to several antifolates.

A comprehensive model for the steady state behavior of xanthine oxidase is proposed. The model includes 36 species which account for all the possible levels of reduction and binding states for the enzyme. A set of seven intrinsic reactions are modulated by the fraction of available reaction center to provide 160 macroscopic rates between these species. Unlike previous steady-state models, our model includes features known from rapid kinetic experiments such as the formation of a bound product species and the ability of the enzyme to react three times sequentially with substrate under anaerobic conditions. The complex reoxidation reactions producing either hydrogen peroxide or superoxide radical depending on the overall state of reduction of the enzyme are also accounted for in our model. Measured room temperature mid-point potentials are used to predict the fraction of available reaction center according to the accepted rapid-equilibrium model and literature values for the intrinsic reactions are used when available. We are able to predict product turnover and superoxide turnover for both xanthine and lumazine using our model. The reductive rapid kinetic reaction of xanthine oxidase with lumazine has been proposed to follow a simple three step reaction sequence of substrate binding, bound-product formation and product release. Absorbance changes during the first turnover with lumazine can be simulated using the three step mechanism. Molar absorbances for the intermediates are obtained from the transient kinetics of substoichiometric lumazine with enzyme. The observed molar absorbance at 650 nm for two electron enzyme with violapterin bound is much larger than expected if only Mo(IV) forms the charge transfer band. To verify that Mo(V) contributes to the charge transfer band at 650 nm, optical and epr potentiometric titrations can be performed with enzyme in excess violapterin. Midpoint potentials for molybdenum with violapterin bound are determined from the epr potentiometric titration. The Mo(VI)/Mo(V) and Mo(V)/Mo(IV) half-reactions are $-$352 and $-$347 mV, respectively. Fitting the optical potentiometric titration data, the molar absorbance of the Mo(V) and Mo(IV) species with violapterin bound are 9.23 and 8.53 mM$\sp{-1}$cm$\sp{-1}$ at 650 nm, respectively.

Crystals were prepared for several studies of the periplasmic maltodextrin-binding protein (MBP) from E. coli. This protein is part of the bacterial transport system for starch metabolites. The complex of MBP with maltotetraose was refined and analysed. The maltotetraose conformation has surprising implications, as it is unusual in terms of the sugar literature but the same as that of maltoheptaose bound to glycogen phosphorylase a (Goldsmith et al., Journal of Molecular Biology 156:411-427, 1983). The complex suggests explanations for an unexpectedly low binding affinity for maltotetraose. MBP was co-crystallized with alpha-cyclodextrin and beta-cyclodextrin; data has been collected from the latter crystal. Crystals and data were also obtained for sugar-free MBP. The current model of transport by this system suggests a conformational change in MBP upon sugar binding which is recognized by the transport system. A mutation in the hinge region of MBP produced a ligand-specific effect on transport; crystals of this mutant with two different ligands were grown. Two further experiments use MBP for general studies of protein structure. First, two deletions on the surface of MBP remove secondary structural elements without affecting the function of the protein (Duplay et al., Journal of Molecular Biology 194:663-678, 1987). Both were crystallized, and data from one was collected and structural analysis begun. Finally, a mutant produced by inserting the C3 neutralizing epitope from type 1 poliovirus into one of the surface deletion sites was crystallized. The structure was examined carefully, but the insertion was found to be disordered.

Recent studies on the tissue distribution and developmental regulation of ADA activity in mice show that very high ADA levels exist in the murine alimentary tract (tongue, esophagus, forestomach, proximal small intestine) and at the fetal-maternal interface. Levels of three other enzymes involved in purine catabolism--purine nucleoside phosphorylase (PNP), guanine deaminase (GDA), and xanthine dehydrogenase (XDH)--were measured and compared with ADA activity. The highest levels of PNP, GDA and XDH were found in the proximal small intestine. Levels of these three enzymes were much lower in the tongue, esophagus, forestomach and fetal-maternal interface in marked contrast to ADA distribution. Tissue-specific differences in PNP, XDH and ADA activity correlated with RNA abundance indicating that the regulation of gene expression is at the level of mRNA production. Thus, ADA is part of a purine catabolic pathway leading to the production of uric acid that is present at the highest known level in the proximal small intestine. ADA may have additional roles in other tissues. We also studied catalytic aspects of ADA. The proposed catalytic mechanism of ADA based on the recent elucidation of its X-ray structure (Wilson et al., Science 252:1278) hypothesized that Glu$\sp{217}$ was involved in protonation of N-1 of the adenosine ring, suggesting a vital role for this residue in tetrahedral intermediate formation. To study the importance of this residue, Glu$\sp{217}$ of murine ADA was mutated to Asp, Gly, Gln and Ser. A novel purification protocol without the use of affinity chromatography was developed. Circular dichroism and zinc analysis showed no change in secondary structure or zinc content, respectively, compared to the native protein. The mutants showed only slight variation in K$\sb{\rm m}$ but dramatically reduced k$\sb{\rm cat}$ ranging from 0.03% (E217G) to 1% (E217S) of wild type activity. K$\sb{\rm i}$'s with purine riboside were similar for wild-type and mutant adenosine deaminases further underscoring the minimal role this residue plays in substrate binding. A characteristic UV difference spectrum of native ADA with purine riboside, indicative of protonation of the purine ring at N-1 of the substrate, was not observed with the mutants. These data confirm the importance of Glu$\sp{217}$ in catalysis as suggested by the crystal structure of mADA.

Proteins were isolated from acidic extracts and alkaline extracts of white kidney beans (Phaseolus vulgaris), navy beans (Phaseolus vulgaris) and baby lima beans (Phaseolus lunatus). The proteins were examined under a light microscope and a Scanning Electron Microscope. The isolated proteins had different microscopic structures; the structures depended on the conditions (pH and normality of extractant) of extraction. The various protein preparations were analysed for phytic acid, carbohydrates and citric acid. The nature of the protein-phytic acid reaction which takes place during the isolation of the proteins was investigated. The proteins that were isolated from the citric acid extracts and sodium hydroxide extracts of the beans were investigated by use of (a) electrophoretic analysis under acidic and under basic conditions in the absence and presence of dissociating agents, such as urea, 2-mercaptoethanol and sodium dodecyl sulphate, (b) ion-exchange chromatography and (c) gel-filtration chromatography.

A simple method for purifying the gene A plus A* proteins of bacteriophage (PHI)X174 is presented. The purified preparation was able to stimulate (PHI)X DNA replication in vitro and was able to nick supercoiled (PHI)X RFI to produce unit-length linear and circular strands of DNA. The kinetics of the nicking reaction were more consistent with a stoichiometric than with a catalytic mechanism. Under non-denaturing conditions the presumably globular A and A* proteins sedimented with apparent molecular weights of 110,000 and 90,000 daltons. A* was unable to nick supercoiled (PHI)X RFI DNA in the absence of A, but participated in the nicking reaction when A was present and both A and A* were found covalently bound to the 5'-end of the resulting nicked DNA strands. The A protein is truly cis-acting in vivo. Relaxation complexes like those observed with the E. coli colicinogenic factors were not found during (PHI)X semi-conservative DNA replication.

Alkaline phosphatase from Escherichia coli has been immobilized on Sepharose CL-4B using low levels of cyanogen bromide activation in order to favour the attachment of the dimer to the support through a single covalent linkage. The level of phosphatase activity in the matrix-bound monomer, obtained after dissociation of the immobilized dimer, was dependent on the extent of cyanogen bromide activation of the support, indicating that this activity is due to contaminating dimers. The quaternary structure of the subunit, reversibly immobilized through a disulfide linkage, was determined by crosslinking the matrix-bound protein, then releasing the protein from the support and analysing it by sodium dodecyl sulfate gel electrophoresis. This method demonstrated the presence of dimeric structures whose concentration was linearly correlated to the amount of phosphatase activity in the matrix-bound subunits. / Titration of matrix-bound ('125)I-labelled subunits with soluble nascent ('131)I-labelled subunits resulted in the recovery of over 60% of the original dimer phosphatase activity and a final isotope ratio of 1.06. Titration of metal-free immobilized dimer and monomer with ('65)Zn('2+) showed that the monomer bound 0.9 equivalents of Zn('2+), while the dimer bound 4.1 equivalents of Zn('2+). These results indicate that although the monomer lacks catalytic activity, it exists as a highly folded structure containing sites for Zn('2+) binding and subunit interactions. Chemical modification with ethoxyformic anhydride has demonstrated that three histidines per subunit are modified in the soluble enzyme with a concomitant loss of catalytic activity. Zn('2+) ions protect the enzyme from modification as well as from inactivation, thus implicating all three histidines in Zn('2+) binding. Zn('2+) also protects the monomer against this modification providing independent evidence for Zn('2+) binding in renatured subunits. These techniques, developed in order to characterize subunits of alkaline phosphatase, are generally applicable to other oligomeric enzymes.