Dead-end Complex Formation Research Articles

Ultrasound-assisted lipase catalyzed synthesis of propyl caprate: Process optimization, kinetic, and thermodynamic evaluation

The present work evaluates the influence of ultrasound on lipase-catalyzed esterification for the synthesis of propyl caprate in a solvent-free system. The optimum conditions were found as molar ratio of capric acid: 1-propyl alcohol 1:3, catalyst loading 1% (w/w), molecular sieves loading 5% (w/w), temperature 60 °C, agitation speed 100 rpm, sonication frequency 22 kHz, duty cycle 80% (10 min) and ultrasound power 100 W yielding 86.86% conversion of capric acid. Compared to conventional agitation, ultrasound technology significantly improved process productivity to around 3% more acid conversion for each batch with 50% less requirement of catalyst amount and 5 fold less reaction time with enhanced enzyme recycling up to 9 cycles and stability. The activation energy requirement was lower (6.22 kcal/mol) with ultrasound than the conventional approach (9.028 kcal/mol). Furthermore, the esterification process was observed to be endothermic and spontaneous, with enthalpy (ΔH), entropy (ΔS), and the Gibbs free energy change (ΔG) values of 83.66 kJ/mol, 250.23 J/mol/K, and -2.74 kJ/mol, respectively. An ordered bi−bi mechanism involving the dead-end complex formation with lipase and inhibition of the reaction by capric acid was presented. Ultrasound application was found to be an influential tool to enhance the performance of biocatalyst in this esterification reaction.

Structural insights into a flavin-dependent dehalogenase HadA explain catalysis and substrate inhibition via quadruple π-stacking

HadA is a flavin-dependent monooxygenase catalyzing hydroxylation plus dehalogenation/denitration, which is useful for biodetoxification and biodetection. In this study, the X-ray structure of wild-type HadA (HadAWT) co-complexed with reduced FAD (FADH–) and 4-nitrophenol (4NP) (HadAWT−FADH–−4NP) was solved at 2.3-Å resolution, providing the first full package (with flavin and substrate bound) structure of a monooxygenase of this type. Residues Arg101, Gln158, Arg161, Thr193, Asp254, Arg233, and Arg439 constitute a flavin-binding pocket, whereas the 4NP-binding pocket contains the aromatic side chain of Phe206, which provides π-π stacking and also is a part of the hydrophobic pocket formed by Phe155, Phe286, Thr449, and Leu457. Based on site-directed mutagenesis and stopped-flow experiments, Thr193, Asp254, and His290 are important for C4a-hydroperoxyflavin formation with His290, also serving as a catalytic base for hydroxylation. We also identified a novel structural motif of quadruple π-stacking (π-π-π-π) provided by two 4NP and two Phe441 from two subunits. This motif promotes 4NP binding in a nonproductive dead-end complex, which prevents C4a-hydroperoxy-FAD formation when HadA is premixed with aromatic substrates. We also solved the structure of the HadAPhe441Val−FADH–−4NP complex at 2.3-Å resolution. Although 4NP can still bind to this variant, the quadruple π-stacking motif was disrupted. All HadAPhe441 variants lack substrate inhibition behavior, confirming that quadruple π-stacking is a main cause of dead-end complex formation. Moreover, the activities of these HadAPhe441 variants were improved by ⁓20%, suggesting that insights gained from the flavin-dependent monooxygenases illustrated here should be useful for future improvement of HadA’s biocatalytic applications.

Purification of Active Peroxidase Isoenzymes and Their Responses to Nitrogen Fertilization and Rotation of Biomass Sorghum

Peroxidases (EC 1.11.1.7) participate in lignin biosynthesis. But peroxidation is not a tool for assaying lignocellulose metabolism because the active cannot yet be separated from the inactive peroxidases. A biochemical tool for assaying plant cell wall responses to agronomic practices is needed in the lignocellulosic feedstock renewable energy industry. Peroxidase of biomass sorghum was purified to 9 - 13 charge isomers by free solution IEF (Rotofor) technique. Free solution IEF was more effective than chromatographic purification of active peroxidase isoenzymes. Native PAGE separated each charge isomer to three anionic and three cationic isoenzymes. Hydrogen peroxide and o-dianisidine assays showed that only 20% - 30% of the isoenzymes displayed normal Michaelis-Menten kinetics. Sorghum planted without nitrogen fertilization induced the hydrogen peroxide noncompetitive inhibition of peroxidase, but 280 kg·ha–1 nitrogen fertilization and 100% sorghum mineral residue return to the soil tripled the concentration of active peroxidase and relieved the inhibition with concomitant increases of 350 kg lignin and 3532 kg·cellulose·ha–1. Nitrogen fertilization without crop rotation induced hydrogen peroxide inhibition of the peroxidase, but nitrogen fertilization and 25% sorghum rotation changed the PI of the active peroxidase from neutral to mildly acidic and relieved the inhibition with concomitant enormous increases of 690 kg lignin and 7151 kg·cellulose·ha–1. Hydrogen peroxide inhibition kinetics is consistent with the known peroxidase-substrate intermediate dead-end complex formation. Lignocellulosic yield was greatest under the agronomic management that combined 280 kg·ha–1 nitrogen fertilizer with 25% sorghum residue, which resulted in a shift of pI value of the active peroxidase due to a reduction in the Km value of the peroxidase. Therefore, up to 75% of sorghum biomass rather than only 50% can be harvested for conversion to bioenergy products.

The Acyclic 2,4-Diaminopyrimidine Nucleoside Phosphonate Acts as a Purine Mimetic in HIV-1 Reverse Transcriptase DNA Polymerization

The acyclic pyrimidine nucleoside phosphonate (ANP) phosphonylmethoxyethoxydiaminopyrimidine (PMEO-DAPym) differs from other ANPs in that the aliphatic alkyloxy linker is bound to the C-6 of the 2,4-diaminopyrimidine base through an ether bond, instead of the traditional alkyl linkage to the N-1 or N-9 of the pyrimidine or purine base. In this study, we have analyzed the molecular interactions between PMEO-DAPym-diphosphate (PMEO-DAPym-pp) and the active sites of wild-type (WT) and drug-resistant HIV-1 reverse transcriptase (RT). Pre-steady-state kinetic analyses revealed that PMEO-DAPym-pp is a good substrate for WT HIV-1 RT: its catalytic efficiency of incorporation (k(pol)/K(d)) is only 2- to 3-fold less than that of the corresponding prototype purine nucleotide analogs PMEA-pp or (R)PMPA-pp. HIV-1 RT recognizes PMEO-DAPym-pp as a purine base instead of a pyrimidine base and incorporates it opposite to thymine (in DNA) or uracil (in RNA). Molecular modeling demonstrates that PMEO-DAPym-pp fits into the active site of HIV-1 RT without significant perturbation of key amino acid residues and mimics an open incomplete purine ring that allows the canonical Watson-Crick base pairing to be maintained. PMEO-DAPym-pp is incorporated more efficiently than (R)PMPA-pp by mutant K65R HIV-1 RT and is not as efficiently excised as (R)PMPA by HIV-1 RT containing thymidine analog mutations. Overall, the data revealed that PMEO- DAPym represents the prototype compound of a novel class of pyrimidine acyclic nucleoside phosphonates that are recognized as a purine nucleotide and should form the rational basis for the design and development of novel purine nucleo(s)(t)ide mimetics as potential antiviral or antimetabolic agents.

The GAP Domain and the SNARE, Coatomer and Cargo Interaction Region of the ArfGAP2/3 Glo3 are Sufficient for Glo3 Function

The ArfGAP Glo3 is required for coat protein I vesicle generation in the Golgi-endoplasmic reticulum (ER) shuttle. The best-understood role of Glo3 is the stimulation of the GTPase activity of Arf1. In this study, we characterized functional domains of the ArfGAP Glo3 and identified an interaction interface for coatomer, SNAREs and cargo in the central region of Glo3 (BoCCS region). The GAP domain together with the BoCCS region is necessary and sufficient for all vital Glo3 functions. Expression of a truncated Glo3 lacking the GAP domain results in a dominant negative growth phenotype in glo3Delta cells at 37 degrees C. This phenotype was alleviated by mutating either the BoCCS region or the Glo3 regulatory motif (GRM), or by overexpression of ER-Golgi SNAREs or the ArfGAP Gcs1. The GRM is not essential for Glo3 function; it may act as an intrinsic sensor coupling GAP activity to SNARE binding to avoid dead-end complex formation at the Golgi membrane. Our data suggest that membrane-interaction modules and cargo-sensing regions have evolved independently in ArfGAP1s versus ArfGAP2/3s.

Mechanisms Associated with HIV-1 Resistance to Acyclovir by the V75I Mutation in Reverse Transcriptase

It has recently been demonstrated that the anti-herpetic drug acyclovir (ACV) also displays antiviral activity against the human immunodeficiency virus type 1 (HIV-1). The triphosphate form of ACV is accepted by HIV-1 reverse transcriptase (RT), and subsequent incorporation leads to classical chain termination. Like all approved nucleoside analogue RT inhibitors (NRTIs), the selective pressure of ACV is associated with the emergence of resistance. The V75I mutation in HIV-1 RT appears to be dominant in this regard. By itself, this mutation is usually not associated with resistance to currently approved NRTIs. Here we studied the underlying biochemical mechanism. We demonstrate that V75I is also selected under the selective pressure of a monophosphorylated prodrug that was designed to bypass the bottleneck in drug activation to the triphosphate form (ACV-TP). Pre-steady-state kinetics reveal that V75I discriminates against the inhibitor at the level of catalysis, whereas binding of the inhibitor remains largely unaffected. The incorporated ACV-monophosphate (ACV-MP) is vulnerable to excision in the presence of the pyrophosphate donor ATP. V75I compromises binding of the next nucleotide that can otherwise provide a certain degree of protection from excision. Collectively, the results of this study suggest that ACV is sensitive to two different resistance pathways, which warrants further investigation regarding the detailed resistance profile of ACV. Such studies will be crucial in assessing the potential clinical utility of ACV and its derivatives in combination with established NRTIs.

Effects of the Translocation Status of Human Immunodeficiency Virus Type 1 Reverse Transcriptase on the Efficiency of Excision of Tenofovir

The ATP-dependent phosphorolytic excision of nucleoside analogue reverse transcriptase inhibitors can diminish their inhibitory effects on human immunodeficiency virus replication. Previous studies have shown that excision can occur only when the reverse transcriptase complex exists in its pretranslocational state. Binding of the next complementary nucleotide causes the formation of a stable dead-end complex in the posttranslocational state, which blocks the excision reaction. To provide mechanistic insight into the excision of the acyclic phosphonate nucleotide analog tenofovir, we compared the efficiencies of the reaction in response to changes in the translocation status of the enzyme. We found that rates of excision of tenofovir with wild-type reverse transcriptase can be as high as those seen with 3'-azido-3'-deoxythymidine monophosphate (AZT-MP). Thymidine-associated mutations, which confer >100-fold and 3-fold decreased susceptibility to AZT and tenofovir, respectively, caused substantial increases in the efficiency of excision of both inhibitors. However, in contrast to the case for AZT-MP, the removal of tenofovir was highly sensitive to dead-end complex formation. Site-specific footprinting experiments revealed that complexes with AZT-terminated primers exist predominantly pretranslocation. In contrast, complexes with tenofovir-terminated primers are seen in both configurations. Low concentrations of the next nucleotide are sufficient to trap the complex posttranslocation despite the flexible, acyclic character of the compound. Thus, the relatively high rate of excision of tenofovir is partially neutralized by the facile switch to the posttranslocational state and by dead-end complex formation, which provides a degree of protection from excision in the cellular environment.

Pyrophosphorolytic Excision of Nonobligate Chain Terminators by Hepatitis C Virus NS5B Polymerase

Nonobligate chain terminators, such as 2'-C-methylated nucleotides, block RNA synthesis by the RNA-dependent RNA polymerase (RdRp) of hepatitis C virus (HCV). Previous studies with related viral polymerases have shown that classical chain terminators lacking the 3'-hydroxyl group can be excised in the presence of pyrophosphate (PP(i)), which is detrimental to the inhibitory activity of these compounds. Here we demonstrate that the HCV RdRp enzyme is capable of removing both obligate and clinically relevant nonobligate chain terminators. Pyrimidines are more efficiently excised than are purines. The presence of the next complementary templated nucleotide literally blocks the excision of obligate chain terminators through the formation of a dead-end complex (DEC). However, 2'-C-methylated CMP is still cleaved efficiently under these conditions. These findings show that a 2'-methylated primer terminus impedes nucleotide binding. The S282T mutation, associated with resistance to 2'-C-methylated nucleotides, does not affect the excision patterns. Thus, the decreased susceptibility to 2'-C-methylated nucleotides appears to be based solely on improved discrimination between the inhibitor and its natural counterpart. In conclusion, our data suggest that the phosphorolytic excision of nonobligate, pyrimidine-based chain terminators can diminish their potency. The templated nucleotide does not appear to provide protection from excision through DEC formation.

Site-directed Mutagenesis in the Fingers Subdomain of HIV-1 Reverse Transcriptase Reveals a Specific Role for the β3–β4 Hairpin Loop in dNTP Selection

HIV-1 reverse transcriptase shares the key features of high fidelity polymerases, such as a closed architecture of the active site, but displays a level of fidelity that is intermediate to that of high fidelity, replicative polymerases and low fidelity translesion synthesis (TLS) polymerases. The β3–β4 loop of the HIV-1 RT fingers subdomain makes transient contacts with the dNTP and template base. To investigate the role of active site architecture in HIV-1 RT fidelity, we truncated the β3–β4 loop, eliminating contact between Lys65 and the γ-phosphate of dNTP. The mutant, in a manner reminiscent of TLS polymerases, was only able to incorporate a nucleotide that was capable of base-pairing with the template nucleotide, but not a nucleotide shape-analog incapable of Watson–Crick hydrogen bonding. Unexpectedly, however, the deletion mutant differed from the TLS polymerases in that it displayed an increased fidelity. The increased fidelity was associated with reduced dNTP binding affinity as measured using the dead end complex formation. In an effort to delineate the specific amino acid residue in the deleted segment responsible for this phenotype, we examined the K65 residue. Two substitution mutants, K65R and K65A were studied. The K65A mutant behaved similarly to the deletion mutant displaying dependence on Watson–Crick hydrogen bonding, increased fidelity and reduced dNTP-binding, while the K65R was more akin to wild-type enzyme. These results underscore the key role of the K65 residue in the phenotype observed in the deletion mutant. Based on the well-known electrostatic interaction between K65 and the γ-phosphate moiety of incoming dNTP substrate in the ternary complex structure of HIV-1 RT, we conclude that non-discriminatory interactions between β3–β4 loop and the dNTP in wild-type HIV-1 RT help lower dNTP selectivity. Our results show that the fidelity of dNTP insertion is influenced by protein interactions with the triphosphate moiety.

The enzymological basis for resistance of herpesvirus DNA polymerase mutants to acyclovir: relationship to the structure of alpha-like DNA polymerases.

Acyclovir (ACV), like many antiviral drugs, is a nucleoside analog. In vitro, ACV triphosphate inhibits herpesvirus DNA polymerase by means of binding, incorporation into primer/template, and dead-end complex formation in the presence of the next deoxynucleoside triphosphate. However, it is not known whether this mechanism operates in vivo. To address this and other questions, we analyzed eight mutant polymerases encoded by drug-resistant viruses, each altered in a region conserved among alpha-like DNA polymerases. We measured Km and kcat values for dGTP and ACV triphosphate incorporation and Ki values of ACV triphosphate for dGTP incorporation for each mutant. Certain mutants showed increased Km values for ACV triphosphate incorporation, suggesting a defect in inhibitor binding. Other mutants showed reduced kcat values for ACV triphosphate incorporation, suggesting a defect in incorporation of inhibitor into DNA, while the rest of the mutants exhibited both altered km and kcat values. In most cases, the fold increase in Ki of ACV triphosphate for dGTP incorporation relative to wild-type polymerase was similar to fold resistance conferred by the mutation in vivo; however, one mutation conferred a much greater increase in resistance than in Ki. The effects of mutations on enzyme kinetics could be explained by using a model of an alpha-like DNA polymerase active site bound to primer/template and inhibitor. The results have implications for mechanisms of action and resistance of antiviral nucleoside analogs in vivo, in particular for the importance of incorporation into DNA and for the functional roles of conserved regions of polymerases.

Nucleotide-induced stable complex formation by HIV-1 reverse transcriptase.

Nondenaturing gel electrophoresis was used to study the nucleotide substrate-induced conformational change in reverse transcriptase (RT) of human immunodeficiency virus type 1 (HIV-1). Dead-end complex was formed between HIV-1 RT, dideoxynucleotide chain-terminated primer, and DNA template in the presence of deoxynucleotide triphosphate (dNTP) complementary to the next position on the template. Complexes which form in the absence of the next complementary dNTP were disrupted by adding excess poly(rA)/oligo(dT) or heparin just prior to electrophoresis. Dead-end complex formation by noncomplementary dNTP's or ribonucleotides was at least 2000-fold less efficient than with the complementary nucleotide. When dA was the next nucleotide on the template, analogues of dTTP supported dead-end complex formation with increased apparent Kd (dTTP < dideoxy-TTP approximately alpha-thio-dTTP < dUTP < 3'-azidothymidine triphosphate). A similar relationship was observed for dGTP analogues across from dC on the template (dGTP < dideoxy-GTP < alpha-thio-dGTP << dITP < dideoxy-ITP). The optimal length of the primer/template duplex region for dead-end complex formation was between 20 and 32 base pairs. Primer-template with a mismatched primer terminus did not support dead-end complex formation, and primer terminated with 3'-azidothymidine formed dead-end complex with 25-fold elevated apparent Kd. By contrast, dead-end complex formation on primer terminated with dideoxy-IMP base paired with dC on the template was more efficient than on primer terminated with dideoxy-GMP. Implications for the mechanisms of discrimination between nucleotide analogues by HIV-1 RT are discussed.

Preferred order random kinetic mechanism for homoserine dehydrogenase of Escherichia coli (Thr-sensitive) aspartokinase/homoserine dehydrogenase-I: equilibrium isotope exchange kinetics.

Isotope exchange kinetics at chemical equilibrium have been used to investigate the kinetic mechanism of homoserine dehydrogenase (EC 1.1.1.3) of the (Thr-sensitive) aspartokinase/homoserine dehydrogenase-I multifunctional enzyme from E. coli. For the reaction (L-ASA + NADPH + H+ = L-Hse + NADP+), at pH 9.0, 37 degrees C, Keq = 100 (+/- 20). Under these conditions, the rate for exchange of [14C]-L-homoserine (Hse) in equilibrium L-aspartate-beta-semialdehyde (ASA) is nearly twice that for the [3H]-NADP+ in equilibrium NADPH exchange. This indicates that covalent interconversion between reactants and products bound in the active site cannot be rate-limiting. Upon variation of the concentrations of all four substrates in constant ratio at equilibrium (to minimize dead-end complex formation), the Hse in equilibrium ASA exchange increased smoothly toward a maximum. In contrast, the NADP+ in equilibrium NADPH exchange rate increased to a maximum value at partial saturation, then decreased to approximately half the maximum rate. These data are consistent with a preferred-order random kinetic mechanism in which the dominant pathway involves association of NADPH prior to L-ASA and dissociation of L-Hse prior to NADP+.

Proline biosynthesis in Escherichia coli kinetic and mechanistic properties of glutamate semialdehyde dehydrogenase

The kinetics of the NADP +- and phosphate-dependent oxidation of glutamic acid 5-semialdehyde are consistent with a rapid-equilibrium random order mechanism. The K m for dl-pyrroline-5-carboxylic acid is 2.5 mM, for NADP + is 0.05 mM and for phosphate is 0.35 mM. The V max is approx. 8.0 units per mg protein. The reaction is highly specific for the dl-pyrroline-5-carboxylic acid and NADP +, but a number of divalent anions can substitute for phosphate. NADPH is competitive with respect to all three substrates and an analog of γ-glutamyl phosphate, 3-(phosphonoacetylamido)- l-alanine, is competitive with respect to dl-pyrroline-5-carboxylic acid and non-competitive with respect to NADP + and phosphate, suggesting dead-end complex formation.

The influence of 2-oxoglutarate on the activity of prolyl 4-hydroxylase

1. Treatment of prolyl 4-hydroxylase (prolyl-glycyl-peptide, 2-oxoglutarate:oxygen oxidoreductase (4-hydroxylating), EC 1.14.11.2) with 2-oxoglutarate in the absence of added Fe 2+ for 10s causes partial inactivation of the enzyme which is not reversed by subsequent addition of Fe 2+. It appears that 2-oxoglutarate prevents loss of enzyme-bound iron and prevents access of added iron to its binding site. 2. For optimal enzyme activity the enzyme should be preincubated for 15s with Fe 2+ (5μM). 3. Under turnover conditions prolyl 4-hydroxylase does not release iron. 4. The inactivation brought about by pre-incubation with 2-oxoglutarate and 0 2 in the absence of ascorbate is partly reversed by removal of 2-oxoglutarate. 5. It is proposed that dead-end complex formation with 2-oxoglutarate is responsible for the inactivation of the enzyme by 2-oxoglutarate in the absence of either ascorbate or Fe 2+. 6. Optimal enzyme activity is obtained if the reactants are added to the reaction medium in the following order, enzyme, iron and ascorbate in any order, then after at least 15s 2-oxoglutarate and finally (Pro-Pro-Gly) 5 to start the reaction.

D-Mannitol dehydrogenase from Absidia glauca. Steady-state kinetic properties and the inhibitory role of mannitol 1-phosphate

Steady-state kinetic studies including initial velocity for mannitol oxidation and fructose reduction and product inhibition for mannitol oxidation using fructose and reduced nicotinamide adenine dinucleotide (NADH) are in accord with a reaction mechanism best described as ordered Bi-Bi with NAD+ and NADH designated as the first substrate, last product, respectively at pH 8.8. All replots of slopes and intercepts from product inhibition studies were linear. Dead-end inhibition studies using mannitol 1-phosphate gave slope-parabolic, intercept-linear noncompetitive inhibition for both NAD+ and mannitol as substrates. The dead-end inhibitor is capable of binding multiply to the E, EA, and EQ forms of the enzyme to an extent that is controlled by the concentration of substrates. The EQ complex is inferred to undergo a conformational change, E'Q equilibrium EQ, since (V1/E1) greater than (KiqV2)/(KqE1), and no evidence for dead-end complex formation with NADH can be adduced. This is interpreted to mean that the release of fructose from the central complex is faster than the isomerization of the E-NADH complex. When mannitol is saturating, the noncompetitive inhibition against NAD+, as the variable substrate, becomes parabolic uncompetitive. A replot of the slopes of the parabola against mannitol 1-phosphate remains concave upward. This situation could arise if the conformational change we infer in the EQ complex opens up additional sites on the protein which can interact with the dead-end inhibitor.

Deviation from Michaelis-Menten kinetics for fumarase.

A study of the steady-state kinetics of fumarase over an extended concentration range, using novel methods of analysis, reveals an initial-rate equation of at least fourth degree for malate as substrate at pH 7.0, with no kinetically significant dead-end complex formation even up to concentrations of 100 mM. In the absence of demonstrable enzyme-aggregation phenomena, this is interpreted as indicating co-operative effects overlooked previously, although a mixture of isoenzymes, each individually of high degree and giving a complex curve, may be a contributing factor.

Oxidation of ferrocytochrome c by mitochondrial cytochrome c oxidase.

Attempts to rationalize the kinetics of cytochrome c oxidation catalyzed by solubilized mitochondrial cytochrome c oxidase (ferrocytochrome c:oxygen oxidoreductase, EC 1.9.3.1) have been based on assumptions of productive complex formation (Michaelis-Menten approach). However, the range of substrate concentrations used has not, in general, been sufficient to establish a general rate equation. Data adequate to derive such a rate expression are presented, as well as a method for estimation of constants which appear in the rate law deduced and reported herein. It is shown that either of two types of mechanisms, one assuming productive complex formation, as opposed to the other postulating dead-end complex formation, accurately predict the rate equation as deduced from experiment.

Properties of halophil nicotinamide-adenine dinucleotide phosphate-specific isocitrate dehydrogenase. True Michaelis constants, reaction mechanisms and molecular weights.

True values of Michaelis constants of the NADP(+)-specific isocitrate dehydrogenase from Halobacterium salinarium were not very different from those of the apparent constants reported by Aitken et al. (1970). The true constants were affected by salt in a similar manner to that of the apparent constants obtained with NADP(+) at fixed concentrations of 1.0-0.2mm and threo-d(s)-(+)-isocitrate at fixed concentrations of 2.0-0.125mm. The response of apparent V(max.) to salt concentration was highly dependent on fixed substrate concentration in solutions of sodium chloride but much less so in solutions of potassium chloride. At several levels the results emphasize the difficulty of generalizing about the salt relations of a halophil enzyme without adequate attention to substrate concentration. The enzyme has at least two different reaction mechanisms depending on salt concentration. In its ;physiological' form (i.e. in 1.0m-potassium chloride), and also in 1.0m-sodium chloride, the reaction mechanism is ordered with NADP(+) the first substrate added and NADPH the last product released. In 0.25m-sodium chloride, however, the mechanism is different and is probably non-sequential. In 4.0m-sodium chloride with low concentrations of either fixed substrate, there was evidence of a co-operative action of the variable substrate. The evidence suggests that salt participates in the reaction mechanism in two ways: one is the reversible addition to the enzyme in a manner analogous to that of a substrate; the other is dead-end complex-formation. The relative contributions of these two types of reaction determine whether salt activates or inhibits the enzyme. In addition, the inhibition caused by high concentrations of sodium chloride is more complex than the corresponding inhibition by potassium chloride. Gel-filtration experiments indicated that at very low salt concentrations the enzyme has an apparent molecular weight of about 70800. In ;physiological' concentrations of potassium chloride the enzyme appears to be a dimer (mol.wt. 122000-135000) and, in 1.0-4.0m-sodium chloride, it behaves as a trimer or tetramer (mol.wt. 224000-251000). A preliminary method of purifying the enzyme is described.