Structure-function relationschip of flavoproteins. With special reference to p-hydroxybenzoate hydroxylase from Pseudomonas fluorescens.

<TT>In this thesis different studies probing the structurefunction relationship of some flavoproteins are dealt with. The attention has been focused on two central themes:</TT><TT>The first part of the thesis deals with studies concerning the application of affinity chromatography in order to allow the large scale preparation of apo flavoproteins.</TT><TT>In the second part of the thesis, different studies are presented concerning the biophysical properties of p-hydroxybenzoate hydroxylase from <u>Pseudomonas</u><u>fluorescens</u> .</TT><TT>Conventional methods for the preparation of apo flavoproteins and general properties of PAD-dependent external monooxygenases are reviewed in <strong>Chapter 1</strong> . Special attention has been paid to the different studies performed with p-hydroxybenzoate hydroxylase from <u>P.fluorescens</u> .</TT><TT>Conventional methods to prepare the apoenzyme of p-hydroxybenzoate hydroxylase from <u>P.fluorescens</u> yield relatively low amounts of apoenzyme showing both variable residual and reconstitutable activity. In <strong>Chapter 2</strong> a new method is described to overcome this problem.</TT><TT>Large amounts of stable apoprotein, showing almost no residual activity, have been obtained by use of DTNB-Sepharose covalent affinity chromatography.</TT><TT>The enzyme can be reconstituted on the column or the dimeric apoprotein can be isolated in the free state. The degree of reconstitution of almost completely recovered enzyme is better than 95% of the original activity.</TT><TT>The affinity of p-hydroxybenzoate for the apoprotein is comparable to native holoenzyme. The substrate protects the apoprotein from inactivation.</TT><TT>The apoenzyme also forms a complex with NADPH. The dissociation constant of this complex is even lower than that of the holoenzyme and is strongly dependent on pH and ionic strength of the solution.</TT><TT>Kinetic experiments show that the enzyme is reconstituted in a fast process, FAD being tightly bound by the apoprotein.</TT><TT></TT><TT>In <strong>Chapter 3</strong> a new and more general applicable method for the large scale preparation of apo flavoproteins is described. Two classes of flavoproteins have been selected to demonstrate the usefulness of the applied hydrophobic interaction chromatography method. In contrast to conventional methods, homogeneous preparations of apoproteins in high yields are obtained.</TT><TT>The holoenzyme of lipoamide dehydrogenase from <u>Azotobacter</u><u>vinelandii</u> can be reconstituted while the apoprotein is still bound to the column or the apoenzyme can be isolated in the free state. The biophysical properties of completely recovered reconstituted lipoamide dehydrogenase compare favorable with the properties of native holoenzyme.</TT><TT>The holoenzyme of butyryl-CoA dehydrogenase from <u>Megasphaera</u> elsdenii cannot be reconstituted when the apoenzyme is bound to the column. However, this is the first report where stable apoprotein can be isolated in the free state. The yield of apoprotein Is more than 50% of starting material. The coenzyme A ligand present in native holoenzyme is removed during apoprotein preparation.</TT><TT>At pH 7.0 apo butyryl-CoA dehydrogenase Is in equilibrium between dimeric and tetrameric forms and reassociates to a nativelike tetrameric structure in the presence of FAD.</TT><TT>Fluorescence-polarization experiments show that the pH- dependent stability of reconstituted enzyme is strongly influenced by the presence of CoA ligands. Unliganded reconstituted enzyme is easily regreened in the presence of a mixture of coenzyme A and sodium sulfide.</TT><TT>In <strong>Chapter 4</strong> the large scale purification of p-hydroxybenzoate hydroxylase from <u>P.fluorescens</u> is described. The highly purified enzyme can be separated into at least five fractions by anionexchange chromatography. All enzyme molecules exhibit the same specific activity and exist mainly in the dimeric form in solution. The observed microheterogeneity of the enzyme can be explained by the (partial) oxidation of Cys-116 in the sequence of the enzyme.</TT><TT>The separation of the different enzymic forms has allowed the development of a kinetic FPLC method to describe the dissociation behaviour of the dimeric enzyme.</TT><TT>By chemical modification studies using maleimide derivatives, DTNB and H ₂ O ₂ , it is shown that sulfenic, sulfinic and sulfonic acid derivatives of Cys-116 are the main products of oxidation.</TT><TT>In <strong>Chapter 5</strong> the chemical modification of cysteine residues in p-hydroxybenzoate hydroxylase from <u>P.fluorescens</u> by several reagents is described. Differential labeling and sequencing radioactive labeled tryptic peptides have allowed the assignment of different cysteine residues involved in enzyme modification.</TT><TT>Cys-116 Is found to react rapidly and specifically with N-ethylmaleimide without inactivation of the enzyme.</TT><TT>The enzyme is easily inactivated by mercurial reagents. Enzyme activity can be fully restored upon addition of dithiothreitol. p-Hydroxybenzoate and also the mercurial compounds themselves inhibit the inactivation reaction.</TT><TT>A spinlabeled derivative of p-chloromercuribenzoate reacts fairly specifically with Cys-152 in N-ethylmaleimide prelabeled enzyme. Modification of Cys-152 decreases drastically the affinity of the enzyme for the substrate p-hydroxybenzoate. The modified enzyme exhibits a somewhat higher affinity for NADPH than the native enzyme.</TT><TT>Modification by p-chloromercuribenzoate leads to absorption difference spectra showing pH-dependent maxima at 290 and 360 nm. The observed pKa value of about 7.6 is tentatively ascribed to at least one of the three tyrosine residues located in the substrate binding site.</TT><TT>From the three-dimensional structure of the enzyme-p-hydroxybenzoate complex it can be deduced that Cys-152 is far away from the active site. The modification results strongly indicate that the substrate binding site and Cys-152 are interdependent.</TT><TT></TT><TT>Both group-specific chemical modification studies and crystallization experiments have not (yet) led to the elucidation of the NADPH binding site of p-hydroxybenzoate hydroxylase from P.fluorescens. In <strong>Chapter 6</strong> the NADPH binding site has been probed using the affinity label p-(fluorosulfonylbenzoyl) adenosine.</TT><TT>The enzyme is slowly inactivated by the reagent in the presence of 20% dimethylsulfoxide. The inactivation, strongly inhibited by NADPH and 2',5' ADP, can be related to the modification of one amino acid residue.</TT><TT>Steady state kinetics and 2',5' ADP Sepharose affinity chromatography of modified enzyme suggest that the essential residue is not directly involved in NADPH binding.</TT><TT>From sequencing radioactive labeled peptides It is shown that Tyr-38 Is the main residue protected from modification in the presence of NADPH.</TT><TT>The refined crystal structure of the enzyme-p-hydroxybenzoate complex at 0.19 nm resolution shows that Tyr-38 is far away from the active site. From model-building studies using computer graphics a potential mode of binding of both NADPH and 5'-(p-sulfonylbenzoyl)adenosine is presented.</TT><TT></TT><TT>Chemical modification studies ( <strong>Chapter 1.8</strong> and <strong>5</strong> ) have Indicated the presence of an ionized tyrosine residue in the vicinity of the flavin prosthetic group of p-hydroxybenzoate hydroxylase from <u>P.fluorescens</u> .</TT><TT>In <strong>Chapter 7</strong> therefore, the pH-dependent spectral properties of free oxidized enzyme in the absence or presence of substrate (analogues) have been studied by various spectroscopic techniques.</TT><TT>The observed pH-dependent transitions are explained by (de)protonation of a tyrosine residue involved in binding of the hydroxyl group of the substrate. From the crystal structure it is deduced that Tyr-201 is the most likely candidate showing this low pKa value.</TT><TT>The exact mechanism of the FAD-dependent aromatic hydroxylation reaction is still unclear. The possible role of ionization of Tyr-201 during catalysis is discussed.</TT><TT>