No description available.

The multiple isomorphous replacement method of protein crystallography has revealed the structures of several proteins, but the method is limited by the need for many isomorphous crystals. This difficulty could be eased to a large extent by using a molecular fragment as a model for the unknown structure, when only one set of X-ray intensities need be measured. This thesis discusses ways to determine the orientation of such a partial structure, and recommends a new method to determine the location of a properly oriented partial structure in the crystal unit cell. These methods were applied to the proteins human lysozyme, hen lysozyme crystallised at a high temperature, and goat α-lactalbumin, for all of which part of the published structure of hen lysozyme was used as a model. Only the application to human lysozyme was successful, and the failure in the case of the high-temperature hen lysozyme is surprising. An application of these methods to the antibiotic thiostrepton, using an almost complete structure from another crystal form as a model was successful. Having found a partial structure which explains the X-ray data moderately well, it is necessary to complete and refine the model. Several of the theories discussed in this thesis assume Wilson statistics, which are shown to apply to human lysozyme surprisingly well. Three types of completion / refinement technique were applied, a least squares refinement of atomic occupancies, direct methods, and the use of maps with Fourier coefficients with phases derived from the partial structure. The first two methods failed, but the third had partial success. The relationship between "difference maps", Sim weighting, and the α-synthesis is discussed; an extension of Luzzati's treatment of maps with Fourier coefficients F exp(ia_c) shows that of these three Fourier syntheses, the α-synthesis has the best (unknown peak height)/(known peak height) ratio, whereas Sim weighting has the best signal/noise ratio - which, incidentally, never exceeds unity for acentric data - where "noise" is defined in a more realistic way than that adopted by Sim. A corrected and extended version of a previous semi-quantitative account of difference maps is given. The application of various Fourier syntheses to human lysozyme structure factor amplitudes proceeded as a test of methods in parallel with the isomorphous replacement study by other workers; it is recommended that both F_obs, and difference maps be calculated, using Sim weighting. Small differences in the main chain conformations of hen and human lysozymes can be determined, missing side chain atoms can often be placed, but rearrangements of the structure as determined by the isomorphous replacement study are not always seen on difference maps. The value of the residual for a randomly wrong structure when a mixture of acentric and centric data is used is shown to be within 0.007 of the value which corresponds to a linear interpolation between the all centric and all acentric limiting cases.

The three-dimensional structure of the glycosidase, hen egg white lysozyme, has been determined by X-ray crystallography; the molecule is roughly ellipsoidal, with a cleft running down one side, and this cleft, which is capable of accommodating up to six sugar residues (in sub-sites A-F) is known to be the site of catalytic action, glycolysis occurring between sites D and E. The only significant enzyme functional groups in this region are the carboxylic acids Asp 52 and Glu 35, and it is generally accepted that Glu 35, which is in a hydrophobic environment and which would be expected to be unionised at the optimum pH for lysozyme action, acts as a general acid catalyst in the cleavage of the glycoside. General acid catalysis alone is however not sufficient to account for the catalytic properties of lysozyme, and controversy exists over the source of the extra catalysis required. Asp 52, which has a low pK value and which would be ionised at the optimum pH for lysozyme activity, is an obvious choice of a residue which could participate in the mechanism, either by stabilising an intermediate carbonium ion by ion-pair formation, or by intramolecularly assisting expulsion of the aglycone, both after protonation of the glycosidic oxygen atom by Glu 35. However, as the best substrates for lysozyme are polysaccharides of N-acetyl glucosamine and its derivatives, it is also possible that, with these substrates, the neighbouring acetamido group could intramolecularly assist glycolysis. Although detailed studies of the glycolysis and transglycosylation reactions catalysed by lysozyme have been made, the question of whether neighbouring group participation assists in the catalytic mechanism remains unresolved. The kinetic analysis of hydrolysis of normal substrates, (for example β(1 → 4) linked polymers of N-acetyl glucosamine), is complicated by non-productive binding and transglycosylation, and use has to be made of model substrates such as aryl glycosides of N-acetyl glucosamine oligomers; although poor substrates these are hydrolysed with Michaelis-Menten Kinetics, but results obtained from them are subject to the criticism that one cannot be certain that they bind exactly as, or even are subject to the same mechanism of hydrolysis as normal substrates. Modification of the enzyme rather than the substrate is therefore a more satisfactory approach, and this thesis describes the attempted conversion of aspartic acid 52 to asparagine, a modification which should not affect the accessibility of the active site, but which should enable a more definite conclusion to be made as to the degree of involvement of this residue in the catalytic mechanism. The reaction chosen for effecting this modification was that of the acid with a carbodiimide in the presence of ammonia, and in order that the modification would be specific for Asp 52 the diimide was incorporated into the compound (I) below, which was synthesised in eight stages. (I) should bind to lysozyme with its sugar residue in site C and with its C-l side chain extending down the cleft to the Asp 52 region. [For the diagram, please consult the PDF.] Reaction of lysozyme with this compound has been shown to be complex; inactivation of the enzyme was found to be independent of added nucleophile (and must therefore involve some intramolecular reaction or the irreversible attachment of the inhibitor), and a variety of products of more or less acidic nature than native enzyme was formed. The separation and purification of these modified enzymes by cation-exchange chromatography is described. The major product which was eluted before native enzyme in such a separation was unmodified at Asp 52 or Glu 35, and, as its lower rate of lysis of M. lysodeikticus cell walls than native enzyme (40-50%) could be accounted for entirely by a loss of binding ability, a binding site modification is suggested for this material. The major modified enzyme of less acidic nature than native enzyme was shown to contain a catalytic site modification, and an active site peptide containing residue 52 was isolated. This residue was shown not to be aspartic acid by electrophoresis at pH 6.5, although the modification was labile to acid hydrolysis and leucine aminopeptidase digestion, both of which released aspartic acid. New methods for the separation and identification of asparagine by amino-acid analysis and on paper are described, and the modified residue 52 was shown not to be asparagine by these methods. The modification of aspartic acid 52 in lysozyme by the diimide I was shown to be associated with a loss of binding ability for small oligosaccharides, the product being completely inactive and showing no ability to bind tri-(N-acetyl)-glucosamine. Possible intramolecular reactions and modes of irreversible attachment of the inhibitor to Asp 52 are considered, and it is concluded that the latter is in all probability responsible for the inactivation of the enzyme. The Asp 52-modified enzyme has been crystallised and it is hoped that X-ray diffraction data can be obtained from the crystals, in which case the nature of the modification should be confirmed. Although the fact that the derivative does not bind substrate makes it unsuitable for investigation of the function of Asp 52 in the catalytic mechanism the fact that it contains a blocked sub-site C (which is normally the strongest binding site) means that it could be used in valuable studies of the binding of saccharide residues to other sites, and in particular to sites E and F. Preliminary evidence as to the strong binding of N-acetyl glucosamine in site E has been obtained.

This thesis is concerned with the physical structure and the protein-protein interactions in the two oligomeric proteins, Phosphofructokinase (PFK) and Aldolase. PFK is known to play an important part in the regulation of glycolysis and the activity of the enzyme is modified by a variety of low molecular weight metabolites including ATP, AMP, F-6-P, FDP and citrate. Physical studies on PFK were, therefore, made with particular reference to the conformational changes which might accompany the regulation of activity. The studies on Aldolase are less extensive and represent a continuation of earlier work (K.R.Leonard, Part II thesis, 1966). In contrast to PFK, aldolase is not a regulatory protein. It is, however, composed of subunits and therefore serves as a useful comparison for PFK. Aldolase is also related to PFK in that they are both enzymes of Glycolysis, having adjacent places in the pathway, so that the product of the PFK reaction, FDP, is also the substrate of the Aldolase reaction. Aldolase was purchased from a commercial source and was found to be quite suitable for physical studies. The commercially available PFK, however, was not suitable for physical studies for a number of reasons, including the presence of a large amount of inactive protein which, was visible in the sedimentation patterns and a strong tendency to precipitate when dialysing against buffer to remove ammonium sulphate. The PFK used was therefore prepared in this laboratory by an adaptation of a published method. The sedimentation velocity experiments using schlieren optics to measure protein concentration confirmed that PFK undergoes reversible self-association above 0.5mg/ml. This aspect of the protein-protein interactions of PFK was investigated in some detail. Evidence for a reversible association was manifested by the appearance of the sedimentation pattern of the pure protein (three peaks with s_app of approx. 13 S, 18 S and 22-30 S). Further evidence was the behavior of the fastest peak which increased in s_app from 22 S at 0.5mg/ml to 30 S at 10mg/ml in direct contrast to the normal decrease in s_app with increasing concentration for a non-interacting species. This association of PFK was found to be prevented by mild alkaline pH, high ionic strength, low concentrations of SDS (detergent) and low concentrations of PCMB (thiol reagent). Addition of any of these reagents to PFK solutions resulted in a change ox the sedimentation pattern to a single boundary with s_app = 12 S (approx). This corresponds to the slowest sedimenting species in the pattern for native PFK at high concentration and is probably the unassociated 'Monomer' of the equilibrium. Measurement of s_app for PFK at the concentration level of the enzyme assay (approx. landmu;g/ml) gave a value of about 11 S for the active species, which indicated that PFK is present as the 'monomer' at this low concentration. The 12 S material, which was obtained by direct pH adjustment to pH 10.5, slowly broke down to give protein sedimenting at lower speed. The presence of 0.5mM FDP, however, had a marked stabilising effect, and protein dialysed into pH 10.5 buffer containing FDP sedimented as a sharp symmetrical boundary with s⁰_20,w of 11.9 ± 0.4. Similar concentrations of F-6-P, ATP and AMP had no stabilising effect on the 12 S protein. The diffusion coefficient of the FDP-stabilised material was measured by the spreading boundary method and a value of D⁰₂₀ equal to 3.2 ± 0.3 andtimes; 10^-7 obtained. This combined with the sedimentation coefficient gave a value of 3.4 ± 0.3 andtimes; 10⁵ for the molecular weight. This agreed with the value of 3.45 ± 0.1 andtimes; 10⁵ obtained from sedimentation equilibrium. Equilibrium centrifugation studies of a series of concentrations of PFK over the range 0.5-10mg/ml enabled curves of weight-average and Z-average molecular weights against concentration for the polymerisation to be constructed. These were found to be sigmoid in shape, and the Z-average tended to a value of 3-4 andtimes; 10⁵ at low concentration (where the 12 S species only will be present) and to a plateau of about 5.5 andtimes; (monomer mol.wt) at about 7mg/ml, followed by a slight reduction in molecular weight caused by non-ideality at higher protein concentration. These curves were compared with calculated theoretical curves for a number of model polymerising systems, and it was deduced, from the shape of the curves, that the association was to a closed polymer rather than to an indefinite, open polymerisation. A calculated curve which was a reasonable fit to the experimental points was obtained by assuming that the degree of polymerisation ,n, was six, that there was a negligible concentration of intermediate polymers present at equilibrium and that BM_w (the non-ideality term) had a value close to the theoretical value for a globular protein. A tentative closed-shell structure was proposed for the hexamer. Assuming that n = 6, the equilibrium constant for the association per mole of hexamer was estimated to be 6 andtimes; 10²⁵ Ltr⁵/Mole⁵ and hence the value of andDelta;G⁰ per mole of hexamer was -35 Kcal. at 20°C. Data for M_z was also obtained at 2°C and this, combined with the results at 20°C gave values of andDelta;H⁰ = 36 Kcal and andDelta;S⁰ = 240 e.u. per mole of hexamer, for the association. These results show that the principle contributing factor to the free energy of the interaction is the gain in entropy which results from the release of bound water as the monomers associate. This would implicate either hydrophobic or electrostatic interactions as chief contributors to the protein-protein binding. The effect of high salt concentration, which prevents the association, unequivocally supports the importance of electrostatic interactions in the polymerisation. PFK at pH 12 in the presence of SDS had a molecular weight of about 8.8 andtimes; 10⁴, estimated by sedimentation equilibrium, which is about one fourth of the molecular weight obtained for the 12 S protein. The molecular weight in 6M GuHCl, 0.1M 2-mercaptoethanol was 7.6 ± 0.5 andtimes; 10⁴ by sedimentation equilibrium and about 8 andtimes; 10⁴ calculated from the value of the sedimentation coefficient. Allowing for experimental error, in particular in the value of the partial specific volume, these results agree with the figure obtained at high pH in the presence of SDS. Thus, the 12 S protein is made up of four subunits, each consisting of a single polypeptide chain. In order to test whether the association of PFK had any effect on the enzymic activity it was necessary to carry out the assay at the 1mg/ml enzyme level. The F-6-P activity of PFK is too high to be measured by normal spectrophotometric methods at this concentration. However, PFK will also catalyse the phosphorylation of G-l-P at a much lower rate and this activity can be assayed at the higher concentration. It was found, that there was no change in the G-l-P specific activity over the concentration range 0.2-2.0 mg/ml where the enzyme is known to associate.

Pseudomonas putida metabolises aromatic carboxylic acids by convergent pathways. The regulons for the synthesis of the enzymes involved are controlled by induction and repression by intermediates of the pathways. The results of previous studies in this system suggest that the organism should exhibit a selective enzymic response to mixed aromatic substrates. It has been the aim of this study to predict the enzymic response to mixtures of mandelate, benzoate and parahydroxybenzoate and then to check these predictions experimentally. The affinities of these three inducer substrates for their regulons have been determined by measuring the initial rates of enzyme synthesis over a range of inducer concentrations in cultures growing on glucose. Glucose itself was shown to have very little catabolite repressor effect in this system. From the results, an induction constant, defined as the inducer concentration for half-maximum rate of induction, was estimated for each substrate. Those for the benzoate and mandelate regulons were similar (c. 0.5 mM) and much higher than that for parahydroxybenzoate hydroxylase (< 0.01 mM). By measuring the repressor effect or each aromatic substrate on the induction by one of the other substrates, repression constants (K_rep) were also obtained that are a quantitative measure of the affinity of the repressor for that regulon. Mandelate had no repressor activity and the benzoate regulon was not repressed by parahydroxybenzoate. The mandelate regulon was severely repressed by p-OH benzoate (K_rep = 0.02 mM) while the effect of benzoate on the parahydroxybenzoate regulon was much less pronounced (K_rep = 1.0 mM) . Benzoate had a transient,severe effect on the mandelate regulon and, at high concentrations, a moderate permanent form of repression was observed. The enzymic reponse of the organism following a changeover from growth on a single aromatic carbon source to growth in the presence of an additional aromatic carbon source was predicted by comparing the induction and repression constants for the substrates concerned. Batch cultures were used to provide an indication of the steady state enzymic reponse, which was obtained using a nitrogen-limited chemostat operated at high growth rates. In general, the predictions were experimentally validated. Benzoate oxidase was always fully induced in mixtures containing benzoate, while the mandelate enzymes were almost completely repressed in the presence of parahydroxybenzoate. The mandelate regulon was only partially induced in mixtures containing mandelate and benzoate as was the parahydroxybenzoate regulon in cells grown on parahydroxybenzoate and benzoate. The extent to which the enzymes for the less preferred carbon source were induced was affected by the relative concentrations of the substrates. The results indicate that comparison of induction and repression constants can be used to predict the enzymic behaviour of this organism towards mixed aromatic substrates and it is suggested that this may form the basis of a general method for predicting enzymic adaptation to mixed substrates. Relative substrate utilisation can only be inferred from the induction and repression constants if complete repression of an enzyme system is predicted. Further data are necessary whenever the predictions indicate that more than one enzyme system will be induced. Thus, in this study, it was predicted, and experimentally verified in the chemostat, that mandelate would not be used by cells growing in the presence of mandelate + p-OH benzoate. When induction of two enzyme systems we predicted, the relative utilisation of the substrates was qualitatively related to the enzymic content of the cells. In contrast to other systems where the reaction to mixed substrates seems to be directed towards the selection of the substrate capable of supporting the fastest growth rate, Pseudomonas putida selects, not the best growth substrate (mandelate), but the substrate requiring the synthesis of fewer enzymes for its metabolism (benzoate). Factors in the environment acting as possible selective agents are discussed. In the course of this study, it was also found that the synthesis of the mandelate enzymes wan delayed when low external inducer concentrations were used. At high concentrations, enzyme synthesis rapidly followed the addition of inducer. A pre-induction or maintenance effect was also observed in which cells induced at saturating concentrations of mandelate were able to maintain a maximum rate of induction when transferred to sub-saturating inducer concentrations. Induced cells were able to accumulate mandelate from low external concentrations. These results are discussed in terms of an inducible mandelate transport system. The severe transient repression of the mandelate regulon by benzoate was greatly reduced by pre-induction with mandelate. Benzoate inhibited the accumulation of mandelate by the cells and the action of benzoate on the mandelate regulon is discussed in relation to the possible induction mechanisms of the mandelate regulon.

The mechanism of papain-catalysed hydrolyses of <u>K</u>-acyl-amino acid derivatives is known to involve initial formation of an enzyme-substrate complex, followed by acylation of the active-site thiol of Cys-25 by the substrate to give a covalent acyl-enzyme intermediate, and subsequent deacylation of this intermediate. This thesis is concerned with an investigation of the various steps in catalysis, mainly by a study of steady-state kinetics of hydrolyses of specific substrates. The acylation step for a number of anilide substrates was found to be completely or partially rate-determining. The acylation rate constants (k₊₂) and the substrate constants (K_s) can be evaluated from the determination of Michaelis-Menten parameters (k₀ and K_m). For hippuryl <u>p</u>-substituted anilides and closely related analogues, k₀=k₊₂ and K_m=K_s. For <u>m</u>- and <u>p</u>-substituted anilides of acetyl-L-phenylalanylglycine (APG), k₊₂ were found to be comparable with the deacylation rate (k_d); k₊₂ and K_s were calculable once k_d was known. The deacylation step for the ester substrates studied was found to be the rate-determining step (k₀=k_d). The specificity of the enzyme for the <u>N</u>-acyl moiety of the substrate is probably determined to a large extent by hydrophobic interactions between the enzyme and the substrate, since (k₀/K_m) for <u>N</u>-acylglycine esters and anilides increase in magnitude in the order acetyl- < benzoyl- < benzyloxycarbonyl < acetyl-L-phenylalanyl-glycine derivatives. The greater reactivity of APG esters and anilides, despite the fact that hydrophobicity of the acyl moiety is comparable with benzoyl- and benzyloxycarbonyl-glycyl moieties, however, indicates that polar interactions are also significant. This conclusion is supported by the observation of differences of values of analogues of hippuric acid esters and anilides, in which the amide link has been modified. These differences are not solely determined by the inherent electronic properties of the acyl group, as measured by hydroxide-ion-catalysed rates of <u>p</u>-nitrophenyl esters. The amide link of <u>N</u>-acylamino acid derivatives may interact through its carbonyl oxygen with a polar group of the enzyme, possibly the γ-amide group of Gln-19, and the interaction may be important for the catalytic activity. As has been found by other workers, the enzyme interacts specifically with the side chain of L-arginine and it has been found that (k₀/K_m) for α-<u>N</u>-benzoyl-DL-arginine <u>p</u>-nitroanilide is 150 times that for hippuryl <u>p</u>-nitroanilide. The values of K_s for hippuryl anilides are in the range of 2-60 mM; those for APG anilides, 1-4 mM. Binding specificity is not a reliable indicator of papain activity for a given substrate, although frequently K_s is smaller for a more specific substrate. Free energy of binding interactions must be utilised in order to bring the enzyme-substrate complex to an active state for subsequent reactions. Comparison of activity among inhibitors and substrates derived from APG, methoxycarbonyl-L-phenylalanyl-glycine and hippuric acid indicates that the enzyme-substrate complex is subject to a strain around the susceptible bond, the amount of which accounts for its lability. An 'induced-fit' mechanism, involving conformational adjustment of the enzyme in order to bring the catalytic groups into proper positions, does not provide a satisfactory explanation for the experimental results. Among individual rate steps represented by K_s, k₊₂ and k_d, the best measure of specificity is k₊₂, as demonstrated by comparison between anilides of APG, hippuric acid and analogues. k₊₂ values for APG anilides are 2-20 sec^-1, while for hippuryl anilides, 3-50 msec^-1. The 1,000-fold difference is due to both enthalpy and entropy of activation, as concluded from studies of the effect of temperature on the acylation rates of APG <u>p</u>-chloroanilide, hippuryl <u>p</u>-chloroanilide and hippuryl <u>p</u>-bromoanilide. In contrast with k₊₂, k_d is not very sensitive to the nature of the <u>N</u>-acyl moiety. For acylglycine esters k_d values lie normally between 1 and 10 sec^-1, although (k₀/K_m) may differ by a factor of more than 10⁴. Therefore, like K_s, k_d is not a good indicator of specificity. A Hammett plot of log k₊₂ versus andsigma; for hippuryl anilides gives a andrho; value of -1.2, indicating electrophilic catalysis in the acylation step. The same andrho; value is obtained from a similar Hammett plot for APG anilides. Electronic effects on k₊₂ for both series of substrates are therefore similar. It is proposed that in the acylation of papain by anilide substrates the leaving groups are protonated through general acid catalysis. The donated proton is probably that originally forming a hydrogen bond between the essential imidazole group of His-159 and the essential thiol of Cys-25, which therefore may attack the carbonyl carbon of the substrate in the form of a thiolate anion. Available results do not allow a determination of the exact sequence of these processes. From the principle of microscopic reversibility the same processes in reverse must take place in the deacylation step. Experiments on kinetics of papain-catalysed hydrolyses of methyl and <u>p</u>-nitrophenyl esters in the presence of nucleophiles (methanol, aminoacetonitrile, glycinamide) support the suggestion of various workers that the rate- determining step in deacylation is the general-base-catalysed nucleophilic attack of water or other nucleophiles on the thiolester intermediate. A tetrahedral intermediate formed is rapidly decomposed to give the acid or other acyl derivatives. For hippuryl-papain and possibly for other acyl-papains, experimental results on the effects of aminoacetonitrile can be interpreted as indicating the existence of an alternative pathway. This may involve an acyl transfer to another group of the enzyme, or a conformational change, followed by relatively rapid hydrolysis and aminolysis. The acylation step depends on two prototropic groups with pK_a(app) of approximately 4.2 and 8.3. The pK_a( values of these groups in the free enzyme, as determined by the pH profile of (k₀/K_m) for p-nitrophenyl esters and APG <u>p</u>-nitroanilide, are very similar to those in the enzyme-substrate complex, as determined by the pH profile of k₊₂ for APG <u>p</u>-nitroanilide. If the two pK_a( values represent protonation and deprotonation of the hydrogen-bonded thiol-imidazole system respectively, the binding of the substrate cannot result in the destruction of this hydrogen bond. As expected, pK₂ was not observed in deacylation of APG-papin. In the deacylation step, pK₁ is sensitive to the nature of the acyl moiety and varies in the range 3.1-4.7. The pK₁ value tends to be low for a specific substrate or a substrate with hydrophobic groups in the acyl moiety. This observation can be interpreted in terms of shielding of the thiol-imidazole system by the acyl moiety from hydronium ions. Apart from steric effects, shielding by hydrophobic groups also gives rise to a decrease in local dielectric constant, which could further depress the pK_a of a cationic acid such as the thiol-imidazolium system.

The functions of biological macromolecules are discussed in relation to their structure as linear polymers adopting specific conformations. If such molecules can be crystallised, their three dimensional structure can be determined at atomic resolution by the methods of X-ray crystallography whose applicability to the analysis of protein structures is considered. This is followed by an account of their application in the study of a glycolytic enzyme, triose phosphate isomerase (TIM), leading up to the calculation of an electron density map at 2.5 Å resolution. The characteristics of crystals of chicken TIM are described and a detailed account is given of the measurement of X-ray diffracted intensities, the determination and refinement of the positions of heavy atom binding in isomorphous derivatives of TIM, and their subsequent use for calculating the phases of the protein structure factors. The search which was conducted for possible heavy atom derivatives produced several which have two sulphydryl sites in common. The problems resulting from these common sites and the systematic errors they can cause in the phase determination are examined in relation to the suitability of different methods of refinement in this situation. The electron density map of TIM at 6 Å Resolution is described and a preliminary discussion of the 2.5 Å map is given.

This thesis is concerned with properties of the excited states of flavins. They may be involved in some photobiological processes and are models for flavoprotein catalysis. Understanding the spectral properties of flavins is important for interpreting the nature of intermolecular interactions in flavoproteins. The absorption spectra of many flavin derivatives are similar, but the two long wavelength bands are sensitive to both solvent and substituents. Two independent methods for finding the angle between these absorption oscillators are described. The first, measurement of the fluorescence polarisation of a random but rigid array of the chromophore, is a well known technique. The second is novel, and requires the preparation of a partially ordered solution of flavin by selective blanching with plane polarised light of wavelengths absorbed by only one of the absorption bands. The theory of this technique is developed, and it shows that the angle between the absorption oscillators can be calculated from the dichroic ratios of the extinctions of the two bands after bleaching. This angle does not vary much for different flavins, and the same value (around 30°) is found from both fluorescence polarisation and dichroic bleaching. The directions of the oscillators relative to the axes of the molecule cannot be measured by either technique, but the effects of substituents on absorption spectra suggest that the visible band is polarised approximately parallel to the long axis of the flavin. The first excited singlet of a flavin loses some of its energy by emission as fluorescence, but it also participates in singlet-to-singlet energy transfer. This is demonstrated by fluorescence for two flavin derivatives, using a methyl cellulose film as the supporting matrix. Optical observations are possible under these conditions at higher flavin concentrations than are easily studied in liquid solutions. Excited singlets may also lose energy by intersystem crossing to the lowest triplet state, but no delayed light emission can be detected from aqueous flavin solutions. In a rigid polyacrylamide matrix, however, there is delayed emission, even at room temperature. This is identified as E-type delayed fluorescence on the basic of its spectrum, its temperature and its exciting light intensity dependence. At lower temperatures the delayed fluorescence decreases, and phosphorescence emission dominates. The intensities and lifetimes of the delayed emissions of 12 flavins have been measured in the range 77°-315°K. On plotting the logarithm of intensity or of the reciprocal of half-life (decay is exponential) against the reciprocal of temperature biphasic Arrhenius plots are obtained. These can be interpreted in terms of the activation energies of the various rate processes associated with triplet production and inactivation. The activation energies from delayed fluorescence intensities and half-lives (5.9 and 2.6 kcal/mole respectively for FMN) add to a value close to the energy separating the spectroscopic fluorescence and phosphorescence maxima. A kinetic scheme is proposed to account for this. The activation energy for non-radiative decay of the triplet to the ground state can be calculated from the temperature dependence of phosphorescence, and values of about 100 cal/mole are found. Photochemical reactions of the flavin triplet both in the presence and absence of oxygen have been studied. Anaerobically flavins photo-oxidise many amino acids and amines in their high pH forms, with production of dihydroflavin. When DL-phenylglycine is the reducing agent equimolar amounts of benzaldehyde, carbon dioxide and dihydroflavin are produced. The dependence of reaction rate on reducing agent concentration shows saturation behaviour at high concentrations, which is consistent with a steady state kinetic treatment for the reaction. The sensitivity of rates to substituents in the DL-phenylglycine can be described by a Hammett andrho;-value of -1.1. On the basis of their photo reactivities towards a series of flavin derivatives it is suggested that DL-phenylglycine and benzylamine react with a similar rate determining step (probably hydrogen abstraction). This is different from the photoaddition of phenylacetate. Anaerobic photoreductions are quenched by some aromatic compounds. From the effects of temperature and by comparison with quenching by potassium iodide it is concluded that inhibition is not simply a collisional process. FAD reacts more slowly in photo seduction than FMN, but in concentrated urea solutions the difference is reduced. Inhibition by aromatic compounds is also reduced under these conditions. This surges to that the forces between flavin triplet and quencher are at least partially hydrophobic. Aerobic flavin photochemistry is followed by measuring oxygen uptake. Two general types of behaviour are found for oxygen concentrations from 2 andmu;M upwards. Either the reaction rate increases as oxygen is consumed (type I) or it falls (type II). The type I reaction pathway is slow photoreduction followed by rapid reoxidation of dihydroflavin. Evidence for this comes from the dependence of reaction rate on oxygen concentration and pH, and from the Hammett andrho;-value for substituted phenylglycines which is similar to that of anaerobic photoreduction. Oxygen is also a triplet quencher. Rate constants for the reduction and quenching of FMN triplets by EDTA, diethanolamine and DL-phenylglycine (all type I substrates) are deduced from the dependence of reaction rates on oxygen concentration. In each case quenching is over 10 times as efficient as reaction. The rate constants for all three substrates are similar, increasing in the order DL-phenylglycine < EDTA < diethanolamine. The values for EDTA are 1.38andsdot;10⁷ M^-1sec^-1 for reaction and 1.6andsdot;10⁸ M^-1sec^-1 for quenching. These rate constants can be used to predict a concentration dependence of anaerobic photoreduction which agrees well with experiment. The most likely mechanism for quenching is a rapidly reversed electron transfer. Whereas type I substrates are reducing agents for flavin triplets, type II substrates are typically aromatic molecules which quench anaerobic photoreduction. Tryptophan, for example, has type II kinetics and reacts by a complex mechanism in which product inhibition of reaction occurs. The potential of flavin photosensitized destruction of amino acids as a method for selective chemical modification of proteins is investigated. Tryptophan, tyrosine, histidine, methionine, cysteine and phenylalanine all react but none of them has the same pH-profile of reactivity. Preliminary experiments on two proteins, L-glutamate dehydrogenase and D-amino acid oxidase show that the method has promise. An investigation of the behaviour of flavins when they act as Hill oxidants with isolated spinach chloroplasts has also been made. In this reaction light is absorbed t-y the chloroplast and flavin reacts in its ground state. The rates of reduction of 10 derivatives have been compared. The order of reactivity la different from that found for some model systems; dark reduction by NADH or dihydrolipoic acid, the redox potentials, and -photo re duct ion by DL-phenylglycine. The rates are independent of flavin concentration, but vary with the concentration of chloroplasts and light intensity. It is suggested that the reactivity to reduction by chloroplasts depends on binding between chloroplasts and flavin as well as on the electronic character of the flavin. The partition coefficients of flavins between water and butan-1-ol lend some support to this hypothesis.

The development of the mammary gland is strictly directed by hormones, local signaling, epithelial-mesenchymal cross talk, as well as participation of innate immune cells. On the other hand, the dysregulation of this orchestration may initiate breast cancer and facilitate its progression. In our study Pax9 was found, for the first time, significantly reversely correlated with breast cancer malignancy, being reduced or absent in human DCISs (96%) and invasive breast cancers (78%), as well as in MMTV-Neu and MMTV-PyMT induced mouse mammary tumours, while it is expressed in normal human and mouse mammary epithelium. By a full-range expression investigation, using semi-quantitative RT PCR and immunohistochemistry, covering all developmental stages of mouse mammary gland, Pax9 was found to be expressed in the ductal epithelium with a strict spatial-temporal pattern, with an expression peak at puberty. Reduction or deletion of Pax9 expression, using Pax9 hypomorphs and mammary gland-specific knockout mouse models, resulted in ductal branching delay during puberty, alveolar formation at the wrong position during pregnancy, disrupted epithelial cell apoptosis and engulfment of excess milk fat globulin during postlactational involution. Mammary ductal epithelial cell detachment, basement membrane disruption and tumour-like structure expansion have been found in the mammary glands of parous mice. Taken together, we found that Pax9 functions in the process of mammary epithelial cell differentiation, basement membrane integrity, apoptosis and epithelial cell engulfment. 4 By immunohistochemistry and western blot analysis of the mammary gland during involution, we identified the Stat signaling pathway as a candidate downstream pathway affected by Pax9 deficiency in the mammary gland, which may be responsible for apoptosis delay. Expression microarray profiling of Pax9 deficient and control mammary glands showed the increase of insulin growth factor binding protein 5 (Igfbp5, an essential regulator of mammary gland involution), monocyte to macrophage differentiation-associated (Mmd, immune and inflammation associated genes), and MMP3 and MMP12 (metalloproteinase) genes, and the decrease of inhibitor of DNA binding 2 (Id-2, functioning in mammary cells with low proliferation and invasiveness). Furthermore, myosin-related genes were strikingly up-regulated, which may be a cellular stress response to the milk stasis from impaired involution in the Pax9 deficient mammary gland. All these phenotypes we discovered in the mutants and molecular changes suggested by immunohistochemistry and gene expression profiling during involution, provided us with candidate networks regulated by Pax9 in mammary gland development. Further elucidation of these clues may help us to understand the multiple pathways in which Pax9 takes part in normal mammary differentiation, and the underlying mechanisms how its dysregulation may promote breast cancer formation and progression.

No description available.