Regulation and control of L-arabinose catabolism in Aspergillus niger

This thesis describes studies on the biochemical properties and regulation of L-arabinose metabolism and arabinan degrading enzymes of Aspergillus niger. We focused on the investigation of the catabolic pathway, firstly by isolating pathway specific regulatory mutants using a newly developed selection system and, secondly, by purifying the enzymes and characterising their kinetics for use in metabolic control analysis. Finally, by cloning genes encoding these enzymes we were able to analyse expression of these genes.<o:p></o:p></span>Using a D-xylulose kinase deficient strain we developed a mutant selection system that identified genes involved in pentose catabolism and their regulation. The A.<st1:place w:st="on"><st1:country-region w:st="on">niger</st1:country-region></st1:place> strain carrying the xkiA1 mutation lacks D-xylulose kinase and cannot grow on pentoses, such as D-xylose and L-arabinose, but these sugars still repress the use of other carbon sources such as D-gluconate. We used this genetic background to select for pentose-derepressed mutants on media containing combinations of gluconate with xylitol, L-arabinose or D-xylose. A subset of these mutants was further analysed and turned out to have rather interesting properties as described in chapters two and three.<o:p></o:p></span>One of the mutants isolated using this method carried a mutation called xtlA36 and is described in chapter two. This mutation results in a severe decrease in xylitol consumption, suggesting that xtlA36 inactivates a xylitol transporter and opens the way of isolation of genes encoding the corresponding genes. Two other A.<st1:place w:st="on"><st1:country-region w:st="on">niger</st1:country-region></st1:place> mutants, carrying araA and araB are specifically disturbed in the regulation of the arabinanase system in the presence of L-arabinose. Expression of three arabinanolytic genes, abfA, abfB and abnA , is substantially decreased or absent in the araA and araB , strains compared to the wild-type when incubated in the presence of L-arabinose or L-arabitol. In addition, the intracellular enzyme activities of L-arabitol dehydrogenase and L-arabinose reductase, involved in L-arabinose catabolism, were decreased in the araA and araB strains.<o:p></o:p></span>L-arabitol, most likely the true inducer of the arabinanolytic and L-arabinose catabolic genes, accumulates to a high intracellular concentration in the araA and araB mutants. This indicates that the decreased expression of the arabinolytic genes is not due to lack of inducer accumulation. Therefore, we propose that araA and araB are mutations in positively acting components of the regulatory system involved in the expression of the arabinanase encoding genes and the genes encoding the L-arabinose catabolic pathway.<o:p></o:p></span>Chapter four describes cloning of the A.<st1:place w:st="on"><st1:country-region w:st="on">niger</st1:country-region></st1:place> D-xylulose kinase encoding gene ( xkiA ) by direct complementation of the strain deficient in D-xylulose kinase activity. This enabled us to investigate the expression of xkiA in the presence of L-arabinose, L-arabitol, and D-xylose. Although XKI is part of the D-xylose catabolic pathway, expression of xkiA appeared not to be mediated by XLNR, the xylose-dependent positively acting xylanolytic regulator. Expression of xkiA is subject to carbon catabolite repression (ccr) but the wide domain regulator CREA is not directly involved. Using the araA and araB strains described in chapter three, we showed that xkiA is under control of the arabinanolytic regulatory system.<o:p></o:p></span>Overexpression of xkiA enabled us to purify the encoded D-xylulose kinase enzyme. The molecular mass, determined using Electrospray Ionization Mass Spectrometry (ESI-MS) concurred with the calculated molecular mass of 62816.6 Da. The activity of D-xylulose kinase is highly specific for D-xylulose. Kinetic parameters were determined, including K_m (D-xylulose), 0.76 mM and K_m (ATP), 0.061 mM.<o:p></o:p></span>In a D-xylulose kinase deficient strain a higher accumulation of intracellular arabitol and xylitol correlated to increased transcript levels of the genes encoding arabinan and xylan degrading enzymes, respectively. This supports the suggestion that L-arabitol may be the specific low molecular weight inducer of the genes involved in arabinan degradation. It also suggests a possible role for xylitol in the induction of xylanolytic genes. Overproduction of XKIA did not reduce the size of the intracellular arabitol and xylitol pools, and had no effect on expression of genes encoding xylan and arabinan degrading enzymes. This suggests that the enzymes preceding D-xylulose kinase in the L-arabinose/D-xylose catabolic pathway probably have more control on the flux through this pathway than D-xylulose kinase itself.<o:p></o:p></span>In chapter five, we cloned the genes encoding</span><span lang=EN-US style='font-family: Sylfaen'>A.<st1:place w:st="on"><st1:country-region w:st="on">niger</st1:country-region></st1:place></span><span lang=EN-US style='font-family:Sylfaen'>L-arabitol dehydrogenase ( ladA ) and xylitol dehydrogenase ( xdhA ), and produced the enzymes in Escherichia coli . Analysis of the substrate specificity showed that LADA is most active on L-arabitol and also has significant activity on xylitol, but only low activity on D-sorbitol and galactitol. XDHA has the highest activity on xylitol, significant activity on D-sorbitol, but very low activity on L-arabitol. The higher activity on sorbitol for XDHA is in agreement with the amino acid similarity of the different enzyme classes, since a phylogenetic tree of L-arabitol dehydrogenases, xylitol dehydrogenases and sorbitol dehydrogenases (SDH) suggests that xylitol dehydrogenases are more similar to sorbitol dehydrogenases than L-arabitol dehydrogenases. Expression analysis of the pentose catabolic pathway genes confirmed the model in which an arabinose specific regulator activates the expression of all genes required for the conversion of L-arabinose to D-xylulose-5-phosphate. In addition, XLNR regulates the first step and, to a lesser extent, the other steps of the conversion of D-xylose into D-xylulose-5-phosphate.<o:p></o:p></span>Using dye-affinity chromatography we isolated enzymes of the L-arabinose and D-xylose catabolic pathways that had not been described previously. Using the complete set of kinetic parameters a metabolic model was constructed which we used to perform steady state metabolic control analysis (chapter six). The metabolic model was used to analyse flux and metabolite concentration control of the L-arabinose catabolic pathway. The model predicts that flux control does not only reside at the enzyme following the intermediate with the highest concentration, L-arabitol, but is distributed over the first three steps in the pathway, preceding and following L-arabitol. Flux control appeared to be strongly dependent on the intracellular L-arabinose concentration. At 5 mM intracellular L-arabinose, a level that resulted in realistic intermediate concentrations in the model, flux control coefficients for L-arabinose reductase, L-arabitol dehydrogenase and L-xylulose reductase were 0.68, 0.17 and 0.14 respectively. This analysis can be used as a guide to identify targets for metabolic engineering aiming at either flux or metabolite level optimisation of the L-arabinose catabolic pathway of A.<st1:place w:st="on"><st1:country-region w:st="on">niger</st1:country-region></st1:place> .<o:p></o:p></span>In chapter seven the results from chapters two through six are discussed in light of possible engineering applications and recent scientific developments such as genomics