The glucose-6-phosphate dehydrogenase encoding genes from Aspergillus niger and Aspergillus nidulans

Glucose-6-phosphate (G6P) is a central metabolite, that can either be metabolised via the glycolytic and tricarboxylic acid cycle to generate ATP, or converted into storage molecules or can be directed to the pentose phosphate pathway to yield NADPH and various pentoses. This thesis focuses on one of these G6P consuming reactions, catalysed by glucose-6-phosphate dehydrogenase (G6PD), in which G6P is oxidised and NADP <sup>+</SUP>acts as electron acceptor. The central theme of this thesis is the role of glucose-6- phosphate dehydrogenase (G6PD) enzyme activity in the generation of the cytosolic NADPH pool. This cytosolic NADPH pool serves as the reducing agent in a large number of biosynthetic reactions, in the protection against oxidation damage and in the utilisation of nitrate and certain pentoses. Only a limited number of catabolic reactions generate NADPH of which the one catalysed by G6PD is usually considered to be the most important.If G6PD were the major producer of NADPH, one would expect 1) G6PD enzyme activity to respond to changes in NADPH demand and 2) G6PD mutations to be extremely deleterious or even lethal. According to these criteria, examples can be found in the scientific literature, where G6PD appears to be either vital for cytosolic NADPH production or appears to be completely superfluous. <em>Aspergillus nidulans</em> and <em>Aspergillus niger</em> were chosen as model systems because of their ability to metabolise a great variety of carbon- and nitrogen sources and their well developed genetics. These characteristics would allow us to use nitrogen- and carbon sources that influence NADPH consumption and to employ molecular biological techniques to study and manipulate the regulation of G6PD expression.The biochemical characteristics of the purified G6PD enzymes from A. nidulans and <em>A. niger</em> suggest the involvement of these G6PD enzymes in NADPH production. For both pure enzyme preparations, the main modulator of G6PD enzyme activity is the redox potential as is reflected by its stimulation by NADP <sup>+</SUP>and its competitive inhibition by NADPH. Additionally, these observations show that at least part of the regulation of G6PD activity in <em>A. nidulans</em> and <em>A. niger</em> is exerted at the enzyme level. In their biochemical characteristics, their primary and secondary structure the G6PDs from A. nidulans and <em>A. niger</em> are virtually identical. Both fungal G6PD enzymes bind their substrates G6P and NADP <sup>+</SUP>in a random order and exhibit a strict specificity towards them. Whether the presence of two proteins in both pure Aspergillus G6PD preparations, has any physiological significance or is just an artefact due to proteolytic degradation of the native G6PD during the purification procedure, remains to be determined. However, the cloning of the G6PD encoding genes does provide a clue to the origin of the two G6PD proteins. In <em>A. nidulans</em> both G6PD proteins should be derived from the <em>gsd</em> A gene, since only a single G6PD encoding gene could be detected on genomic Southern blots. For the two <em>A. niger</em> G6PD proteins, the situation is less clear because genomic Southern blots revealed the presence of additional DNA bands that cross-hybridised strongly to the <em>gsd</em> A gene and which could represent a second G6PD encoding gene.The <em>gsd</em> A genes from <em>A. nidulans</em> and <em>A. niger</em> have an identical structure; both contain nine introns, which are located at exactly equivalent positions with respect to the coding region. Furthermore, both genes exhibit strong DNA sequence homology in the coding region, but in the introns, 5'- and 3'-flanking sequences, this homology drops significantly. Alignment of the deduced <em>gsd</em> A amino acid sequences with other eukaryotic G6PDs reveals strong homology at the amino acid level and allows localisation of important domains like the putative catalytic site and NADP <sup>+</SUP>bindings sites. The cloning of the <em>gsd</em> A genes a] lows us to manipulate the G6PD activity directly. The increase of G6PD activity by the introduction of multiple copies of the <em>gsd</em> A gene results in <em>A. niger</em> in grossly disturbed growth especially on media containing reduced nitrogen sources. The degree of growth inhibition and the increase in G6PD activity are directly related to the number of functionally integrated <em>gsd</em> A genes. This observation suggests the reduced growth is a direct consequence of the increased G6PD activity. One can explain these phenomena by assuming that increased G6PD activity results in overproduction of NADPH. If for this excess of NADPH no acceptor is available (e.g. nitrate), the redox potential is disturbed and growth is inhibited. A similar experiment with the <em>A. nidulans</em> gene resulted in cotransformants with a single additional <em>gsd</em> A copy, that only slightly overexpressed G6PD end did not exhibit reduced growth.This already shows that despite the similarity in the biochemical characteristics and in primary structure of the enzymes and genes, there are remarkable differences in the regulation of the <em>A. niger</em> and <em>A. nidulans</em><em>gsd</em> A expression. In <em>A. nidulans</em> , but not in <em>A. niger</em> , G6PD enzyme activity and <em>gsd</em> A transcription respond to the increased NADPH demand during growth on nitrate. These observations indicate that in <em>A. nidulans</em> G6PD does play an important role in NADPH production, whereas in <em>A. niger</em> it probably does not. Further evidence for this statement comes from the observation that in <em>A. niger</em> G6PD enzyme activity and steady state mRNA levels do not change in response to growth on xylose or in response to oxidation stress.The homology in the <em>gsd</em> A 5'-upstream regions is limited to four 11 to 22 bp long sequence blocks, named <em>gsd</em> A-boxes, which in both promoters appear in the same order. Homologues of <em>gsd</em> A-boxes 1, 2 and 4 are also encountered in the same order in the <em>S. cerevisiae zwf</em> l promoter. In both Aspergilli, the <em>gsd</em> A-box region contains all transcription start sites: in the case of <em>A. niger</em> there are four and in <em>A. nidulans</em> there are as much as eight. We did not find any conservation, neither in the number of transcription sites nor the site of initiation. In <em>A. niger</em> all deletions within the region containing the <em>gsd</em> A- boxes, reduce transcription to background level while such a deletion in <em>A. nidulans</em> retains 25% of its original transcription. These observations indicate that despite the <em>gsd</em> A-box homology, these regions still differ functionally.From the deletion study of the <em>A. nidulans</em><em>gsd</em> A promoter, it is clear that the nitrate induced increase in G6PD activity is caused by an increased <em>gsd</em> A transcription. Furthermore, evidence has been obtained that the NIRA transcription factor, which mediates induction of the nitrate utilisation pathway, is also involved in the nitrate stimulation of <em>gsd</em> A transcription in <em>A. nidulans</em> . Furthermore our data show clearly, that NIRA plays an important role in the regulation of <em>gsd</em> A transcription in the absence of nitrate. This phenomenon is probably yet another difference in the regulation of <em>gsd</em> A expression between <em>A. nidulans</em> and <em>A. niger</em> . Since <em>A. niger</em><em>gsd</em> A expression does not respond to nitrate induction, it is unlikely that the uninduced level of its <em>gsd</em> A transcription is under the control of NIRA. This NIRA dependence of <em>gsd</em> A transcription might also explain the remarkably different behaviour of the two <em>gsd</em> A genes in cotransformation experiments.In <em>A. nidulans</em> , titration of the NIRA protein by the cotransformed <em>gsd</em> A gene might limit the expression of the cotransformed <em>gsd</em> A copies. At the same time cotransformants with multiple copies of <em>gsd</em> A would be lost on the nitrate containing selection medium, since NIRA titration would render them incapable of nitrate utilisation. These two phenomena would explain the low cotransformation frequency of the <em>A. nidulans</em><em>gsd</em> A gene and low G6PD overproduction observed in <em>A. nidulans</em> cotransformants. Conversely, the absence of NIRA regulation in the <em>gsd</em> A expression in <em>A. niger</em> does not limit overproduction of G6PD in <em>gsd</em> A cotransformants. Furthermore, the absence of NIRA titration in A. <em>niger g</em> sdA cotransformants allows the use of nitrate as an acceptor for the excess NADPH.In contrast to the other members of the nitrate utilisation regulon (e.g. <em>nia</em> D, <em>nii</em> A and <em>crn</em> A) is <em>gsd</em> A expression not subject to nitrogen catabolite repression exerted by the areA gene product. Still, AREA seems to act on the <em>A. nidulans</em><em>gsd</em> A promoter. The physiological significance of this phenomenon remains to be established.The data presented in this thesis show that although the primary structure and biochemical characteristics of an enzyme from two different organisms are virtually identical, there can be significant differences in the regulation of their expression and hence in their physiological function. Of course our data provide only an outline of the regulation of G6PD activity in two Aspergilli and, unfortunately, raise more questions then they answer.Our data indicate that G6PD does not play a crucial role in the generation of the cytoplasmic NADPH pool in <em>A. niger</em> . What is then the physiological function of G6PD enzyme activity in A. <em>niger?</em> Furthermore, as it is difficult to assign a physiological function to the <em>A. niger</em><em>gsd</em> A gene, is it reasonable to assume that the fragments that cross-hybridise to the <em>gsd</em> A gene in digests of <em>A. niger</em> genomic DNA could represent a second G6PD encoding gene? Clearly, the construction and characterisation of a <em>gsd</em> A null mutant should provide important clues to these questions. Since G6PD is not the main producer of NADPH in <em>A. niger</em> , a <em>gsd</em> A null mutation should be viable. This of course raises the question which pathway provides the bulk of the cytoplasmic NADPH in <em>A. niger</em> ?In <em>A. nidulans</em> G6PD enzyme activity is clearly involved in the maintenance of a proper cytoplasmic NADP <sup>+</SUP>/NADPH ratio as its expression responds to the increased NADPH consumption during growth on nitrate. From this observation we would predict that a <em>gsd</em> A null mutant, if not lethal, would be either unable to utilise nitrate or would exhibit reduced growth on this nitrogen source. However, the fact that no G6PD mutants have been found among the nitrate non-utilising mutants suggests that such mutations are in fact lethal. How does <em>A. nidulans</em><em>gsd</em> A expression respond to NADPH consuming processes other than nitrate utilisation? Are these responses also regulated at the transcriptional level and if so which transacting factors are involved? What is the function of the individual <em>gsd</em> A-boxes in <em>A. niger</em> and <em>A. nidulans</em> ? Why do the regions containing the <em>gsd</em> A-boxes differ functionally, despite the obvious homology? Does titration of NIRA rely occur in <em>A. nidulans</em><em>gsd</em> A cotransformants? In any case these questions demonstrate the necessity for further research!