Lactose metabolism and cellulase production in Hypocrea jecorina: the gal7 gene, encoding galactose-1-phosphate uridylyltransferase, is essential for growth on galactose but not for cellulase induction.

Lactose is at present the only soluble carbon source which can be used economically for the production by Hypocrea jecorina (= Trichoderma reesei) of cellulases or heterologous proteins under the control of cellulase expression signals. However, the mechanism by which lactose triggers the formation of cellulases is unknown. To enhance our understanding of lactose metabolism and its relationship to cellulase formation, we have cloned and characterized the gal7 gene (for galactose-1-phosphate uridylyltransferase) of H. jecorina. The gene encodes a polypeptide of 43.8 kDa, the sequence of which exhibits a moderate level of identity (about 50%) to that of the Gal7 proteins of Saccharomyces cerevisiae and Kluyveromyces lactis, and contains an active-site signature typical for galactose-1-phosphate uridylyltransferase family 1. H. jecorina gal7 is not clustered with other genes of galactose metabolism. A single 1.7-kb transcript is synthesized constitutively during the rapid growth phase and accumulated to twice this level during incubation in the presence of D-galactose and L-arabinose and the corresponding polyols (dulcitol, arabitol). A gal7 deletion mutant, constructed by replacing the gal7 reading frame by the H. jecorina pyr4 gene, was unable to grow on D-galactose between pH 4.5 and 7.5, thus proving that in H. jecorina gal7 is essential for metabolism of D-galactose, whereas the growth rate of the mutant on lactose was only reduced by about 50%. The rate of formation of cellobiohydrolase Cel7A and the abundance of the corresponding (cbh1) transcript during growth on lactose was only slightly lower in the absence of gal7, but a significant delay in decay of the cbh1 transcript was noted during later stages of growth. The results suggest that H. jecorina uses only the Leloir pathway for metabolism of D-galactose and lactose. Furthermore, we conclude that metabolism of lactose past the galactose-1-phosphate step is not essential for cellulase formation.

Identification and characterization of a family of secretion-related small GTPase-encoding genes from the filamentous fungus Aspergillus niger: a putative SEC4 homologue is not essential for growth.

DNA fragments containing genetic information for five secretion-related small GTPases of Aspergillus niger (srgA-E) were isolated and identified as members of different Rab/Ypt subfamilies. This isolation and the search for similar sequences in fungal genomic and EST databases showed that, in contrast to Saccharomyces cerevisiae, filamentous fungi also possess homologues of mammalian Rab2 GTPases. Multiple transcripts with unusually long 5' and 3' untranslated regions were found for all srg genes. Their level of expression was independent of the type of carbon source used for growth. Although the transcripts of srgA and srgB were abundant to the same extent throughout the cultivation, that of the other genes peaked during the early growth phase and then declined. Two genes, srgA and srgB, were characterized further. The protein encoded by srgA exhibited relatively low identity (58%) to its closest S. cerevisiae homologue SEC4, whereas the protein encoded by srgB showed 73% identity with S. cerevisiae YPT1. In contrast to other SEC4 homologues, srgA was unable to complement an S. cerevisiae sec4 mutant, and its disruption was not lethal in A. niger. SrgA mutants displayed a twofold increase in their hyphal diameter, unusual apical branching and strongly reduced protein secretion during growth on glucose.

Glucose does not activate the plasma-membrane-bound H+-ATPase but affects pmaA transcript abundance in Aspergillus nidulans.

The addition of glucose to starved cells of Aspergillus nidulans increased the abundance of the pmaA transcript only transiently (15 min) and to a very low degree (1.3-fold), but strongly decreased its abundance during further incubation. This down-regulation was CreA (carbon catabolite repressor protein)-dependent. Glucose failed to stimulate the plasma membrane (PM)-ATPase activity of A. nidulans, whereas under the same experimental conditions the activity of the enzyme from Saccharomyces cerevisiae was enhanced four-fold within 5-10 min following glucose addition. Glucose stimulated the PM-ATPase of Neurospora crassa only 1.3-fold. Sequence comparison of the C-terminal end of the PM-ATPase from S. cerevisiae, N. crassa, A. nidulans, Fusarium sporotrichoides and Penicillium simplicissimum showed that the two regulatory sites necessary for glucose stimulation in S. cerevisiae are conserved in N. crassa and F. sporotrichoides but not in A. nidulans and P. simplicissimum, and their presence therefore does not correlate with glucose stimulation. We conclude that, in contrast to S. cerevisiae, which has become a paradigm of fungal glucose metabolism, glucose does not up-regulate the activity of the plasma membrane ATPase in the filamentous fungi examined.

Role of four major cellulases in triggering of cellulase gene expression by cellulose in Trichoderma reesei.

The relative contributions of four major cellulases of Trichoderma reesei (1,4-beta-D-glucan cellobiohydrolase I [CBH I], CBH II, endo-1,4-beta-D-glucanase I [EG I], and EG II) to the generation of the cellulase inducer from cellulose were studied with isogenic strains in which the corresponding genes (cbh1, cbh2, egl1, and egl2) had been deleted by insertion of the Aspergillus nidulans amdS marker gene. During growth on lactose (a soluble carbon source provoking cellulase gene expression), these strains showed no significant alterations in their ability to express the respective other cellulase genes, with the exception of the strain containing delta cbh1, which exhibited an increased steady-state level of cbh2 mRNA. On crystalline cellulose as the only carbon source, however, significant differences were apparent: strains in which cbh2 and egl2, respectively, had been deleted showed no expression of the other cellulase genes, whereas strains carrying the cbh1 or egl1 deletion showed these transcripts. The delta cbh1-containing strain also showed enhanced cbh2 mRNA levels under these conditions. A strain in which both cbh1 and cbh2 had been deleted, however, was unable to initiate growth on cellulose. Addition of 2 mM sophorose, a putative inducer of cellulase gene expression, to such cultures induced the transcription of egl1 and egl2 and restored the ability to grow on cellulose. We conclude that CBH II and EG II are of major importance for the efficient formation of the inducer from cellulose in T. reesei and that removal of both cellobiohydrolases renders T. reesei unable to attack crystalline cellulose.

The bgl1 gene of Trichoderma reesei QM 9414 encodes an extracellular, cellulose-inducible beta-glucosidase involved in cellulase induction by sophorose.

We have investigated the effect of disruption of the bgl1-(beta-glucosidase l-encoding) gene of Trichoderma reesei on the formation of other beta-glucosidase activities and on the induction of cellulases. To this end the bgl1 locus was disrupted by insertion of the Aspergillus nidulans amdS (acetamidase-encoding) gene. The bgl1-disrupted strain did not produce the 75 kDa extracellular beta-glucosidase on cellulose or lactose, but still formed beta-glucosidase activity on glucose, cellobiose, xylan or beta-1,3-glucan, suggesting that the enzyme(s) exhibiting this beta-glucosidase activity is (are) not encoded by bgl1. The cellulase-inducer sophorose induced the bgl1-encoded beta-glucosidase, whereas the remaining beta-glucosidase activity was induced by methyl-beta-D-glucoside. The bgl1-gene product was mainly secreted into the medium, whereas the other beta-glucosidase activity was mainly associated with the cells. A bgl1-multicopy strain formed higher amounts of cellulases than the parent strain. Nonsaturating concentrations of sophorose efficiently induced cellobiohydrolase l formation in the bgl1-multicopy strain, but less efficiently in the bgl1-disrupted strain. The multicopy strain and the parent strain were comparably efficient at saturating sophorose concentrations. The beta-glucosidase inhibitor nojirimycin strongly inhibited induction in all strains. These data suggest that the bgl1-encoded beta-glucosidase is not identical to the plasma-membrane-bound, constitutive, methyl-beta-glucoside inducible beta-glucosidase, but represents an extracellular cellulose-induced enzyme. Both enzymes contribute to rapid induction of cellulases by modifying the inducer sophorose.