An enzyme with mannosyl glycoprotein endo-N-acetyl-beta-D-glucosaminidase (ENGase)-type activity was partially purified from the extracellular medium of the mould Hypocrea jecorina (Trichoderma reesei). Internal peptides were generated and used to identify the gene in the T. reesei genome. The active enzyme is processed both at the N- and at the C-terminus. High-mannose-type glycoproteins are good substrates, whereas complex-type glycans are not hydrolysed. The enzyme represents the first fungal member of glycoside hydrolase family 18 with ENGase-type activity. Bacterial ENGases and the fungal chitinases belonging to the same family show very low homology with Endo T. Database searches identify several highly homologous genes in fungi and the activity is also found within other Trichoderma species. This ENGase activity, not coregulated with cellulase production, could be responsible for the extensive N-deglycosylation observed for several T. reesei cellulases.

SUMMARYFor genetic transformation of plants, floral dip with Agrobacterium often results in integration of multiple T-DNA copies at a single locus and frequently in low and unstable transgene expression. To obtain efficient single-copy T-DNA transformants, two CRE/loxP recombinase-based simplifying strategies for complex T-DNA loci were compared. A T-DNA vector with oppositely oriented loxP sites was transformed into CRE-expressing and wild-type control Arabidopsis thaliana plants. Of the primary CRE-expressing transformants, 55% harboured a single copy of the introduced T-DNA, but only 15% in the wild-type plants. However, 73% of the single-copy transformants in the CRE background showed continuous somatic inversion of the DNA segment between the two loxP sites. To avoid inversion of the loxP-flanked T-DNA segment, two T-DNA vectors harbouring only one loxP site were investigated for their suitability for CRE/loxP recombinase-mediated resolution upon floral-dip transformation into CRE-expressing plants. On average, 70% of the transformants in the CRE background were single-copy transformants, whereas the single-copy T-DNA frequency was only 11% for both vectors in the wild-type background. Both resolution strategies yielded mostly Cre transformants in which the 35S-driven transgene expression was stable and uniform in the progeny and remarkably, also in Cre transformants with multiple T-DNA copies. Therefore, a role is proposed for the CRE recombinase in preventing inverted T-DNA repeat formation or modifying the locus chromatin structure, resulting in a reduced sensitivity for silencing.