Hemicellulose biosynthesis and degradation in tobacco cell walls

Natural fibres have a wide range of technological applications, such as in paper and textile industries. The basic properties and the quality of plant fibres are determined by the composition of the plant cell wall. Characteristic for fibres are thick secondary cell walls, which consist of cellulose microfibrils in a matrix of lignin and hemicelluloses. The major hemicellulose component of the dicot secondary wall is xylan. This component can establish tight interactions with cellulose and covalent bonds with lignin and is thereby involved in strengthening of the wall. Since high amounts of hemicelluloses as well as lignins have a negative effect on the industrial processing of fibres, the development of plants with altered cell wall composition is of great value. This thesis focuses on modification of the hemicellulose (in particular xylan) content and thereby also the attachment of lignin in the cell wall.Different approaches to generate transgenically modified plants with altered xylan composition have been examined. (i) Decreasing the expression levels of putative xylan synthases, responsible for the polymerisation of the xylan backbone. (ii) Degradation of already deposited xylan polymers by the introduction of xylan degrading enzymes (xylanases). In our research, tobacco ( Nicotiana tabacum ) is used as a model species, since it is widely used in fundamental cell wall research. Once interesting results are obtained in tobacco, technologies can be transferred into economically more important species, such as flax ( Linum usitatissimum ) and poplar ( Populus spp). At the start of the work described in this thesis, genes involved in hemicellulose backbone biosynthesis were not identified yet. This favoured the use of a set of candidate genes, the cellulose synthase-like genes ( Csl s), which are generally suggested to be involved in the biosynthesis of the non-cellulosic cell wall polysaccharides. Plants contain at least eight families of Csl s ( CslA though H ), which all show the conserved motifs common to polymerizing β-glycosyltransferases. In order to investigate the function of six Csl families ( CslA , B , C , D , E and G )we introduced Inverted Repeat (IR) constructs, based on potato ( Solanum tuberosum ) Csl cDNA sequences into tobacco. If one of the Csl family members is coding for a xylan synthase, this strategy consequently provides us with transgenic tobacco plants with altered xylan levels.Sugar compositional analysis of cell wall material isolated from in vitro grown IR transformants revealed a reduction of xylose exclusively in the CslG transformants. These data led us to speculate about the involvement of CslG in xylan biosynthesis. Microscopic analysis of stem samples from mature CslG transformants revealed some very local effects like the multiplication of ray cells, enlargement of the size of vessels and the occurrence of abnormal fibres with thin walls and unusually high starch content. This result could indicate the tissue- or cell-type specificity of the CslG gene. Sugar compositional analyses of mature CslG plants were rather ambiguous and a second series of in vitro grown CslG transformants did not confirm the decrease of xylose that was observed during the initial screening. Additionally, the absence of tobacco CslG sequence information did not allow us to link the reduced xylose levels with altered CslG mRNA expression levels. Therefore, it required searching for tobacco CslG genes.We report the isolation of full length tobacco CslE and CslG cDNA sequences from a cDNA library, constructed from mRNAs isolated from a xylogenic suspension cell culture of tobacco. The proteins encoded by the NtCslE and NtCslG cDNA clones, contain eight putative transmembrane domains, alternating conserved and variable domains, and the processive glycosyltransferase signature. Southern blot analysis revealed that the tobacco CslEand CslG gene families consist of two to four and at least three genes, respectively. Gene expression studies in wild type tobacco tissues showed that CslE expression was highest in tissues associated with secondary wall biosynthesis, whereas CslG mRNA levels were highest in tissues undergoing primary wall formation. An involvement for CslG in xylan biosynthesis in tobacco, as indicated by the heterologous IR approach, is not supported by the gene expression analysis.Apart from modifying the xylan content by down regulation of putative xylan backbone synthesizing enzymes, a second strategy to generate transformants with altered xylan composition was applied. This strategy involved the specific degradation of the xylan polymers, which are already deposited in the cell wall, by the introduction of a xylanase. We describe the heterologous expression of AtXyn2 , a modular enzyme from Arabidopsis thaliana , consisting of four contiguous CBM22 modules at the N-terminus, joined to a glycoside hydrolase Family 10 (GH10) catalytic domain, into tobacco. The full length AtXyn2 as well as a derivative, comprising the catalytic GH10 domain lacking the four CBMs, was ectopically expressed in tobacco.An N-terminal signal peptide and RGS×(HIS) 6 -tag mediatedtranslocation into the cell wall and detection of the protein. Although both transgenes were successfully expressed in tobacco, presence of the corresponding proteins could not be demonstrated. We speculate that this could be caused by removal of the RGS×(HIS) 6 -tag due to either post-translational processing of AtXyn2 in tobacco or due to cleavage of the epitope-tag by a tobacco protease. Our data showed that tobacco transformants, expressing the complete AtXyn2 protein, possessed a significantly higher wheat xylan and HE-cellulose degrading activity than wild type tobacco. This result indicated that AtXyn2 displays a dual hydrolytic activity towards specific β-1,4-linked xylans and glucans. Since the catalytic activity was inversely proportional to the AtXyn2 transcript level, we speculate that high amounts of AtXyn2, or released oligosaccharides, could result in a negative feedback mechanism and subsequent inactivation or degradation of the enzyme. The cellulose degrading activity of AtXyn2 appeared to be even higher than the xylan degrading activity. This raises the question whether AtXyn2 should be considered as a cellulase instead of a xylanase. Transgenic tobacco plants expressing the truncated AtXyn2 protein did not show an altered hydrolytic activity against any of the substrates tested. It can therefore be concluded that the presence of the CBM22 domains adjacent to the catalytic module seems to play an essential role in the enzyme activity.Based on our results and additional information found in literature, we propose a model, which might explain the regulation of AtXyn2 activity in the plant cell wall. This model suggests that the CBM22 domains are responsible for immobilization of the enzyme in proximity of the substrate and for correct folding and stability of the protein. The enzyme activity is in turn suggested to be regulated by post-translational cleavage by an endoprotease.   The thesis is concluded with a general discussion on the outcome of the different approaches to generate tobacco transformants with reduced xylan content. Although both approaches initially seemed to be very promising, we were not successful in altering the xylan level in the cell wall. The drawbacks related to the down regulation of Csl candidate genes are discussed. Alternative approaches are therefore essential to identify xylan synthases and to modulate the xylan content in the cell wall. Various strategies to identify and characterise genes involved in hemicellulose biosynthesis are discussed. The heterologous expression of AtXyn2 was expected to result in a reduction of the xylan level in the cell wall. However, the dual substrate activity and the regulation of AtXyn2 activity, make this enzyme less suitable for reduction of the xylan content. Heterologous expression of well-studied microbial xylanases could be a suitable alternative to degrade xylan polymers in planta