Functional analysis of the Arabidopsis thaliana AtEP3 endochitinase

Chitinases are enzymes that are capable of catalyzing the hydrolysis of chitin, a homopolymer of N-acetylglucosamine. Chitin is the main constituent of the exoskeleton of insects, of crustacean shells and of the cell wall of many fungi but is absent in plants. This led to the commonly accepted hypothesis that plant chitinases are involved in defense against pathogens with chitin in their cell wall such as certain classes of fungi. Yet their role is not restricted to responding to pathogen attacks since plant chitinases are also induced by various types of stress, for instance after treatment with heavy metals or after UV irradiation. Chitinases can also be induced by plant hormones and they have been associated with a number of developmental processes, most notably in embryogenesis and during pollination. In addition, some chitinases may play a role in defense as well as in development, depending on their expression at particular stages in the plant life cycle. Plant chitinases belong to a relatively large protein family, which has hampered attempts to gain a better understanding of their role. A detailed study of individual chitinases is a prerequisite to unravel their precise role as well as to determine the function the different members of the five classes in which plant chitinases are subdivided.

In this thesis, we have addressed the role of one particular chitinase, AtEP3, in the model plant Arabidopsis thaliana . The work presented illustrates some of the difficulties inherent to the study of individual genes that belong to fairly large gene families. Chapter 1 gives a detailed overview of all chitinase genes present in the Arabidopsis genome. The genomic distribution and the sequences of these genes revealed interesting evolutionary relationships between the different classes. We discuss the possible significance of some of their sequence characteristics in light of their predicted role and propose a number of functions based on chitinases studied in other plants.

In Chapter 2, we present an elaborate analysis of the expression pattern of the AtEP3 gene. The expression pattern of the AtEP3 chitinase gene suggested possible functions in somatic embryo development, pollen maturation and/or germination, pollen tube growth, seed germination and root hair growth. All of these aspects have been looked at when searching for morphological aberrations (Chapters 5 and 6). The analogy between the expression of the gene found in pollen and in embryogenic cultures suggested there may be a correlation between gametogenesis and embryogenic cell formation. This notion was further taken into account by a study of GUS markers for specific cells of the female gametophyte and for the male gametophyte (Chapter 3). This work clearly indicated that a number of genetic programs specific for both gametophytes in planta are reproduced in tissue culture and that they are regulated in a spatially and temporally manner.

In Chapter 4, we were confronted with some of the pitfalls of reverse genetics. We performed a molecular and phenotypic analysis of several mutant plants in which the AtEP3 gene had been disrupted. However, we did not succeed in identifying any phenotype that could be directly linked to the absence of the AtEP3 chitinase. This was mainly due to the genetic instability of the material we studied, combined with the growth conditions in which we performed our analysis. It became clear from this work that small errors introduced while generating the available mutant plant collections can prevent the recovery of the desired individual mutant plants. In addition, the number of elements inserted for mutagenesis greatly influences the ease by which the phenotypic analysis of the mutant plants can be performed since multiple insertions can lead to several unlinked phenotypes.

In Chapter 5, we describe several transgenic lines in which the expression levels of the AtEP3 gene had been manipulated. An increase in AtEP3 expression did not result in any visible change in plant morphology, nor in embryogenic potential in vitro . However, a reduction of AtEP3 expression to 10% the level of the wild-type resulted in a defect in root hair morphology. Similarly, complete knockout of the gene produced a root hair phenotype in a mutant plant now renamed ep3-1 (Chapter 6). Both phenotypes suggest a role for AtEP3 in root hair formation. Complete absence of AtEP3 mRNA also gave rise to the direct germination of fresh seeds without prior stratification, indicating that the chitinase could also be involved in the maintenance of seed dormancy. An additional defect was a strong reduction of pollen development in vitro . Surprisingly, the reduced pollen germination phenotype in the ep3-1 mutant could be compensated by stigma exudates in vitro . This makes it very unlikely that such a phenotype could ever be observed in planta . The absence of AtEP3 in pollen is most likely compensated by the presence of other chitinases in the stigma. This is probably the case in embryogenic cultures as well, where the absence of a single chitinase might not be sufficient to hinder embryo development. Previous work done in carrot is in line with these observations.

We have previously proposed the involvement of arabinogalactan proteins (AGPs) as a possible substrate for the AtEP3 chitinase. We base this hypothesis on the findings that carrot EP3 chitinases can cleave specific AGPs and that as a result the promotive effect of these now "cleaved" AGPs on somatic embryo development is enhanced. Second, AGPs are often found at the same location as chitinases and finally AGPs have been shown by others to be involved in pollen and root development as well as in seed germination. Taken together these observations suggest a role for the AtEP3 chitinase in intercellular communication through N-acetylglucosamine-containing signal molecules. The work presented in this thesis provides the groundwork that is essential to address the role of plant chitinases by molecular, genetic and biochemical means.