INSIGHTS INTO THE EVOLUTION OF EXPANSIN FUNCTION USING IN VITRO, IN VIVO, AND PHYLOGENETIC METHODOLOGIES
Open Access
- Author:
- Hepler, Nathan Kent
- Graduate Program:
- Plant Biology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- September 04, 2019
- Committee Members:
- Daniel J Cosgrove, Dissertation Advisor/Co-Advisor
Daniel J Cosgrove, Committee Chair/Co-Chair
Michael Axtell, Committee Member
Claude Walker Depamphilis, Committee Member
Yinong Yang, Outside Member
Teh-Hui Kao, Program Head/Chair - Keywords:
- expansin
plant cell wall
gene family evolution
Spirodela polyrhiza
Zostera marina
Utricularia gibba
CRISPR/Cas9
root hair
in vitro evolution - Abstract:
- A key aspect of plant cell growth is a process known as stress relaxation, which is mediated in part by proteins known as expansins. Expansins comprise a gene superfamily which can be divided into four families: EXPA, EXPB, EXLA and EXLB. All expansins share a two domain structure, where Domain 1 (N-terminus) has homology with family 45 glycoside hydrolases, while Domain 2 (C-terminus) is a carbohydrate binding module (CBM63). In the case of EXPAs and EXPBs, which are ubiquitous among land plants, evidence for cell wall-loosening (creep) has been observed, although this activity is not lytic. EXLAs and EXLBs have been found only in more derived plant lineages, specifically spermatophytes (‘seed plants’), and while activity has not been demonstrated for either, their structural homology with other expansins suggests they likely contain wall-loosening activity as well. More recently, xenologous expansins have been identified in various bacteria and fungi, the result of horizontal gene transfer. Unlike their plant homologs, microbial expansins can be produced using heterologous expression. Studies conducted using an expansin from Bacillus subtilis (BsEXLX1) elucidated key residues in both domains which contribute to function and polysaccharide binding. Specifically, deeply conserved aromatic residues (W125, W126, Y157) and K119 on the surface of Domain 2 are responsible for cellulose binding, and when mutated result in non-functional protein. In addition, basic residues on the ‘back face’ of Domain 2 (opposite the cellulose-binding region) mediate electrostatic binding to acidic polysaccharides, an interaction which reduces wall-loosening activity, apparently by competing with cellulose interactions. The interaction between basic residues and acidic polysaccharides has been termed ‘nonproductive binding’. Despite the insights gained using BsEXLX1 and other proteins, microbial expansins display low sequence similarity to plant homologs and exhibit relatively weak activities. One possibility for this reduced activity is a lower affinity for cellulose, a hypothesis which we tested using in vitro evolution. Through random selection of BsEXLX1 mutants which displayed stronger cellulose binding, we identified a general trend between enhanced cellulose affinity and increased protein activity, as assessed by the weakening of filter paper and irreversible extension of plant tissue. In addition, we also identified a mutation (E191K), which while increasing cellulose affinity, also resulted in enhanced nonproductive binding, ultimately abolishing creep activity. Subsequent experiments revealed E191K retains function, and the absence of activity under normal test conditions is interpreted as an issue with protein navigation within the cell wall. Detailed phylogenetic analyses have further subdivided the four plant expansin gene families into 17 orthologous clades (12 EXPA, 2 EXPB, 1 EXLA, 2 EXLB) shared among monocots and eudicots. Of the 17 clades, 16 have also been identified in Amborella trichopoda, and at least a few are reported in gymnosperms. Expansin superfamily size tends to be large in plant species, maintained predominantly through whole genome duplications. Currently, the smallest expansin superfamily is reported for the lycophyte Selaginella moellendorffii (15 EXPA and 2 EXPB), with the relatively small size attributed to the lack of any lineage-specific whole genome duplications. Exploration of the Spirodela polyrhiza (greater duckweed) genome, revealed an even smaller expansin superfamily, totaling 14 members (10 EXPA, 3 EXPB, 1 EXLA and 0 EXLB). A similar but less extensive reduction was seen for the marine plant Zostera marina (eelgrass), a close relative of duckweed. As opposed to Selaginella, whose expansin superfamily is small due to an absence of gene birth events, Spirodela and Zostera have reduced superfamilies due to gene loss, as both species have evidence for lineage-specific genome duplications. Another aquatic plant, the carnivorous eudicot Utricularia gibba (bladderwort), has also lost numerous expansin clades indicating expansin loss is a common occurrence during the adaptation to an aquatic environment. In at least one case, clade loss (EXPA-X) can be correlated with the loss of a specific phenotype (root hairs). Furthermore, several Utricularia expansins fail to branch among any of the 17 orthologous clades, which may be related to the derived trap structures used by bladderworts to capture prey. Plant expansins have been notoriously difficult to study, primarily due to a recalcitrance for recombinant expression and extensive functional redundancy across the superfamily. In order to address this issue, we developed an in vivo system for exploring expansin function by generating an EXPA-X loss-of-function mutant (atexpa7/atexpa18) in Arabidopsis thaliana (Col-0). atexpa7/atexpa18 mutants fail to form elongated root hairs, a phenotype which can be complemented through reintroduction of either paralog. Interestingly, root hair initiation was unaffected in the EXPA-X mutants. Subsequent complementation assays using paralogous and orthologous expansins revealed a deeply conserved biochemical function for EXPAs. However, no expansins from any of the other three families (EXPB, EXLA, EXLB) were able to restore root hair elongation in atexpa7/atexpa18, indicating EXPA function is distinct from the remaining three families. Whether this is a true difference in biochemical function, or simply differences in substrate specificity, is unknown and needs to be addressed. The development of this system will allow for further analysis of plant expansins via targeted mutagenesis experiments, as well as exploration of functional conservation among EXPAs throughout the plant kingdom. Our results indicate EXPA function evolved as early as the lycophytes, but how far this functional conserved extends into more basal plant species, such as bryophytes and green algae, requires additional testing.