Selaginella moellendorffiihas a reduced and highly conserved expansin superfamily with genes more closely related to angiosperms than to bryophytes
© Carey et al; licensee BioMed Central Ltd. 2013
Received: 6 August 2012
Accepted: 22 December 2012
Published: 3 January 2013
Expansins are plant cell wall loosening proteins encoded by a large superfamily of genes, consisting of four families named EXPA, EXPB, EXLA, and EXLB. The evolution of the expansin superfamily is well understood in angiosperms, thanks to synteny-based evolutionary studies of the gene superfamily in Arabidopsis, rice, and Populus. Analysis of the expansin superfamily in the moss Physcomitrella patens revealed a superfamily without EXLA or EXLB genes that has evolved considerably and independently of angiosperm expansins. The sequencing of the Selaginella moellendorffii genome has allowed us to extend these analyses into an early diverging vascular plant.
The expansin superfamily in Selaginella moellendorffii has now been assembled from genomic scaffolds. A smaller (and less diverse) superfamily is revealed, consistent with studies of other gene families in Selaginella. Selaginella has an expansin superfamily, which, like Physcomitrella, lacks EXLA or EXLB genes, but does contain two EXPA genes that are related to a particular Arabidopsis-rice clade involved in root hair development.
From sequence-based phylogenetic analysis, most Selaginella expansins lie outside the Arabidopsis-rice clades, leading us to estimate the minimum number of expansins present in the last common ancestor of Selaginella and angiosperms at 2 EXPA genes and 1 EXPB gene. These results confirm Selaginella as an important intermediary between bryophytes and angiosperms.
KeywordsExpansin Selaginella moellendorffii Cell wall loosening Gene family evolution Plant phylogenetics
Expansins are plant proteins discovered via their involvement in pH-dependent wall extension . In land plants these proteins are encoded by a large superfamily of genes. Expansins act non-enzymatically in the cell wall to disrupt the interactions between cellulose microfibrils and hemicelluloses [2, 3]. This is thought to contribute to turgor-driven cell wall expansion during cell growth [3–5]. The original proteins characterized in this way are now known as the EXPA family of expansins. A group of grass pollen allergens was later discovered that was also capable of causing cell wall creep and became the founding members of the group now known as the EXPB family of expansins . The expansin superfamily in plants has four constituent families named EXPA, EXPB, EXLA, and EXLB. While members of the EXPA and EXPB families have been shown to have characteristic expansin activity, the functions of the EXLA and EXLB (expansin-like) families, discovered via their similarity to other expansin sequences, have not yet been characterized.
Expansins are usually expressed in a tissue -specific pattern and are involved in many processes where cell wall loosening is crucial, such as growth [7–9], fruit ripening , pollen tube penetration of the stigma , root hair elongation , and others . The proteins encoded by these genes share certain characteristic features including a signal peptide for secretion and a two-domain structure [14, 15]. Expansins have been identified in all land plants that have been examined and several related but highly divergent sequences exist in unicellular green algae .
Previous work has demonstrated that expansin family sizes remain relatively constant among species even when the individual genes have a distinct evolutionary history [17, 18]. This suggests that there is some selective advantage to having a relatively large superfamily of expansins. The evolutionary relationships between the members of this large superfamily are complicated and have proved difficult to elucidate , but understanding of the superfamily in angiosperms (specifically Arabidopsis, rice, and Populus) has improved through the use of genomic history to complement phylogenetic analysis [18, 20]. The analysis by Sampedro et al. indicated 17 orthologous expansin gene clades between Arabidopsis and rice, and revealed a dynamic gene superfamily with large numbers of gene births (due to polyploidy and segmental duplications) and deaths shaping the distribution of sequences within these clades.
An additional study elucidated the composition of the expansin superfamily in Physcomitrella patens and compared these sequences with angiosperm expansins . Although these Physcomitrella expansins do not show a clear relationship to specific Arabidopsis-rice clades defined by the work of Sampedro et al., they do show a gene superfamily of similar size and complexity arising from a minimum of 2 EXPA and 1 EXPB genes in the common ancestor of Physcomitrella and angiosperms . The genome sequencing of Selaginella moellendorffii, an early diverging vascular plant  offers an opportunity to extend our understanding of this large gene superfamily into the lycophytes, a key intermediate between bryophytes and seed plants. Selaginella, a vascular plant with true roots and shoots has a far greater morphological similarity to angiosperm species than mosses like Physcomitrella. Thus, the likelihood of relating expansins of an early diverging lineage to the expansin genes of angiosperms seems greater in a study of lycophytes than bryophytes.
Expansin superfamily in Selaginella moellendorffii
The expansin sequences revealed via searches of the Selaginella genome comprise a superfamily whose composition is similar to what has been observed in angiosperm genomes with a few notable exceptions.
Expansin and other selected plant gene family compositions
22, 136 scaffolds
Phylogenetic analysis of Selaginellaexpansins
Two Selaginella EXPA genes, SmEXPA5 [XM_002961012.1] and SmEXPA6 [XM_002980135.1], appear to be a sister group to the Arabidopsis-rice clade EXPA-X (AtEXPA7 [NM_101127.3] and OsEXPA30 [AC092697.6]). This grouping is corroborated by the two alternate tree-building methods (see Additional files 4 and 5). The remaining Selaginella EXPA sequences can be divided into five groups that have been named A-E (Figures 1, 2 and Additional files 4 and 5).
Group A contains Selaginella sequences SmEXPA1 [XM_002974852.1], SmEXPA2 [XM_002981819.1], SmEXPA3 [XM_002974112.1], SmEXPA4 [XM_002988923.1], and SmEXPA11 [XM_002973901.1]. This group of Selaginella expansins, while not grouping consistently with a specific Arabidopsis-rice clade, does have the smallest pairwise distances to an angiosperm expansin of any Selaginella gene group (Additional file 6).
These low distances are always to the members of Arabidopsis-rice clades I-IV, the most conserved of all Arabidopsis-rice clades (indicating that they are under strong purifying selection). This group also branches (although with weak support) on all trees with Physcomitrella group D (Figure 2 and Additional files 7 and 8). In previous work it was observed that this Physcomitrella group branched with the members of Arabidopsis-rice clades EXPA I-III in the Bayesian trees . Although it is still very poorly resolved phylogenetically, it is possible that Selaginella group A, Physcomitrella group D, and angiosperm clades EXPA I-III are orthologous groups based on the low distances and phylogenetic results described here. It is certain, however, that the genes of Selaginella group A are more closely related to angiosperm EXPA genes and Physcomitrella groups D-F than to Physcomitrella groups A-C.
Group B consists of five Selaginella EXPA genes (SmEXPA7 [XM_002994463.1], SmEXPA8 [XM_002968976.1], SmEXPA13 [XM_002980028.1], SmEXPA14 [XM_002990586.1], and SmEXPA15 [XM_002994463.1]). While it is not possible to state with any confidence that this group of Selaginella expansins is a sister to a specific Arabidopsis-rice clade, it does seem clear that these genes group more closely with angiosperm expansins and not, for example, with the genes of Physcomitrella groups A-C (Figure 2 and Additional files 7 and 8).
The placement of SmEXPA9 [XM_002963656.1], SmEXPA10 [XM_002981332.1], and SmEXPA12 [XM_002966496.1] is poorly resolved in all phylogenetic trees. They do not clearly branch with any known rice, Arabidopsis or Populus clade. SmEXPA10 [XM_002981332.1] does consistently branch (Figure 2 and Additional files 7 and 8) with AtEXPA12 [XM_002882892.1], but with uniformly poor support. These Selaginella expansins also do not group consistently with a known pine specific group  or with each other. SmEXPA9 [XM_002963656.1], SmEXPA10 [XM_002981332.1], and SmEXPA12 [XM_002966496.1] do not consistently branch with any known Selaginella or Physcomitrella expansin either, regardless of the tree-building method employed (Figures 1, 2 and Additional files 4, 5, 7 and 8).
Thus, Selaginella EXPA sequences can be divided into 6 groups ranging from 1 to 5 sequences. One of these is clearly orthologous to a clade seen in Arabidopsis, rice, and Populus (EXPA-X). The five remaining groups (A-E) seem to be more closely related to angiosperm expansins than to bryophyte specific groups, but do not group consistently with any specific Arabidopsis-rice clade.
The two Selaginella EXPB genes (SmEXPB1 [XM_002970263.1] and SmEXPB2 [XM_002983273.1]) branch as a sister group to the representatives of Arabidopsis-rice EXPB clade II (AtEXPB3 [NM_118965.3] and OsEXPB16 [AK240809.1]) in the Bayesian and Neighbor-joining (but not the maximum parsimony) trees with relatively good support (Figure 3 and Additional files 10 and 11).
As was noted for the Physcomitrella expansin superfamily , the Selaginella expansin superfamily seems to be evolving quite independently and yet a large multigene family is maintained. This may indicate that the size of the expansin gene families is somehow critical, with the advantage of a large family becoming redundant at some maximum number.
Without including substantial gymnosperm and other intermediary expansin sequence (such as fern sequences), phylogenetic analyses comparing taxa as distantly related as Selaginella and angiosperms inevitably become inaccurate. Although there are some EST sequences available for Pinus taeda, including this limited set of gymnosperm expansins in these phylogenetic analyses does not help to resolve the placement of Selaginella EXPA groups (Additional files 12 and 13). Adding the very few fern expansins available in GenBank also does not improve these phylogenies (data not shown). As it becomes available, extensive gymnosperm and fern sequence will need to be included in these analyses in order to improve the reliability of the phylogenies. At present, there is substantial EST data available for loblolly pine, but no whole-genome data from any gymnosperm or fern. It should also be noted that even within the angiosperms, the difficulty in using traditional phylogenetic methods to elucidate relationships between members of the expansin families is well known . This is not surprising as the expansin superfamily shows evidence of rapid diversification with many gene births and deaths .
Notwithstanding these difficulties, sequence-based phylogenies still offer some insight into the evolutionary relationships between the expansin sequences of Selaginella and angiosperms, especially when used in the light of the well-supported classifications proposed in previous work [18, 20]. The classification of Sampedro et al. will be used here to discuss the relationship of Selaginella expansins to their angiosperm counterparts.
Distances of Selaginellaexpansins to angiosperm expansins
Intron-based analysis of Selaginellaexpansins
In addition to having conserved amino acid sequence, expansins have been shown to have a fairly conserved intron pattern (see Additional file 15). Sampedro et al. hypothesized the ancestral intron patterns for the angiosperm expansin families. Based on what was known of the intron patterns seen in Arabidopsis and rice, the intron pattern for ancestral EXPA and EXPB sequences was estimated using parsimony as a basis for determining the pattern (the number of gains/losses was maximized). In this way, it was hypothesized that the ancestral EXPA intron pattern likely consisted of introns ‘A’ and ‘B.’ The likely ancestral EXPB intron pattern was hypothesized to consist of introns ‘A,’ ‘B,’ ‘C’ and ‘F.’ The intron patterns of Selaginella lend support to the hypothesized ancestral EXPA and EXPB patterns  and indicates these patterns pre-date the divergence of lycophytes and angiosperms.
The fifteen Selaginella EXPA genes all contain introns ‘A’ and ‘B,’ which obviously supports the idea of an ‘AB’ ancestral intron pattern. Four EXPAs also contain additional novel introns: both SmEXPA2 [XM_002981819.1, XM_002994136.1] haplotypes contain an intron in the 5′-untranslated region, SmEXPA11b [XM_002973901.1] contains an intron in the 5′-untranslated region and SmEXPA1b [XM_002988863.1] contains an intron in the 3′-untranslated region.
All three EXPB genes discovered in Selaginella have introns ‘A,’ ‘C,’ ‘B’ and ‘F.’ This data suggests that in the last common ancestor between Arabidopsis, rice and Selaginella, the intron pattern for EXPB genes may well have been ACBF, which is congruent with the findings of Sampedro et al.. This is in contrast to the more variable intron patterns seen in Physcomitrella, and is further evidence of Selaginella’s value as an intermediary taxon between bryophytes and angiosperms.
Selaginella expansins appear to have reduced intron sizes when compared to their Physcomitrella and Arabidopsis counterparts. For example, the average size of intron A in Selaginella EXPA genes is 85 bp while it is longer in both Physcomitrella (387 bp) and Arabidopsis (158 bp). Intron B shows a similar pattern, with Selaginella having the shortest introns (although Arabidopsis has longer average length than Physcomitrella for intron B). This reduction of intron size is consistent with previous observations that reduced genome size correlates with a decreased size of non-coding regions .
The sequencing of the Selaginella moellendorffii genome allows us to fill in some of the gaps in our knowledge of early land plant expansin evolution. Using phylogenetic analyses, it has been possible to predict some of the types of expansins found in the last common ancestor of Selaginella and Arabidopsis. The pattern of introns seen in Selaginella is also useful for determining the pattern of intron evolution in the EXPA and EXPB families, with Selaginella having a pattern consistent with previous predictions about expansin intron evolution.
As seen in Table 1 there are some differences between the compositions of the expansin superfamily in Selaginella compared to what is seen in angiosperms or Physcomitrella. One of the most obvious differences is the apparent lack of members of the EXLA and EXLB family in Selaginella and Physcomitrella. This likely indicates that these families arose after the divergence of Physcomitrella and Selaginella, as the presence of these genes was not detected by a tBLASTX search of the JGI Selaginella moellendorffii v1.0 genome. It is doubtful that these families are ancestral to all land plants, as they would have to have been lost in multiple independent lineages. It is more likely that EXLA and EXLB gene families arose after the divergence of lycophytes and bryophytes. The sequencing of basal vascular plants, ferns, and gymnosperms will help clarify this issue. It will be interesting to see at what point these gene families first appeared as more sequence becomes available.
We also see in Table 1 that the proportion of genes belonging to the EXPA and EXPB families in Selaginella is similar to what is found in Arabidopsis, Populus, and Physcomitrella but not in rice [18, 20]. The diversity of cell wall composition among land plants is likely important in shaping the diversity of the expansin superfamily, and it may be that the expansion of the EXPB family in rice is related to the unique cell walls of grasses [23, 24]. The overall size of the Selaginella superfamily is smaller than is seen in the other plants studied here. Reduced gene family size is not limited to expansins in Selaginella but has also been observed in the non cell wall-related gene family of major intrinsic proteins  and the cell wall-related gene families of callose synthase [26, 27] and xyloglucan endo-transglycosylase/hydrolase (XTH) . Current research also shows no evidence that Selaginella has undergone whole genome duplication or a polyploidy event , which would limit the number of expansin genes as compared to other plant species as polyploidy is known to be an important driving force in expansin evolution . This may mean that the smaller expansin superfamily of Selaginella has changed much more slowly and may represent a more “essential set” of expansins.
Phylogenetic analysis of the Selaginellaexpansin superfamily
From the phylogenetic trees for the Selaginella EXPA genes with selected EXPA sequences from rice, Arabidopsis, and Populus (Figure 1), we see that one group of Selaginella EXPA genes branches clearly as a sister group to the Arabidopsis-rice clade EXPA – X. SmEXPA5 [XM_002961012.1] and SmEXPA6 [XM_002980135.1] clearly branch sister to Arabidopsis-rice clade EXPA – X, a clade whose Arabidopsis genes have a well-characterized expression pattern  that is root hair specific. Selaginella does have root hairs,  and it would be a confirmation of the functional orthology of these genes if they were expressed there. It is possible that this particular type of expansin may have evolved from the need to regulate root hair development once these organs arose in land plant lineages (Selaginella has true roots while Physcomitrella does not).
The members of group A have the smallest pairwise distances of any group to the three most conserved Arabidopsis-rice clades (EXPA I – III). These Arabidopsis-rice clades along with Arabidopsis-rice clade EXPA IV are what were initially characterized as Subgroup A  and may function in vasculature tissue, specifically xylem . The members of Arabidopsis-rice clades I, II, III, and IV are consistently the genes with the smallest pairwise distances to Selaginella EXPA sequences (the exceptions being SmEXPA5 [XM_002961012.1] and SmEXPA6 [XM_002980135.1]). They also have the smallest within and between group mean distances when compared with the other Arabidopsis-rice clades. These data suggest that the members of these Arabidopsis-rice clades are under strong purifying selection. Despite this overall similarity to many of the basal land plant EXPA genes seen in Selaginella, the members of group A have noticeably smaller distances to the members of Arabidopsis-rice clades EXPA I – III (Additional file 6). If members of this Selaginella EXPA group were shown to be expressed in vascular tissue, it might indicate that these genes are orthologous to the members of Arabidopsis-rice clades I – IV and raise the possibility that these genes have developed a function important in xylem development in tracheophytes. Group A seems to also group consistently, with relatively good support, with Physcomitrella group D. This may support an association of group A and angiosperm clades EXPA 1 – III as Physcomitrella group D shows weak branching with Arabidopsis-rice clades EXPA I – III on some trees .
The remaining Selaginella EXPA groups are not clearly sister groups to any particular angiosperm clade or Physcomitrella grouping, but do seem to be more closely related to angiosperm expansins and Physcomitrella groups D – F than to Physcomitrella groups A – C (Figures 1, 2 and Additional files 4, 5, 7 and 8). When all Selaginella EXPA genes are constrained as a monophyletic group and a parsimony analysis is performed, maximum parsimony trees of length 1939 (38 steps longer than the tree in Additional file 4) are obtained. When all of the Selaginella EXPA genes except for SmEXPA5 [XM_002961012.1] and SmEXPA6 [XM_002980135.1] are constrained as a monophyletic group, maximum parsimony trees of length 1910 are obtained. This would seem to indicate that there are relationships amongst these Selaginella groups and angiosperm clades that phylogenetic analyses do not yet clearly resolve, and it also is consistent with the idea that SmEXPA5 [XM_002961012.1] and SmEXPA6 [XM_002980135.1] are sister to clade EXPA-X.
Thus, although it is likely an underestimation, we conclude that the last common ancestor of Selaginella and angiosperms had two EXPA genes, one that gave rise to SmEXPA5 [XM_002961012.1] and SmEXPA6 [XM_002980135.1] and one that gave rise to the rest of the Selaginella EXPA gene family.
The two Selaginella EXPB genes group with Arabidopsis-rice clade EXPB – II in Bayesian and Neighbor Joining trees, indicating at least one EXPB in the common ancestor of Selaginella and angiosperms that is more similar to the vegetative EXPBs of angiosperms than to Physcomitrella EXPBs.
Selaginellaexpansin distance analysis
Additional file 6 shows that nearly all Selaginella EXPA genes have their lowest pairwise distance to a member of Arabidopsis-rice clades I – IV, again potentially suggesting that they are under greater purifying selection. The genes of group A have particularly small distances to the members of these clades, perhaps suggesting some relationship that is not yet apparent in phylogenetic analyses. It is also interesting to note that nearly the only exceptions to this pattern are SmEXPA5 [XM_002961012.1] and SmEXPA6 [XM_002980135.1] whose smallest pairwise distances are to members of clade EXPA – X, the one with which they branch as a sister group to in phylogenetic analyses.
Selaginella EXPA and EXPB genes have surprisingly small average distances to their angiosperm counterparts (Figure 4a,b). These rather small evolutionary distances do not alleviate the difficulty of phylogenetic analysis mentioned previously, however.
Intron analysis of the Selaginellaexpansin superfamily
All Selaginella EXPA genes show an ‘AB’ intron pattern, with four haplotypes showing additional introns. Both haplotypes for SmEXPA2 [XM_002981819.1, XM_002994136.1] and one haplotype for SmEXPA11 (designated SmEXPA11b [XM_002973901.1]) contain an intron in the 5′ – untranslated region. The SmEXPA2 [XM_002981819.1] introns and SmEXPA11b [XM_002973901.1] intron are relatively the same length, located in the same area of the 5′ - UTR and are nearly a 45% match on the nucleotide level, so we’ve decided that they are probably the same. They have been designated novel intron prime, n’. Arabidopsis-rice EXPA clades I – II do contain an intron in the 5′ – UTR , so it is possible that n’ is that same intron. However, since none of the other Selaginella EXPA genes contain n’, that is not likely the case. More likely n’ is a novel intron that arose in a subset of Selaginella group A and has been lost in one SmEXPA11 [XM_002973901.1] haplotype. Also, one haplotype of SmEXPA1 (designated SmEXPA1a [XM_002974852.1]) contains an intron in the 3′ – untranslated region. This intron has been designated novel intron, n.
Both Selaginella EXPB genes show an ‘ACBF’ intron pattern, which is the ancestral intron pattern predicted in Sampedro et al. for these families based on a parsimony model of intron gain and loss in angiosperms. The ‘AB’ intron pattern seen in all the Selaginella EXPA genes is also the predicted ancestral intron pattern . These data therefore support this predicted ancestral intron pattern at least as far back as the last common ancestor of Selaginella and Arabidopsis.
Conservation of amino acid sequence
As was seen for the EXPA gene family in Physcomitrella, Selaginella also shows conservation at all normally conserved expansin amino acid residues. In contrast to the EXPB family in Physcomitrella, the Selaginella EXPB gene family also shows conservation at these sites. This would seem to imply that the biochemical function of Selaginella EXPA and EXPB genes is not altered from the biochemical function of these gene families in angiosperms. It is worth noting that recent work has demonstrated the importance of xyloglucan in both acid growth and expansin activity assays  and that lycophytes have a very different xyloglucan composition than eudicots, gymnosperms, and some ferns . It is possible that these differences in xyloglucan composition are not important for expansin function in lycophytes, or that there is some subtle systematic difference in lycophyte expansins that is not immediately obvious.
With the extensive analysis of rice, Arabidopsis, and Populus as a guide, the classification of Selaginella expansins into groups and the inference of the relationship of these groups to known orthologous groups in Arabidopsis and rice, and homologous groups of genes observed in Physcomitrella has been attempted. What is seen is an expansin superfamily in Selaginella that is somewhat more easily related than Physcomitrella expansins to the groups of expansin genes seen in higher plants. Indeed, Selaginella expansins seem to have much more in common with their Arabidopsis and rice counterparts than they do with Physcomitrella. Evidence indicates that some Selaginella genes are sister groups to Arabidopsis-rice clades. In addition, all Selaginella expansins seem to be more closely related to angiosperm expansins and Physcomitrella groups D – F than to the bryophyte – specific groups described previously . Thus a picture emerges of morphological similarity potentially reflecting expansin superfamily development, with morphologically similar plants having more similarities in their expansin families. This makes sense given the closer evolutionary relationship of morphologically similar plants and the importance of expansins in growth and developmental processes. The smaller and less diverse Selaginella expansin superfamily may prove useful as a vehicle for understanding the “essential set” of expansins needed for plant growth and development. As more and more plant species are sequenced in the genomics age, what are now mere outposts of data will be interconnected, hopefully with the result of elucidating the dynamic evolutionary past of gene superfamilies such as expansins.
Trace archive searches
Trace archives for Selaginella moellendorffii (1,814,554 traces on 10/08/2005) were searched using the “Cross-species Mega BLAST” on the NCBI Trace Archive Nucleotide BLAST website . All Arabidopsis, rice, and known Physcomitrella sequences were used as BLAST queries under default parameters. The traces identified by these searched were downloaded in .scr trace format for assembly into contigs. All Selaginella expansins isolated in this way were then used to search the archive. An additional tBLASTX search of the archives was done using EXLA and EXLB genes from Arabidopsis, rice, and pine as search queries (thanks to K. Wall).
Assembly of contigs
Trace files were assembled into contigs with the SeqMan application in the DNASTAR software package. The ends of the traces were trimmed on the ‘high’ quality setting (quality score = 16). The alignments were created with a minimum match percentage of 90% over 50 base pairs. Assembly was performed after the completion of all searches.
The Selaginella genes originally assembled from the trace archive were used to search the Selaginella moellendorffii v1.0 genome . A tBLASTX search was also conducted using all Arabidopsis, rice, and Physcomitrella expansin sequences. The traces identified by these searches were downloaded in .fasta format and cross checked to eliminate duplicate results. The Selaginella genome (both haplotypes) was analyzed using the resulting sequences to identify expansin genes. Sequences that did not encode genes were discarded. Sequences that correctly encoded expansin genes were downloaded in .fasta format, compared to previously isolated Selaginella expansins and named accordingly (see Additional file 16). All expansin annotations were inspected for intron patterns. Sequences were then trimmed for alignment.
Phylogenetic tree construction
Selaginella sequences (Additional file 1) were aligned with selected Arabidopsis, rice, and sometimes Physcomitrella sequences . Alignments were generated via the Clustal W function of the MegAlign application of the DNASTAR 9 software package with default alignment parameters (Gonnet Series protein weight matrix, gap penalty of 15, gap length penalty of 6.66, delay Divergent Seqs 30%). These alignments (Additional files 2, 3 and 15) were then used as the input to generate Bayesian, parsimony, and neighbor-joining phylogenies trees.
MrBayes version 3.1.2p [36, 37] was utilized using the POOCH software application  to generate the Bayesian trees (Jones amino acid model, gamma rates, 2 runs, 4 Markov chains – number of generations and burnin as indicated in figure legends) from an alignment trimmed from a conserved tryptophan following the signal peptide to a conserved phenylalanine at the carboxyl terminus of the expansin genes. MCMC convergence was assessed graphically using the AWTY web service . The consensus trees were then visualized using the Tree Graph 2 software application  and manually rooted.
Protein parsimony trees were made using the same alignment with the Phylogenetic Analysis Using Parsimony software package (PAUP* version 4.0) . Maximum parsimony trees were generated by a heuristic search with 100 random sequence additions. A bootstrap analysis with 500 replicates was then performed with 10 search replicates with random additions per bootstrap replicate. The Tree Graph 2 software application  was then used to visualize the consensus trees and manually root them. If the bootstrap consensus contained adequate information it is used in the figure. If many branches in the consensus tree were poorly resolved then one of the maximum parsimony trees was used with bootstrap values manually added to nodes with good support in the bootstrap consensus tree.
Neighbor-Joining trees were constructed using the MEGA Phylogeny software version 5.05 . The alignments were trimmed as described previously. Poisson-corrected amino acid distance with complete deletion of gaps was the distance method employed in the trees constructed. Confidence values given are bootstrap values based on 1000 bootstrap replicates. The trees were manually rooted.
Calculation of between and within group average distances
Amino acid alignments of all Populus, Arabidopsis, rice, Selaginella, and Physcomitrella EXPA and EXPB sequences were used to determine the between group and within group mean Poisson-corrected amino acid distances using MEGA 5.05. Standard error was also calculated for these values using 500 bootstrap replicates.
REC participated in the design of the study, performed the original trace archive searches, generated alignments, built phylogenetic trees, and drafted the manuscript. NKH performed genome searches, generated alignments and phylogenies, generated the between and within group distance analysis, and helped draft the manuscript. DJC conceived the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.
The authors acknowledge the contributions of P. Kerr Wall and Claude W. dePamphilis for invaluable advice on the archive searches and phylogenetic analyses presented here. REC was supported by NSF grant IBN-9874432 to DJC for the initial genomic analysis and by an LVC Arnold Experiential Grant to REC and NKH for refined genomic and phylogenetic analysis. Computational analysis was supported by DOE Office of Science grant DE-FG02-84ER13179.
- McQueen-Mason S, Durachko DM, Cosgrove DJ: Two endogenous proteins that induce cell wall extension in plants. Plant Cell. 1992, 4: 1425-1433.PubMedPubMed CentralView ArticleGoogle Scholar
- McQueen-Mason S, Cosgrove DJ: Disruption of hydrogen bonding between plant cell wall polymers by proteins that induce wall extension. Proc Natl Acad Sci USA. 1994, 91: 6574-6578. 10.1073/pnas.91.14.6574.PubMedPubMed CentralView ArticleGoogle Scholar
- Cosgrove DJ: Loosening of plant cell walls by expansins. Nature. 2000, 407: 321-326. 10.1038/35030000.PubMedView ArticleGoogle Scholar
- Jung J, O’Donoghue EM, Dijkwel PP, Brummell DA: Expression of multiple expansin genes is associated with cell expansion in potato organs. Plant Sci. 2010, 179: 77-85. 10.1016/j.plantsci.2010.04.007.View ArticleGoogle Scholar
- Cho HT, Kende H: Expression of expansin genes is correlated with growth in deepwater rice. Plant Cell. 1997, 9: 1661-1671.PubMedPubMed CentralView ArticleGoogle Scholar
- Cosgrove DJ, Bedinger P, Durachko DM: Group I allergens of grass pollen as cell wall-loosening agents. Proc Natl Acad Sci USA. 1997, 94: 6559-6564. 10.1073/pnas.94.12.6559.PubMedPubMed CentralView ArticleGoogle Scholar
- Lee Y, Kende H: Expression of beta-expansins is correlated with internodal elongation in deepwater rice. Plant Physiol. 2001, 127 (2): 645-654. 10.1104/pp.010345.PubMedPubMed CentralView ArticleGoogle Scholar
- Pien S, Wyrzykowska J, McQueen-Mason S, Smart C, Fleming A: Local expression of expansin induces the entire process of leaf development and modifies leaf shape. Proc Natl Acad Sci USA. 2001, 98: 11812-11817. 10.1073/pnas.191380498.PubMedPubMed CentralView ArticleGoogle Scholar
- Lee D, Ahn JH, Song S, Choi YD, Lee JS: Expression of an expansin gene is correlated with root elongation in soybean. Plant Physiol. 2003, 131: 985-997. 10.1104/pp.009902.PubMedPubMed CentralView ArticleGoogle Scholar
- Brummell DA, Harpster MH, Dunsmuir P: Differential expression of expansin gene family members during growth and ripening of tomato fruit. Plant Mol Biol. 1999, 39: 161-169. 10.1023/A:1006130018931.PubMedView ArticleGoogle Scholar
- Valdivia ER, Stephenson AG, Durachko DM, Cosgrove D: Class B beta-expansins are needed for pollen separation and stigma penetration. Sex Plant Reprod. 2009, 22: 141-152. 10.1007/s00497-009-0099-y.PubMedView ArticleGoogle Scholar
- Cho H, Cosgrove DJ: Regulation of root hair initiation and expansin gene expression in Arabidopsis. Plant Cell. 2002, 14: 3237-3253. 10.1105/tpc.006437.PubMedPubMed CentralView ArticleGoogle Scholar
- Cosgrove DJ, Li LC, Cho HT, Hoffmann-Benning S, Moore RC, Blecker D: The growing world of expansins. Plant Cell Physiol. 2002, 43: 1436-1444. 10.1093/pcp/pcf180.PubMedView ArticleGoogle Scholar
- Sampedro J, Cosgrove DJ: The expansin superfamily. Genome Biol. 2005, 6: 242. 10.1186/gb-2005-6-12-242.PubMedPubMed CentralView ArticleGoogle Scholar
- Choi D, Cho H, Lee Y: Expansins: expanding importance in plant growth and development. Physiol Plantarum. 2006, 126: 511-518.Google Scholar
- Vannerum K, Huysman MJ, De Rycke R, Vuylsteke M, Leliaert F, Pollier J, Lutz-Meindl U, Gillard J, De Veylder L, Goossens A, Inze D, Vyverman W: Transcriptional analysis of cell growth and morphogenesis in the unicellular green alga Micrasterias (Streptophyta), with emphasis on the role of expansin. BMC Plant Biol. 2011, 11: 128. 10.1186/1471-2229-11-128.PubMedPubMed CentralView ArticleGoogle Scholar
- Carey RE, Cosgrove DJ: Portrait of the expansin superfamily in Physcomitrella patens: comparisons with angiosperm expansins. Ann Bot. 2007, 99: 1131-1141. 10.1093/aob/mcm044.PubMedPubMed CentralView ArticleGoogle Scholar
- Sampedro J, Carey RE, Cosgrove DJ: Genome histories clarify evolution of the expansin superfamily: new insights from the poplar genome and pine ESTs. J Plant Res. 2006, 119: 11-21. 10.1007/s10265-005-0253-z.PubMedView ArticleGoogle Scholar
- Li Y, Darley CP, Ongaro V, Fleming A, Schipper O, Baldauf SL, McQueen-Mason SJ: Plant expansins are a complex multigene family with an ancient evolutionary origin. Plant Physiol. 2002, 128: 854-864. 10.1104/pp.010658.PubMedPubMed CentralView ArticleGoogle Scholar
- Sampedro J, Lee Y, Carey RE, de Pamphilis C, Cosgrove DJ: Use of genomic history to improve phylogeny and understanding of births and deaths in a gene family. Plant J. 2005, 44: 409-419. 10.1111/j.1365-313X.2005.02540.x.PubMedView ArticleGoogle Scholar
- Banks JA, et al: The Selaginella genome identifies genetic changes associated with the evolution of vascular plants. Science. 2011, 332: 960-963. 10.1126/science.1203810.PubMedPubMed CentralView ArticleGoogle Scholar
- Wang H, Singer GAC, Hickey DA: Mutational bias affects protein evolution in flowering plants. Mol Biol Evol. 2004, 21: 90-96.PubMedView ArticleGoogle Scholar
- Sarkar P, Bosneaga E, Auer M: Plant cell walls throughout evolution: towards a molecular understanding of their design principles. J Exp Bot. 2009, 60: 3615-3635. 10.1093/jxb/erp245.PubMedView ArticleGoogle Scholar
- Popper ZA, Fry SC: Primary cell wall composition of Bryophytes and Charophytes. Ann Bot. 2003, 91: 1-12. 10.1093/aob/mcg013.PubMedPubMed CentralView ArticleGoogle Scholar
- Anderberg HI, Kjellbom P, Johanson U: Annotation of Selaginella moellendorffii major intrinsic proteins and the evolution of the protein family in terrestrial plants. Front Plant Sci. 2012, 3: 33.PubMedPubMed CentralView ArticleGoogle Scholar
- Schuette S, Wood AJ, Geisler M, Geisler-Lee J, Ligrone R, Renzaglia KS: Novel localization of callose in the spores of Physcomitrella patens and phylogenomics of the callose synthase gene family. Ann Bot. 2009, 103: 749-756.PubMedPubMed CentralView ArticleGoogle Scholar
- Geisler-Lee J, Geisler M, Coutinho PM, Segerman B, Nishikubo N, Takahashi J, Aspeborg H, Djerbi S, Master E, Andersson-Gunneras S, Sundberg B, Karpinski S, Teeri TT, Kleczkowski LA, Henrissat B, Mellerowicz EJ: Poplar carbohydrate-active enzymes. Gene identification and expression analyses. Plant Physiol. 2006, 140: 946-962. 10.1104/pp.105.072652.PubMedPubMed CentralView ArticleGoogle Scholar
- Eklof JM, Brumer H: The XTH gene family: an update on enzyme structure, function, and phylogeny in xyloglucan remodeling. Plant Physiol. 2010, 153: 456-466. 10.1104/pp.110.156844.PubMedPubMed CentralView ArticleGoogle Scholar
- Banks JA: Selaginella and 400 million years of separation. Annu Rev Plant Biol. 2009, 60: 223-238. 10.1146/annurev.arplant.59.032607.092851.PubMedView ArticleGoogle Scholar
- Link BM, Cosgrove DJ: Acid-growth response and alpha-expansins in suspension cultures of bright yellow 2 tobacco. Plant Physiol. 1998, 118: 907-916. 10.1104/pp.118.3.907.PubMedPubMed CentralView ArticleGoogle Scholar
- Gray-Mitsumune M, Mellerowicz EJ, Abe H, Schrader J, Winzell A, Sterky F, Blomqvist K, McQueen-Mason S, Teeri TT, Sundberg B: Expansins abundant in secondary xylem belong to subgroup A of the alpha-expansin gene family. Plant Physiol. 2004, 135: 1552-1564. 10.1104/pp.104.039321.PubMedPubMed CentralView ArticleGoogle Scholar
- Park YB, Cosgrove DJ: Changes in cell wall biomechanical properties in the xyloglucan-deficient xxt1/xxt2 mutant of Arabidopsis. Plant Physiol. 2012, 158: 465-475. 10.1104/pp.111.189779.PubMedPubMed CentralView ArticleGoogle Scholar
- Hsieh YS, Harris PJ: Structures of xyloglucans in primary cell walls of gymnosperms, monilophytes (ferns sensu lato) and lycophytes. Phytochemistry. 2012, 79: 87-101.PubMedView ArticleGoogle Scholar
- Trace Archive Nucleotide BLAST. [http://www.ncbi.nlm.nih.gov/blast/tracemb.shtml]
- Selaginella moellendorffii v1.0. [http://genome.jgi-psf.org/Selmo1/Selmo1.home.html]
- Huelsenbeck JP, Ronquist F, Nielsen R, Bollback JP: Bayesian inference of phylogeny and its impact on evolutionary biology. Science. 2001, 294: 2310-2314. 10.1126/science.1065889.PubMedView ArticleGoogle Scholar
- Ronquist F, Huelsenbeck JP: MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics. 2003, 19: 1572-1574. 10.1093/bioinformatics/btg180.PubMedView ArticleGoogle Scholar
- Dauger Research, Inc: [http://daugerresearch.com/index.shtml]
- Nylander JAA, Wilgenbusch JC, Warren DL, Swofford DL: AWTY (are we there yet?): a system for graphical exploration of MCMC convergence in Bayesian phylogenetics. Bioinformatics. 2008, 24: 581-583. 10.1093/bioinformatics/btm388.PubMedView ArticleGoogle Scholar
- Stover BC, Muller KF: TreeGraph 2: combining and visualizing evidence from different phylogenetic analyses. BMC Bioinformatics. 2010, 11: 7. 10.1186/1471-2105-11-7.PubMedPubMed CentralView ArticleGoogle Scholar
- Swofford DL: PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Sunderland Mass. USA: Sinauer Associates Inc.; 2003.Google Scholar
- Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S: MEGA5: Molecular Evolutionary Genetics Analysis using likelihood, distance, and maximum parsimony methods. Mol Biol Evol. 2011, 28: 2721-2739. 10.1093/molbev/msr062.View ArticleGoogle Scholar