Results

Genomic distribution of poplar MAPK and MAPKK genes
Previous analysis of the genome sequence of poplar (Populus trichocarpa) had identified robust gene models corresponding to all the MAPK (PtMPK) and MAPKK (PtMKK) family members [39]. With this information, we were able to obtain an overview of the chromosomal distribution of these important signaling components. PtMPK genes are distributed over 12 of the 19 poplar chromosomes (Figure 1). Chromosomes I and II both carry three divergent PtMPK genes, whereas chromosomes V, VII and X display two PtMPK genes each. The remaining poplar MAPK genes are unique with respect to their chromosomal location. PtMPK7 and PtMPK18 have not yet been assigned to any linkage group, and therefore remain positioned on their respective scaffolds. Interestingly, although there are numerous PtMPK paralogs displaying high levels of sequence similarity, they are distributed all across the genome and do not form clusters containing closely related genes as may have been expected if they originated from local duplication events. This pattern probably reflects the series of whole genome, chromosomal and large segmental duplication events that typify the poplar genome (G. Tuskan, personal communication).
PtMKK genes also display a scattered genomic distribution (Figure 2) across six of the 19 poplar chromosomes, with three (PtMKK6, PtMKK7 and PtMKK10) of the 11 gene models located on unattributed scaffolds. Only chromosomes VIII and X contain more than one PtMKK gene.

Exon and intron organization of poplar MAPK and MAPKK genes
Analysis of the pattern of exon-intron junctions can provide important insights into the evolution of gene families. Therefore, we extracted data regarding predicted exon and intron distribution for the coding regions of all PtMPKs and PtMKKs (Figures 3 and 4) as well as for all Arabidopsis putative orthologs (AtMPKs, Figure 5 and AtMKKs, Figure 6). Group A PtMPKs exhibit a highly conserved distribution of exons and introns (Figure 3) consisting of six exons of conserved length, and five introns of conserved or variable sizes. PtMPKs belonging to group B also possess six exons, with lengths similar to those found in group A PtMPKs, while the associated introns vary in size between the different members of group B. Group C PtMPKs are each composed of only two exons with strictly conserved or very similar sizes. PtMPK14 is the only group C member with a shorter intron (398 vs ~ 1200 base pairs for the other three members).
In contrast to these three highly conserved structural patterns, group D PtMPKs possess a complex distribution of exons and introns, including different pattern subsets within the same phylogenetic group. For instance, PtMPK9-1 and PtMPK9-2 are both composed of 11 exons, whereas PtMPK16-1, PtMPK16-2, PtMPK19, PtMPK20-1 and PtMPK20-2 all possess ten. Despite some modest differences in the length of particular exons, it is clear that the exon structural pattern is well conserved not only between close paralogs (e.g. PtMPK16-1 and PtMPK16-2), but also between group D PtMPKs that apparently diverged following earlier duplication events (e.g., PtMPK16-1 and PtMPK19). These same patterns are also found in the Arabidopsis MPK gene family with the exception of group B MPKs, which display three different patterns of exon-intron distribution (Figure 5).
The MKK genes display two strikingly different structural patterns in both poplar and Arabidopsis (Figures 4 and 6). Members of group C and D MKKs have a completely intronless configuration, whereas the group B MKK3s and all group A MKKs possess numerous exon and intron junctions. In poplar, PtMKK2-1, PtMKK2-2 and PtMKK6 show quite strong exon length conservation, with the exception of an additional 17 base pairs exon in PtMKK2-1. PtMKK3, on the other hand, has a completely unique exonic structure, consistent with both its evolutionary distinctiveness [39] and the presence of an unusual C-terminal NTF2 domain that is not found in any other MKK group.
Intron phase (i.e., the position of an intron within a codon; phase 0 when lying before the first base, phase 1 when lying after the first base and phase 2 when lying after the second base) was also assessed for all PtMPK and PtMKK gene models (Figures 3 and 4), as well as for all AtMPK and AtMKK gene models (Figures 5 and 6). For both PtMPKs (58%) and PtMKKs (73%), the majority of introns are within phase 0, while 23% of introns found in both poplar protein kinase families are within phase 2. Phase 1 introns represent 19% of all PtMPKs introns and only 4% of all PtMKKs introns. For Arabidopsis, similar numbers are found (Figures 5 and 6), except that there are no phase 1 introns predicted within AtMKK genes.
The association of two adjacent introns in eukaryotic genes can be in any of nine different intron phase combinations, leading to two classes of exons: symmetric exons (0-0), (1-1), (2-2) and asymmetric exons (0-1), (0-2), (1-0), (1-2), (2-0), (2-1). For poplar MPKs, 51% of all exons are symmetric and the great majority of these are (0-0) exons (79%). A similar picture is found in Arabidopsis, where 54% of AtMPKs exons are symmetric. Once again the majority of these are (0-0) exons (85%). No symmetric exons harboring the (1-1) configuration are found in any of the poplar or Arabidopsis MPK gene models. For the PtMKK and AtMKK gene families, we respectively found 58% and 68% of symmetric exons and for both species, all of these symmetric exons are in the (0-0) configuration. For both plants, the most frequent asymmetric exons found in MPKs and MKKs are those belonging to the (0-1), (0-2) and (2-0) configurations.

Real-time quantitative PCR data normalization and general considerations
The real-time, fluorescence-based reverse transcription polymerase chain reaction (RTqPCR) technique has become a method of choice because of its wide dynamic range, sensitivity and robust quantification of mRNA levels. In contrast to microarray profiling, transcript accumulation (TA) of closely-related gene members can also be easily discriminated in RTqPCR by using oligonucleotide primers specific to unique gene signatures. Such comparisons are only valid, however, if the chosen primer sets display comparable efficiencies in their ability to amplify targeted amplicons. We therefore tested all the selected primer sets against genomic DNA from three different genetic backgrounds, namely Populus trichocarpa (Nisqually-1), Populus trichocarpa X Populus deltoides (H11-11) and Populus deltoides (ST-70). In these assays, most primer sets yielded a very similar Ct value (around 21 for most genes) across the three different genetic backgrounds (see additional file 1). This indicates that these primer pairs can equally hybridize to the P. trichocarpa alleles in the Nisqually-1 background, to the P. deltoides alleles in the ST-70 background and most importantly to both the P. trichocarpa and the P. deltoides alleles in the H11-11 hybrid, which was used in this study to monitor PtMPK and PtMKK expression profiles. On the other hand, primer pairs specific for PtMPK2, PtMPK4, PtMPK9-2, PtMPK16-2, PtMKK4 and PtMKK7 all display higher Ct values in the P. deltoides background. This reflects that these gene specific primer pairs preferentially hybridize to the P. trichocarpa allele in the hybrid genotype, and that TA obtained on cDNA (see below) might slightly underestimate the actual level of expression since the P. deltoides allelic contribution is not perfectly captured. This observation could be related to lower primer set efficiency, resulting from polymorphism within the 3'UTR region of the different alleles. Nevertheless, melting curve and gel electrophoresis analysis confirmed single product amplification for all respective MPKs and MKKs in the three poplar genotypes (data not shown). Moreover, within one particular genetic background, the steady Ct value obtained for the different MPKs and MKKs confirmed that for a constant number of target sequences (e.g., 5 ng of DNA), each primer set gave a similar Ct value (see additional file 1). This clearly demonstrates the similar efficiencies of the various primer sets used in this study, and allows direct gene-to-gene comparisons of the levels of expression detected in the various organs sampled.
Finally, the use of internal standard candidate genes that display relatively stable expression over time and between tissues or organs allows sample-to-sample normalization of the RTqPCR expression data. In the present study, the use of RTqPCR primers specific for a cyclin-dependent kinase 2 (cdc2) gene revealed generally consistent levels of cdc2 TA in different organs from poplar (Figure 7A, B, C), with less than two-fold variation observed across most vegetative organs, and in suspension-cultured cells (Figure 8). Slightly higher transcript abundance of cdc2 was detected in actively growing organs, such as floral and foliar buds, where CDC2 is likely involved in cell-cycle progression [40,41]. Therefore, cdc2 expression levels provided a good normalization baseline. In addition, since Actin 2 (Act2) has been stated to be a good internal control gene for RTqPCR studies across various poplar organs [42], we have used this candidate as housekeeping gene. TA for Act2 proved to be quite stable across most organs, with a maximum five-fold increase from mature leaf to upper stem (Figure 8). This level of variation seems relatively low considering the wide diversity of tissues tested in our survey (from meristematic organs such as buds to mature leaves). Moreover, our results revealed that two poplar MAPKs (PtMPK6-1/6-2) display quite constant TA in all assayed organs (Figure 8). The highest TA was detected in the female floral bud sample, but this level of expression represents only a one-fold increase in comparison to the average TA in all organs. Finally, PtMKK2-2 was also detected at low but constant levels in all tested organs. The respective TA levels for cdc2, Act2, PtMPK6-1, PtMPK6-2 and PtMKK2-2 across many different organ types provide confirmation that an appropriate and consistent dosage of cDNA was used in the various RTqPCR reactions. In our analysis, TA levels corresponding to <100 transcript molecules per ng total RNA were scored as 'very low', values of 100–400 as 'moderate', values >400 as 'high', and values >1000 as 'very high'. Levels <10 were treated as effectively zero.

Poplar MAPK and MAPKK gene expression patterns
Virtually all PtMPKs and PtMKKs are expressed in all organs analyzed, but their level of expression varies considerably. Regardless of the phylogenetic groups or organs examined, the TA for PtMPKs generally fluctuates between moderate to very high levels (See additional file 2). Members of group D MPKs show TA levels of ~1500 transcript molecules per ng of total RNA, while members of the group A, B and C MPKs show lower levels, around 400–600. For the PtMKK gene family, the most highly expressed members are those belonging to group C (TA >1600, see additional file 3). Group B and D MKK genes have similar levels of TA (~1100–1600). Finally, group A MKK genes are the most weakly expressed, with on average 550 transcript molecules per ng of total RNA.

MAPKs

Group A MPKs
The most extensively studied plant MAPKs belong to Group A which, in Arabidopsis, consists of three members, AtMPK3, AtMPK6 and AtMPK10 [4]. No direct putative ortholog of AtMPK10 has been detected in the poplar genome, but phylogenetic analysis [39] revealed the presence of two closely-related poplar presumed orthologs for the defense-related genes AtMPK3 (PtMPK3-1 and 3-2) and AtMPK6 (PtMPK6-1 and 6-2). In most organs, the expression of both PtMPK3-1 and 3-2 is lower than that of PtMPK6-1 and 6-2 (Figure 9; see additional file 2). PtMPK3-1 is relatively strongly expressed in roots and xylem, in comparison to other samples. In most organs PtMPK3-1 tends to be slightly more expressed than PtMPK3-2. However, this pattern is reversed in the four types of buds (male and female floral buds, lateral and terminal foliar buds) as well as in both types of catkins. PtMPK6-1 and 6-2 both show similar expression profiles across many organs, but TA for PtMPK6-1 becomes slightly more pronounced than that of its paralog in all four types of buds, in both types of catkins and in suspension-cultured cells.

Group B MPKs
There is less information on the biological roles of the other MAPK groups in plants, although some reports have suggested the potential involvement of group B MPK genes in response to environmental stresses as well as in cell development [4]. In poplar, PtMPK11 and PtMPK5-2, the most highly expressed of the four group B MAPK genes, are particularly actively transcribed in male and female floral buds (Figure 9). By contrast, the paralogous PtMPK5-1 gene shows the lowest level of TA within this group.

Group C MPKs
Among the group C MPKs, it has been reported that the tobacco Ntf3 gene is expressed in pollen [43] and that the Arabidopsis AtMPK7 gene has circadian rhythm-regulated patterns of expression [44]. The most highly expressed of the four poplar group C MPK genes is PtMPK7, with elevated TA levels detected in female catkin, buds, phloem, xylem, mature leaves (LPI 12) and roots (Figure 9). This gene is also differentially expressed in particular developmental stages of specific organs, with more abundant transcripts detected in floral buds (male and female) than in either type of catkin. A similar situation is observed for leaves, where PtMPK7 TA is more pronounced in mature leaves than in young leaves. This is similar to what has been reported for the rice putative ortholog OsMAPK4 (now annotated as OsMPK7 [39]), whose expression is higher in mature leaves than in young leaves [45].

Group D MPKs
The group D MPKs represent the largest group of MPKs in poplar, as they do in Arabidopsis and rice [4,39]. In rice and alfalfa (Medicago sativa), two group D MAPK genes (OsBMWK1 and MsTDY1) are induced transcriptionally by pathogen challenge and wounding, respectively [36,46] Activated OsBMWK1 (now annotated as OsMPK17-1 [39]) has also been shown to phosphorylate a transcription factor that binds a cis-acting element in the promoter of defence-related genes [47].
In poplar, PtMPK17 is the most highly expressed of the group D MPK genes and, indeed, it is the most highly expressed among all the MPK and MKK genes (Figure 9). On the other hand, the most weakly expressed group D MPK genes, PtMPK9-1 and PtMPK9-2, display low transcript levels in all organs with the noteworthy exception of all types of buds, and cell suspensions. As previously observed in other MPK groups, some members of the group D MPK genes seem to represent the paralogous products of recent genomic duplication events, since they possess a very high degree of sequence similarity, are located on different chromosomes and have only one direct putative ortholog in Arabidopsis [39]. PtMPK16-1 and PtMPK16-2 are particularly interesting paralogs, since they have very similar expression profiles in most organs, including male and female catkins. On the other hand, PtMPK16-1 is strongly expressed in male and female floral buds, whereas expression of PtMPK16-2 is barely detectable in these reproductive organs.

MAPKKs

Group A MKKs
In other plant species, some group A MKKs appear to be functionally associated with group B MPKs [48,49] These phosphotransfer relationships have been involved in responses to abiotic stresses in Arabidopsis, and in cell development in tobacco. As in Arabidopsis, there are three group A MKK genes found in poplar. PtMKK2-2 shows low but constant levels of TA in all tested organs (Figure 10; see additional file 3), while the paralogous PtMKK2-1 is much more highly expressed. Higher expression of PtMKK6, on the other hand, seems to be associated with proliferating organs such as apex, floral and terminal buds, cell suspensions, and young leaves (LPI 1), with a 25-fold decrease in PtMKK6 expression levels observed along the foliar developmental gradient from young to mature leaves. Interestingly, the presumed Arabidopsis and tobacco orthologs of PtMKK6 have been involved in regulation of cytokinesis and cell division [13], suggesting that this protein may play an analogous role in poplar tissues.

Group B MKKs
Group B MKKs in poplar are represented by a single gene, PtMKK3. This is also seen in Arabidopsis (AtMKK3) and rice (OsMKK3) [39]. MKKs of this class are unique in encoding a characteristic MKK protein kinase domain fused in C-terminal to a putative nuclear transport factor 2 (NTF2) domain [4]. No biological functions have yet been assigned to plant MKK3s, but PtMKK3 has moderate to relatively high expression levels across all organs (TA between 400 to 2600), with the highest levels detected in female floral buds, lateral foliar buds and cell suspensions (Figure 10).

Group C MKKs
Among the plant MKKs, attention has been largely focused on those found in group C because of their demonstrated roles in stress signaling. Ectopic expression of constitutively-activated versions of the tobacco NtMEK2 protein, and of other group C MKKs, has been used to demonstrate that they are capable of phosphorylating and thus activating stress-responsive Group A MPKs [50-52]. Poplar possesses two group C MKKs, PtMKK4 and PtMKK5, both of which are expressed. Of the two, PtMKK5 shows higher TA in most organs, while the expression of PtMKK4 only predominates in suspension-cultured cells (Figure 10).

Group D MKKs
Limited functional information is available for group D MKKs, of which there are five encoded representatives in the poplar genome (PtMKK7, PtMKK9, PtMKK10, PtMKK11-1 and PtMKK11-2). However, our RTqPCR analysis suggests that only PtMKK7 and PtMKK9 are clearly expressed (Figure 10). The other three genes may therefore be pseudogenes, or be expressed only under circumstances that were not tested in our survey. The pseudogene hypothesis is also supported by the observation of structural differences in normally conserved motifs within the predicted PtMKK10, 11-1 and 11-2 protein kinase domains, and by the apparent absence of expression data for the Arabidopsis (AtMKK10) and rice (OsMKK10-1/10-2) putative orthologs [39].
While both PtMKK7 and PtMKK9 expression could be detected, the patterns differ significantly across organs and developmental stages (Figure 10). PtMKK9 expression is generally more pronounced except in secondary xylem, cell suspensions, primary phloem and xylem cambium-enriched, where PtMKK7 predominates. PtMKK9 is particularly highly expressed in mature leaves (LPI 12), where it reaches close to 10 000 transcript molecules per ng of total RNA, in contrast to the younger leaf sample (LPI1). The levels of transcript accumulation for PtMKK7 are on the other hand relatively constant across most organs. This striking difference in expression pattern has also been observed in Arabidopsis expression databases for AtMKK7 and AtMKK9 [53] where AtMKK9 transcription is most strongly associated with mature or senescing leaves.