MADS about MOSS

Abstract

Classic MIKC-type MADS-box genes (MIKC^c) play diverse and crucial roles in angiosperm development, the most studied and best understood of which is the specification of floral organ identities. To shed light on how the flower evolved, phylogenetic and functional analyses of genes involved in its ontogeny, such as the MIKC^c genes, must be undertaken in as broad a selection as possible of plants with disparate ancestries. Since little is known about the functions of these genes in non-seed plants, we investigated the developmental roles of a subset of the MIKC^c genes present in the moss, Physcomitrella patens, which is positioned informatively near the base of the land plant evolutionary tree. We observed that transgenic lines possessing an antisense copy of a MIKC^c gene characteristically displayed knocked-down expression of the corresponding native MIKC^c gene as well as multiple diverse phenotypic alterations to the haploid gametophytic and diploid sporophytic generations of the life cycle.¹ In this addendum, we re-examine our findings in the light of recent pertinent literature and provide additional data concerning the effects of simultaneously knocking out multiple MIKC^c genes in this moss.

The moss, Physcomitrella patens, is the only non-seed plant that is amenable to an investigation of MADS-box gene function comparable to that achieved in angiosperms. P. patens possesses six MIKC^c genes which cluster into two distinct phylogenetic clades.² We recently reported a functional genetic analysis of the three genes (PPM1, PPM2 and PpMADS1) within the PPM2-like clade as an initial contribution towards gaining an understanding of the role(s) of MIKC^c genes in this moss.¹

By fusing the respective MIKC^c promoters to a GUS reporter gene, we found that both PPM1 and PpMADS1 exhibited fairly ubiquitous expression patterns in both gametophytic and sporophytic tissues. The levels of PPM1 expression were generally higher than those of PpMADS1, and PpMADS1 was not expressed in antheridia, suggesting subtle differences in the functions of these genes. The observed patterns of widespread expression resemble those characterising the majority of vascular, non-seed plant MIKC^c genes^3–7 and accord with RT-PCR results of Quodt and coworkers.² Our in situ GUS expression and RT-PCR results¹ showed that PPM2 was not expressed or was expressed at levels too low to be detected by these methods. Conversely, the original isolation of PPM2 cDNA⁸ and data from more recent expression studies² indicated that PPM2 is expressed (albeit inconsistently and weakly) ubiquitously with elevated levels of expression sometimes observed in gametangia, sporophytic feet and basal portions of sporophytic setae.² The contradictory expression data for PPM2 may derive from differences between the PPM2-reporter gene constructs used by the respective research groups^1,2 or perhaps from variations in moss culture conditions.

We also employed an antisense approach designed to knock down expression of PPM1, and perhaps closely related MIKC^c genes, in order to discern MADS-box gene function in P. patens. Knocked-down strains displayed a complex mutant phenotype comprising delayed gametangia formation and sporophyte production, diminished sporophyte yields, and morphological abnormalities in both leaves and sporophytes, findings that are generally consistent with the ubiquitous expression pattern of PPM1¹ and PPM2's expression as described by Quodt et al.²

The phenotypes of strains with single gene knockouts of PPM1, PPM2 or PpMADS1 appeared to be perfectly normal, not displaying any of the phenotypic alterations observed in PPM1 gene knock-down mutants. While it is possible that subtle, transient or conditional phenotypic changes went unnoticed, it seems more probable that genetic redundancy is responsible for these results since the PPM2-like genes exhibit a very high level of sequence similarity. In an effort to circumvent the problem of functional redundancy, we generated all double knockout combinations for PPM1, PPM2 and PpMADS1. However, the double mutants were also phenotypically unchanged. Finally we attempted to produce triple mutants by co-transforming single PPM2 knockout lines with PPM1 and PpMADS1 linear knockout constructs. Of the 31 stable transformants from two transformation experiments, 55% were shown to be double mutants in which the original PPM2 knockout was accompanied by a second gene knockout in either PPM1 or PpMADS1. However, no triple knockouts were obtained. Given the knockout frequencies generally observed in batch transformation experiments in our laboratory and those of others,⁹ between two and five of the transformants had been expected to be triple mutants. These preliminary data, albeit involving a relatively small sample of transformants, suggest that PPM1, PPM2 and PpMADS1 triple knockouts may be lethal.

We have related compelling evidence that functionally redundant PPM2-like MIKC^c genes are involved in several aspects of the moss developmental program. It has been argued that broad expression patterns like theirs represent the ancestral state of MADS-box genes in land plants, and that the sporophytic- and organ-specific expression patterns that characterise many MIKC^c genes in extant spermatophytes, including those that specify floral organ identity, correspond to a derived condition that evolved in the spermatophyte lineage following its separation from lineages that led to bryophytes and ferns and fern allies.¹⁰ Nevertheless, it is the apparent participation of PPM2-like genes in the formation of gametangia (the differentiation of reproductive organs from non-reproductive tissues at the gametophore apex) that is particularly interesting and assumes a special significance because of its analogy to the proposed role for ancestors of seed plant C-function MADS-box genes (identifying those regions of the vegetative SAM that will become reproductive organs).11 Furthermore, expression studies of MIKC^c genes in two charophycean algae, the presumed progenitors of all terrestrial plants,^12–14 suggest that they too are involved in haploid reproductive cell differentiation.¹⁵ While these functional similarities do not infer orthology and may be coincidental, we should not discount yet the admittedly controversial hypothesis that some MIKC^c genes in non-seed plants, for example PPM2-like genes of Physcomitrella, are homologous to spermatophyte class C genes and that the ancient role proposed for ancestral class C genes¹¹ has been conserved, in some form, in all major terrestrial plant taxa.

Addendum to: Singer SD, Krogan NT, Ashton NW. Clues about the ancestral roles of plant MADS-box genes from a functional analysis of moss homologues. Plant Cell Rep. 2007;26:1155–1169. doi: 10.1007/s00299-007-0.