Genetic Regulation of the 2D to 3D Growth Transition in the Moss Physcomitrella patens

Laura A Moody,Steven Kelly,Ester Rabbinowitsch,Jane A Langdale

doi:10.1016/j.cub.2017.12.052

Abstract

SummaryOne of the most important events in the history of life on earth was the colonization of land by plants; this transition coincided with and was most likely enabled by the evolution of 3-dimensional (3D) growth. Today, the diverse morphologies exhibited across the terrestrial biosphere arise from the differential regulation of 3D growth processes during development. In many plants, 3D growth is initiated during the first few divisions of the zygote, and therefore, the genetic basis cannot be dissected because mutants do not survive. However, in mosses, which are representatives of the earliest land plants, 3D shoot growth is preceded by a 2D filamentous phase that can be maintained indefinitely. Here, we used the moss Physcomitrella patens to identify genetic regulators of the 2D to 3D transition. Mutant screens yielded individuals that could only grow in 2D, and through an innovative strategy that combined somatic hybridization with bulk segregant analysis and genome sequencing, the causative mutation was identified in one of them. The NO GAMETOPHORES 1 (NOG1) gene, which encodes a ubiquitin-associated protein, is present only in land plant genomes. In mutants that lack PpNOG1 function, transcripts encoding 3D-promoting PpAPB transcription factors [1] are significantly reduced, and apical initial cells specified for 3D growth are not formed. PpNOG1 acts at the earliest identified stage of the 2D to 3D transition, possibly through degradation of proteins that suppress 3D growth. The acquisition of NOG1 function in land plants could thus have enabled the evolution and development of 3D morphology.

Highlights

To discover novel regulators of 3D growth, we designed a forward genetic screen to identify mutants of the moss Physcomitrella patens that could grow as 2D filaments, but not as 3D shoots
In wild-type (WT) P. patens, haploid spores germinate to produce apical initials that divide in a single plane to generate filaments of chloronemal and caulonemal cells that are collectively referred to as protonema [2,3,4] (Figure 1A)
Most side-branch initials develop into secondary caulonemal filaments (Figure 1A), but $5% are specified to cleave in three planes, producing 3D leafy shoots known as gametophores (Figure 1B)

Summary

Introduction

Because the Ppnog1-R mutant does not produce gametophores, egg-producing archegonia and sperm-producing antheridia cannot develop, and the causative mutation could not be mapped by conventional genetic crosses. A previously published technique [9] was used to generate somatic hybrids between the infertile Ppnog1-R mutant and fertile lines of the Gransden (Gd) strain of P. patens. The presumed diploid hybrids produced phenotypically normal gametophores that generated sporophytes after fertilization (Figures 2A–2C). Spores obtained from three independent Ppnog1-R/Gd hybrid sporophytes exhibited phenotypic segregation ratios consistent with meiosis from a tetraploid, confirming that the original hybrids (and the generated spores) were diploid (Figures 2A, 2B, and 2D). Genomic DNA was extracted from diploid segregants (120 mutant, no 3D growth; 120 wild-type, 3D growth) and sequenced in two separate pools alongside both Vx and Gd parental lines. When allele frequencies for the 120 mutant individuals were plotted across all 27 chromosomes in the P. patens genome assembly, a single allel

Methods

Results

Conclusion