The stele - a developmental perspective on the diversity and evolution of primary vascular architecture

The stele concept is one of the oldest enduring concepts in plant biology. Here, I review the history of the concept and build an argument for an updated view of steles and their evolution. Studies of stelar organization have generated a widely ranging array of definitions that determine the way we classify steles and construct scenarios about the evolution of stelar architecture. Because at the organismal level biological evolution proceeds by changes in development, concepts of structure need to be grounded in development to be relevant in an evolutionary perspective. For the stele, most traditional definitions that incorporate development have viewed it as the totality of tissues that either originate from procambium - currently the prevailing view - or are bordered by a boundary layer (e.g. endodermis). Consensus between these two perspectives can be reached by recasting the stele as a structural entity of dual nature. Following a brief review of the history of the stele concept, basic terminology related to stelar organization, and traditional classifications of the steles, I revisit boundary layers from the perspective of histogenesis as a dynamic mosaic of developmental domains. I review anatomical and molecular data to explore and reaffirm the importance of boundary layers for stelar organization. Drawing on information from comparative anatomy, developmental regulation, and the fossil record, I propose a stele concept that integrates both the boundary layer and the procambial perspectives, consistent with a dual nature of the stele. This dual stele model posits that stelar architecture is determined at the apical meristem by two major cell fate specification events: a first one that specifies a provascular domain and its boundaries, and a second event that specifies a procambial domain (which will mature into conducting tissues) from cell subpopulations of the provascular domain. If the position and extent of the developmental domains defined by the two events are determined by different concentrations of the same morphogen (most likely auxin), then the distribution of this organizer factor in the shoot apical meristem, as modulated by changes in axis size and the effect of lateral organs, can explain the different stelar configurations documented among tracheophytes. This model provides working hypotheses that incorporate assumptions and generate implications that can be tested empirically. The model also offers criteria for an updated classification of steles in line with current understanding of plant development. In this classification, steles fall into two major categories determined by the configuration of boundary layers: boundary protosteles and boundary siphonosteles, each with subtypes defined by the architecture of the vascular tissues. Validation of the dual stele model and, more generally, in-depth understanding of the regulation of stelar architecture, will necessitate targeted efforts in two areas: (i) the regulation of procambium, vascular tissue, and boundary layer specification in all extant vascular plants, considering that most of the diversity in stelar architecture is hosted by seed-free plants, which are the least explored in terms of developmental regulation; (ii) the configuration of vascular tissues and, especially, boundary layers, in as many extinct lineages as possible.