Reconstructing the transcriptional ontogeny of maize and sorghum supports an inverse hourglass model of inflorescence development

Samuel Leiboff, Sarah Hake

Evolving a body plan: Transcriptional ontogeny of maize and sorghum supports an inverse hourglass model for inflorescence development

Selected by Alexa Sadier


How novel body plans are generated, conserved and diversify has been a fascinating question in evolutionary biology from the first embryological comparisons in the XIX century to now. Among the models that have proposed to approach this question, the hourglass model (1) has shed a lot of interest. This model predicts an hourglass-like divergence of animal embryogenesis when comparing species from the same phylum. In other words, it proposes that embryos are more divergent (morphologically or in term of gene expression) at early and late stages of development but conserved at mid-embryonic development at a stage called the “phylotypic stage”. At this stage, embryos show maximum morphological similarity and are hypothesized to show a conserved expression of developmental genes.

The advent of molecular and genomic tools led to many investigations to demonstrate the existence of an hourglass model from a transcriptional point of view in animals (2, 3). In plants (4), which exhibit a radically different mode of development with organ formation occurring largely post-embryonically, the search for hourglass patterns has been extended outside of embryonic development. Hence, recent studies suggest the existence of an hourglass pattern of development in flowering plants, not only during development but also during the floral transition which is in general representative of a taxa. In this preprint, the authors test the existence of an hourglass model for the development of the inflorescence in maize and sorghum, which exhibit morphological variation while sharing a same inflorescence organization.

Key findings

The preprint starts by establishing developmental stages of inflorescence development for both species. To this end, they combined key morphological structures of inflorescence development with the establishment of individual transcriptional profiles that span inflorescence maturation and meristem differentiation. When correlating the two metrics, they were able to assign a precise transcriptional signature to a given inflorescence developmental stage and identified 5 stages for the maize and 4 for the sorghum. The differences observed between species suggest differences in the timing of appearance and persistence (i.e. heterochronies) of some meristems.


To find new molecular markers of inflorescence development, they used a random-forest machine learning approach to link developmental stages with developmental age and their already known markers. Remarkably, this approach allowed them not only to find new markers but to accurately predict the developmental stages from other RNAseq datasets, confirming the accuracy of the model and providing an in silico staging application in plants.

Based on these results, they then compared the different phases of the expression of the key architectural genes by dynamic time warping in both species. Interestingly, they found that while the overall sequence is conserved, specific regulators associated with maize inflorescence architecture development exhibit some heterochrony (which eventually leads to morphological changes). When analyzing these factors, they showed that while general features of growth are regulated similarly between the two species, the floral organ programs are among the most heterochronic, with increased indeterminacy in sorghum being correlated with the high-order branching morphology of this species.

Finally, they tested for the presence of hourglass patterns during the development of the inflorescence by looking at transcriptional conservation for both species. These analyses revealed some hourglass patterns at different distant developmental stages for the two species, suggesting that the selective pressures underlying these patterns have changed during sorghum evolution. As a result, the two species unexpectedly follow an inverse hourglass pattern that is generally observed in distant species with highly dissimilar body plans, suggesting a short time evolutionary history of selective pressures that have led to the hourglass patterns to be re-positioned differently in the two species. More broadly, these results demonstrate that hourglass patterns are applicable to post-embryonic development in plants and can evolve fast during evolution.

What I like about this work

One key issue when comparing the development of two different species is to find equivalent stages. This can be tricky since organs can be morphologically very different, or heterochronies can occur. As a result, the description of a given developmental stage has to be based on multiple non-redundant, non circular evidences which is not often the case in papers comparing developmental stages and testing for the presence of hourglass patterns. I really like that this preprint was really strong in defining and then comparing stages, even proposing a model able to test data from external datasets, which allowed them to discuss shifts in development such as heterochronies in a very accurate and convincing manner, as well as providing tools for defining developmental stages for future research.

What I also liked a lot in this preprint is the distance the authors took from their model, allowing them to pinpoint interesting hypotheses and to question the hourglass model and its evolution. I particularly like the idea of shifts in hourglass patterns in responses to different selective pressures.


Future directions/Questions for authors

  • Do you expect to find hourglass patterns of development outside flowering plants, for example in taxa with very different life cycle?
  • Do you think that post-embryonic developmental hourglass patterns are more suceptible to change more rapidly than the embryonic ones?
  • Do you have any interesting “behind the preprint” story you would like to share?


Additional references

  1. Irie N, Kuratani S. 2014 The developmental hourglass model: a predictor of the basic body plan? Development 141, 4649–4655.
  2. Kalinka AT, Varga KM, Gerrard DT, Preibisch S, Corcoran DL, Jarrells J, Ohler U, Bergman CM, Tomancak P. 2010. Gene expression divergence recapitulates the developmental hourglass model. Nature468:811–814.
  3. Levin M, Anavy L, Cole AG, Winter E, Mostov N, Khair S, Senderovich N, Kovalev E, Silver DH, Feder M, Fernandez-Valverde SL, Nakanishi N, Simmons D, Simakov O, Larsson T, Liu S-Y, Jerafi-Vider A, Yaniv K, Ryan JF, Martindale MQ, Rink JC, Arendt D,Degnan SM, Degnan BM, Hashimshony T, Yanai I. 2016. The mid-developmental transition and the evolution of animal body plans. Nature531:637–641
  4. Quint M, Drost H-G, Gabel A, Ullrich KK, Bönn M, Grosse I. 2012. A transcriptomic hourglass in plant embryogenesis. Nature490:98–101.


Tags: development, evo-devo, evolution, hourglass, maize, plant biology

Posted on: 11th June 2019

Read preprint (No Ratings Yet)

  • Have your say

    Your email address will not be published. Required fields are marked *

    This site uses Akismet to reduce spam. Learn how your comment data is processed.

    Sign up to customise the site to your preferences and to receive alerts

    Register here