Leaf shape development

---

Leaf shape development

On of the things that intrigues me most in biology is the development of organisms. How does that single cell that is just fertilised knows what to do. To get its polarity established, initiate cell division at the right time, place and direction. What makes it go on developing into recognisable plant and not just a mass of cells. It simply fascinates me. I guess that is what is exciting of the research coming from Enrico Coen lab. I was first introduced to what his lab was doing while working at the John Innes Centre, through our departmental seminars and the annual research days. Here I was introduced to the concepts of how you could compute the development of a leaf.

This they did through painstakingly following the developments of Arabidopsis leaves, tracking cell divisions, from the tiny leaf primordia up to the fully grown leaf. Studying the relationship between cell division, cell size and growth rate. This information they then used to feed into computational models. Which as explained in a recent publication that leaf shape could be brought back to a few parameters and growth factors. The important parameters included growth rate (perpendicular and proximodistal) division rate (division competence and mean threshold cells size). And the growth factors could be brought back to a graded proximodistal factor (PGRAD), a mediolateral factor (MID), a factor distinguishing lamina from petiole (LAM), timing factor (LATE),  and proximal mobile factor (PMF). Interestingly by slightly tweaking the influence these growth factors have an effect on the leaf growth the size of the leaf. This is corresponding to what is observed in mutants with a leaf size phenotype. Another exciting part of this research is that it shows that it is possible to obtain variation in leaf size with just a defined set of growth factors, corresponding to transcription factors in the plant.

These initial studies focused mainly on simple leaves, flat and round. But now they taken this to a whole new level, going from a 2D to a 3D leaf shape. There they show how carnivorous plant carnivorous plant Utricularia gibba adjust their planar (simple) leaves into needle-like structures and cup-shaped traps. First using the same through approach  they used 3D imaging to track cell divisions and to obtain growth rate measurements in three dimensions to use to build a model of how the cup-shaped traps develop. This resulted in a model similar to that of an Arabidopsis leaf, with parameters for growth rate and division rate and growth factors representing a mediolateral factor (MID), ventral midline factor (VEN), and a stalk diameter factor (STK). Using the observed parameters they showed that model was able to reproduce the observed development of an Utricularia gibba cup-shape trap. In addition they showed that by adjusting the parameters they could generate trap morphologies similar to other Utricularia species.

However, a limiting point of the model is in that it considers only the later stages of development of the trap. Not explaining how the curvature of the trap primordium originates. One way of how the initial curvature for the cup-shaped trap might be initiated from a two dimensional origin is through abaxial and adaxial patterning. Comparing the cup-shaped traps with simple leaves it was deducted that the outer side of the cup corresponds to the abaxial (lower) side, and the inner side to the adaxial (upper) side of a simple leaf. Therefor the latest work of the Coen lab investigates the influence of the effect of adaxial and abaxial domains on the polarity field that orientates growth. The decision if cells are part of the adaxial or abaxial site of the leaf occur via gene activity in the leaf primordia. However, solely based on leaf morphology you would not able to distinguish between future leaves or future traps as they are all dome-shaped.

To understand how adaxial and abaxial domain specification might influence trap development the adaxial-abaxial domains in developing traps were identified. While in primordia of leaves the adaxial and abaxial domains each take up approximately halve of the dome-shaped primordia, in primordia of traps the adaxial domain was much smaller than the abaxial domain. To investigate the influence of this on trap development, they build a model in which they could specify the adaxial and abaxial regions. Finding that having an equal adaxial-abaxial distribution would result in either simple leaves or needle like leaves, depending of the settings of the parameters for the different growth rates. However, when they restricted the adaxial region to only a small region on one side of the primordia they found that the resulting leaf would develop into a cup-shaped trap. Furthermore, by adjusting the size and form of the adaxial region, traps of different shapes could be generated.

With developing these models the Coen lab has illustrated the universal mechanism that is behind leaf development. The possibility to generate completely different leaves by adjusting only a few parameters points towards a small group of transcription factors whose differences in expression and activity between species might explain the variation in leaf shapes we see.

Finally, even if you did not understand much about what I have been talking about in this post, I would like to point out that they made lots of movies, both of the models as well as of the imaging they did in order to obtain all the data to feed into the models. You can have a look in the supplemental data of the discussed papers, as well as on the Coen lab website.