Alfonso Martinez Arias (http://bit.ly/14zBcke)
An EMBO workshop organized by Alex Schier and James Briscoe, assembled a cast of young and seasoned biologists in Oxford to discuss progress and controversies on Morphogen Gradients, a central topic in Developmental Biology. I could only stay for half of the meeting but the message was clear: there is a new era in Developmental Biology. We could call it Systems Developmental Biology or, if you do not like the association with this vague word Systems, call it Quantitative Developmental Biology. IN any event, a change in the way problems are addressed was clear: measurement, quantification, modelling and more important, analytical approaches and a critical eye on the problem were central to each and every talk. You may want to say that this is a credit to the organizers, which it is, but it is also the fact that this is the new way of an interesting field and the way forward. Developmental Biology is coming of age. This is not the place to go through the customary list of speakers and talks that made up the meeting; you can get a glimpse of this in the website (http://events.embo.org/13-morphogen/). In any case, it was the general tone and drift that caught my eye.
The notion of ‘Morphogen’ was introduced by Alan M Turing in 1952 to consider the potential of diffusible chemicals (Morphogens) to generate spatial and temporal patterns through chemical reactions. In the 70s Lewis Wolpert and his disciples brought together Gradients and Morphogens in a much more biological context, that of the generation of specific patterns, and in doing so ushered an important new era in Developmental Biology. Central to the wolpertian view was the notion of Positional Information: that there were molecules which would diffuse from a source, generate gradients and elicit specific responses in a concentration dependent manner; cell would respond to a given concentration according to its position. The gradients became “gradients of Morphogens’ and were said to be Universal. The response was more flexible and was a cell type specific. The blend of Positional Information and Morphogen gradients placed itself at the center of pattern formation and allowed a conceptualization of developmental biology in terms of signals, responses, fields, scaling that is still with us. But there was a problem, the theoretical framework lacked data.
Jeremy Gunawardena has written recently on the role of theory in Biology (1, 2) and while he chose transition states in enzyme kinetics and the gene as his examples –their notions helped Biology well before they were identified experimentally- I would add “Morphogens’ as a third theoretical entity that helped the growth, in this case of developmental biology. By the late 70s the concept of Morphogen gradients was THE conceptual reference to deal with problems in pattern formation but………. there were no Morphogens. Those with long memories will remember the rise and fall of cAMP and DIF in Dictyostelium and of Retinoic acid in the specification of digits in the vertebrate limb, and how we had to adjust our lectures through the 80s as experiments and new findings wiped the slate clean on the role of these molecules. But finally, two molecules broke in and established themselves as bona fide Morphogens: Activin in Xenopus and Bicoid in Drosophila. To do this Activin, used a very traditional blend of Biochemistry and embryology, whereas Bicoid came through the route of Genetics. It was the Genetics that would become the tool of choice in Developmental Biology and that was going to reel in other candidates; the reason for this was and is that some of these substances are in concentrations which elude the detection thresholds of the biochemical techniques but the loss of function approached of Genetics do not have this limitation.
The 90s was a good time for Morphogen gradients as genes were identified encoding molecules that, with some leniency, would fulfil the wolpertian view; in particular the criteria of inducing patterns in a concentration dependent manner. A sense of triumph was in the air. Intriguingly, and this has been a sin of the genetic approaches to developmental biology, in the euphoria we forgot some of the basis of Morphogen gradients. What is the shape of the gradients and what consequences this would have for their interpretation? If they were based on diffusion, how reproducible are they? What is the role that time, the dynamics of gradient formation and read out, played in the development of pattern? How robust is the response? The theory of Morphogen gradients created a quantitative framework that needed to be explored. Moreover, Wolpert had been explicit in addressing the problem of pattern scaling, which needed to be related to the mechanism of Morphogen activity. These issues which, for the most part have been swept under the carpet, were at the center of the EMBO meeting.
In the early 00s, three laboratories, all of them working with Drosophila, rose to the challenge of dealing with the real questions of Morphogen gradients. They did so by blending physical approaches with an extension of the potential of genetics beyond the identification of genes challenge. John Reinitz and Eric Wieschaus in the US tackled the early pattern formation of Drosophila embryos and Marcos Gonzalez Gaitan, addressed the issue of the presumed gradients of Dpp in the wing imaginal discs and its consequences. This work was at first seen as esoteric, but in the course of time has established standards, reformulating questions and generating new perspectives on old problems; in the process, it has revealed some surprises. Many other people have followed and the meeting in Oxford was a lively exponent that these arguments are now at the heart of the discussions.
If the gradients required detailed interrogation and analysis, the response of the cells is in need of no lesser scrutinity. Wolpertian gradients posed linear gradients and linear interpretations but biological systems are different. Linear approximations might work in some instances but the response, particularly when signal transduction is in the middle, is more complicated. Here the work of James Briscoe (one of the organizers) on the translation of the continuous gradient of Sonic Hedgehog into discrete states in the vertebrate neural tube, paves the way for progress. Gradients need to be decoded and time, as well as the traditional spatial variable, plays an important role. On the side of the interpretation, more classical molecular biology techniques are important and the development of genomics is central to understanding this important part of the action of Morphogen gradients and the programme of the second day, which I had to miss, had much bearing on this. Networks and models linking patterns and genes are the elements of this analysis. Most importantly, as has been emphasized by Arthur Lander, there has to be a feedback between the input and the output and we are far from undertanding it. Early days, but moving forward.
Whereto Genetics in this new era? I have expressed my thoughts elsewhere (http://bit.ly/10fZkmK) but the meeting made one thing clear: Genetics has been useful to break the system into component parts and to identify those parts, but it is not the tool of choice to put the system back together; it cannot be. Genetics still works well to identify components but in the new era, it is a perturbation tool to challenge the system out of its comfort zone (the one chosen by Evolution) to test some of the predictions of the models. This much was also clear, though perhaps not explicit, at the meeting.
With a few exceptions, every talk I sat through in Oxford was quantitative, analytical and critical of itself and of the field. While there were glimpses of solutions to particular problems, there was a general feeling that there is a way to go in some of the important questions and this critical look at the problems is, it seems to me, a good thing. What we have in front of us, whether Morphogen gradients or decision making during development, are formidable tasks which cannot be ‘solved’ by joining gene names with arrows or drawing linear gradients that outline French flags (even Lewis Wolpert admits this). Over the last few years interactions with the Physical Sciences are changing Biology. The result goes beyond Biophysics, while this looks at Biology for problems that look like Physics, the new interactions looks at physics for inspiration and methods to address biological problems which, often, physics has never encountered. The benefits of this new approach were in display at the workshop in Oxford and make us look forward to a future that will not only give us more a satisfactory understanding of the system at the mechanical level, but also will begin a more mature and quantitative biology. After all, if there is a frustration in Biology (and this was an undercurrent of the meeting and where it differs from Physics) is that Natural Selection has ensured that there is fine tuning at all levels of description, from the levels and timing of the expression of a gene, to the scaling and dimensionality of tissues and organs. Such a system will require much that is new in terms of concepts and techniques. What a formidable challenges but also, what fun!
1. Gunawardena J. (2012) Some lessons about models from Michaelis Menten. Mol. Biol. Cell 23, 517-519.
2. Gunawardena J. (2013) Biology is more theoretical than physics. Mol. Biol.Cell 24, 1827-1829.