Shigeru Kondo Osaka University

Making a pattern from no pattern

In 1952, Alan Turing presented a theoretical framework for biological pattern formation that explained how positional information can be generated in an embryo. At the beginning of the 21st century, the “Turing model” has become widely accepted throughout the scientific community.

Fig 1. In 1952, Alan Turing presented a theoretical framework for biological pattern formation that explained how positional information can be generated in an embryo. At the beginning of the 21st century, the “Turing model” has become widely accepted throughout the scientific community.

Be it the staggering beauty of bird feathers, the iridescent coloring of a butterfly wing or the breathtaking skin motifs of some fish, the immense variety of animal pigmentation patterns has at the same time amazed und puzzled natural scientists from ancient Greece till modern times. Which controlling mechanisms are hidden behind such exquisite design? By which pencil are these patterns written down on the virgin pages of a developing embryo? Answers have eluded biologists for centuries.

Until 1952, when a period of little scientific progress in the field came to an end through the work of a single man, brilliant mathematician and theoretical biologist: Alan Turing. During that year, his now legendary paper, entitled “The Chemical Basis of Morphogenesis” offered a theoretical framework for biological pattern formation, thereby obviating any further need for speculation. By developing a mathematical model, Turing demonstrated that a combination of chemical reaction and diffusion - two diffusible substances interacting with each other - can generate a wide variety of spatial motifs, such as stripes, spots and spirals. He proposed that such reaction-diffusion mechanism can serve as the basis of pattern formation during morphogenesis. In mid 20th century, such a hypothesis was nothing short of amazing - and indeed, for many years to come, only few would grasp the scope of Turing's concepts - as it offered a biochemical basis for generating positional information without the existence of any pre-pattern. Or in other words, it described how non-uniformity can arise biochemically out of a homogeneous, uniform state - the scientific enigma that kept developmental biologist awake at night, in those days.

Turing patterns come alive

Stripe formation in the emperor angelfish (Pomacanthus imperator)- Top: Picture of the same individual, as a juvenile (left) and an adult (right): note that the number of horizontal stripes has doubled while the spacing between the stripes remains unchanged./- Bottom lef: Simulation, based on the Turing model, explaining the generation of stripes./- Bottom right: The bifurcation of the horizontal stripes proceeds exactly as predicted by the simulation model.

Fig 2. Stripe formation in the emperor angelfish (Pomacanthus imperator)
- Top: Picture of the same individual, as a juvenile (left) and an adult (right): note that the number of horizontal stripes has doubled while the spacing between the stripes remains unchanged.
- Bottom lef: Simulation, based on the Turing model, explaining the generation of stripes.
- Bottom right: The bifurcation of the horizontal stripes proceeds exactly as predicted by the simulation model.

Nowadays the Turing model is widely recognized as a robust mathematical theory, capturing a biological phenomenon in a unique way, but that has not always been the case: for many years biologists were not convinced of its in vivo relevance, no matter how faithfully it could simulate a broad repertoire of patterning events. Convinced of its fundamental importance however, Prof. Kondo has directed his main scientific effort for the past two decades towards finding evidence for the “Reaction- Diffusion” model - from a theoretical and experimental research perspective - and identifying molecular key players controlling animal skin pattern formation. About twenty years ago, he could indeed convincingly demonstrate that the dynamic changes in the striped pattern of the marine angelfish Pomacanthus imperator were being unfolded, exactly as predicted by the Turing model.

Moreover, Prof. Kondo and his colleagues have recently unraveled the mechanism underlying skin pattern formation in zebrafish: a well known striped motif primarily composed of melanophores (producing a black pigment) and xanthophores (producing a yellow pigment). Observing isolated pigment cells in a Petri-dish - being actually the first research group to do so - they could demonstrate a causal relationship between different cell movements and naturally occurring skin patterns in zebrafish. Through careful in vitro analysis of the behavior of pigment cells,

Fig 3. “Run & Chase” movements of zebrafish pigment cells in vitro (left: movie; right: schematic illustration of the different steps)

isolated from either wild-type or genetic mutants with different skin patterns (e.g. leopard and jaguar) a theoretical framework could be formulated: the interaction between melanophores and xanthophores (so called “Run & Chase” movements) proceed by local activation and long-range inhibition, the essential mechanism by which Turing patterns come alive…

Intriguingly, a whole array of different skin patterns could be mimicked by constructing a series of transgenic zebrafish, carrying different small deletions in the connexin 41.8 gene, encoding a gap junction protein involved in direct cell-cell interaction. For example, loss of function of connexin 41.8 (as in the mutant leopard) reduces the spatial periodicity and changes the pattern from stripes to spots.

Connexin: bridging the gap between 2D and 3D patterning?

Zebrafish stripe formation is essentially a patterning process that occurs in a 2D field. Since many years however, Prof. Kondo has been directing a major part of his research efforts towards elucidating the mechanisms that control 3D patterning. One such project deals with understanding how the 3D shape of bone is generated. Bones come in all sizes and forms, ranging from whale vertebrae over alligator skull bones to the humerus of a colibrie or the ulna of a poison dart frog. What controls their shape? By which rules are they molded into the stupendous variety that can be observed among vertebrates?

In search for answers, Prof. Kondo and his team are studying bone growth in zebrafish. A major step forward in understanding this process has been made by focusing on the stoepsel mutant (stp), which affects normal vertebra formation. Indeed, the stp mutant has disturbed bone proportions, showing distinctly shorter vertebrae than wild-type, at later growth stages. Surprisingly, positional cloning revealed that stp encodes Connexin 43 (Cx43), thereby demonstrating that connexins are apparently major players during 2D and 3D patterning. It is important to notify that Cx43 has two independent functions: a gap junction (involved in initial bone formation) and a hemichannel activity (involved in later bone growth). Prof. Kondo has proposed a mechanism, whereby Cx43 hemichannels function in osteocytes as sensors of mechanical stress: a major factor in determining the 3D shape of a bone. In the stp mutant the hemichannel activity is abnormally enhanced - mimicking the effect of high mechanical stress - leading to increased osteoclast differentiation and enhanced rates of bone breakdown.

Unfolding the origami: lessons from the beetle horn

Fig 4. Pupa formation in the Japanese beetle (Popillia japonica) takes about 2 hours. Elongation of the horn however is completed within approximately 30 minutes.

The body shape of a vertebrate animal closely follows the size and form of the bones, which collectively form the endoskeleton. Insects on the other hand possess an exoskeleton - “the bones are outside” so to speak - and consequently must employ different patterning strategies. To unravel those mechanisms, a new research project has been started up at the Kondo lab with the aim of unmasking the principles that control morphogenesis of the beetle horn.

Many scarab beetles (Scarabaeoidea) develop horns: conspicuous skeletal outgrowths from the head - often used as weaponry during intraspecies combat over females - that come in an astonishing diversity of bizarre shapes and sometimes ridiculously exaggerated sizes. Morphogenesis of these beetle horns proceeds according to intriguing rules: they originate from localized patches of epidermal tissue that have been “kept aside” during larval development, as disc-like structures. During metamorphosis, these epidermal discs unfurl into an adult horn: a process which requires no more than 30 minutes and proceeds in the absence of any cell proliferation or major cell rearrangements.

In other words: at the end of the pupal stage, the 3D shape of the adult horn is already present, pre-patterned as a folded epithelial tissue - “hidden” in the disc, one could say - waiting to be unfolded at the appropriate developmental moment. But how is this possible? How does the tissue know into which 3D shape it has to develop, before the event has actually taken place? Which morphogenetic machinery is utilized here? Who are the instructors? And in which chapter of the Manual for Animal Development can one find the guidelines? During the next years, Prof. Kondo and his team will try to dissolve some of the clouds that befog the answers and provide explanations for this intriguing folding problem. Will it turn out to follow A Turing pattern? Or not?

A look into the Future
Science Fiction
Science or Fiction?
Prof. Kondo, in which directions are your research projects developing?
Prof. Kondo:
Our work on skin pattern formation is approaching its final stages: at present, we are in the process of identifying molecular factors that are instrumental in this process. Now, a natural and logical way to continue our research is to proceed from 2D to 3D patterning formation. As such, we are focusing on 3D patterning of the zebrafish bone and the beetle horn, which have become the spear point projects in our laboratory. In our search for unraveling the mechanisms, we will shift from the experimental towards a more theoretical approach.
But looking at patterning from a theoretical angle has always been a major research direction in your laboratory, no?
Prof. Kondo:
Certainly. We have always combined experimental and theoretical research - and we will continue to do so - but during the years ahead, I expect computer modeling to become more prominent in our lab.
Prof. Kondo:
Because we want to understand the principles behind the patterning of a biological shape. Using a computer model to simulate morphogenetic phenomena - like the unfolding of an imaginal disc into an adult beetle horn - can be a very powerful tool to recapture its essential features and understanding the basic parameters. And that is what we are aiming for: uncovering the patterning principles which guide the enormous diversity of beetle horn shapes. Ultimately, we want to know whether there is a common logic - and maybe I should say “bio-logic” - behind the diversity of 3D morphogenetic programs observed in the animal world.
Could one say that you are searching for an all-encompassing theory of 3D pattern formation, like Alan Turing elucidated the principles behind 2D patterning?
Prof. Kondo:
That is indeed our final goal: a theory that explains biological shape.
At the moment, you have not yet formulated a theory that explains 3D shaping of the beetle horn.
Prof. Kondo:
That is correct.
Why not? What are the questions you need to have answered?
Prof. Kondo:
We need to understand how these insects control position, angle, depth of the folding and bifurcation of their horns. I believe there exists a common mechanism that controls these 4 parameters in a 3D coordinate field: differences in one or more parameter settings can give rise to entirely different shapes. At the moment we are developing an algorithm that recapitulates these phenomena.
Do you expect to find a universal shaping principle?
Prof. Kondo:
I believe that the mechanisms employed by Nature to shape a biological body - how diverse and complex they may seem - are the emanation of some universal principle that can be understood in a few mathematical equations.
  1. Misu A, Yamanaka. H, Aramaki T, Kondo S, Skerrett IM, Iovine MK, Watanabe M. Two Different Functions of Connexin43 Confer Two Different Bone Phenotypes in Zebrafish. J Biol Chem. 10;291(24):12601-11. 2016.
  2. Yamanaka H. and *Kondo S. In vitro analysis suggests that difference in cell movement during direct interaction can generate various pigment patterns in vivo (Article). Proc Natl Acad Sci USA, 111, 1867-1872, 2014.
  3. Inaba M., Yamanaka H., *Kondo S. Pigment pattern formation by contact-dependent depolarization. Science, 335, 677, 2012.
  4. *Kondo S. and Miura T. Reaction-Diffusion Model as a Framework for Understanding Biological Pattern Formation. Science, 329, 1616-1620, 2010.
  5. *Kondo S. and Arai R. A reaction-diffusion wave on the skin of the marine angelfish Pomacanthus. Nature, 376, 765-768, 1995.
Who is Who?

Prof. Kondo

Prof. Shigeru Kondo

Prof. Shigeru Kondo was born in Tokyo and started his scientific career as an immunologist at the University of Kyoto. However, unraveling immunological mechanisms - a research domain with a strong link to clinical practice - did not give him intellectual satisfaction and he therefore decided to enter the field of Developmental Biology, in search for new scientific challenges. Consequently, he started a post-doctoral training at the Biocenter in Basel, Switzerland, under supervision of Prof. Walter Gehring. During that period, Prof. Kondo got intrigued by biological pattern formation, a fascination which would never leave him. As a matter of fact his research has led to a major breakthrough in the field: Prof. Kondo was the first scientist to provide convincing experimental evidence for the patterning mechanisms proposed by Alan Turing, more than half a century earlier. At present, Prof. Kondo is leading a research team at the University of Osaka, exploring the mechanisms that guide morphogenesis and patterning, from both a theoretical and an experimental perspective.