Embryology - Biology 104, Spring 2006 - Albert Harris and Corey Johnson   OUTLINE OF SEVENTEENTH LECTURE: Feb 15, 2006, by Corey Johnson Developmental Patterning One problem that has provoked the curiosity of developmental biologists is how patterns arise in living creatures. A pattern might be defined as anything with some sort of symmetry. That may be too broad of a definition... I don't know. But we know one when we see it. One obvious pattern is pigmentation; zebras have a repeatable pattern of stripes, which is a form of displacement symmetry. We see segmentation in insects and other arthropods, even humans. Spots are common pigmentation patterns, as we see on leopards, cheetahs, birds, dogs, fish. Then there's the pattern we see among animals in a given species. You can identify a dog as a dog because one looks similar to the next. The anatomy of my dog is quite similar to that of another because they were patterned the same way. There's a 'dog pattern' that is produced from the embryo. Snail shells have a spiral that has predictable mathematical properties. A starfish has arms that are equally spaced around the center giving it radial symmetry. The sum of all of the genes in an animal could not possibly be sufficient to specify the location of every neuron in the brain, the location of every stripe on the back of a zebra, or coordinate the growth of a nautilus shell to mathematical precision. Rather, the embryo can utilize simple developmental rules to make patterns. These "rules" which can be seen at work in inorganic matter may just as easily work in biological organisms. Many nonliving patterns can be observed in the word. Crystals, waves in the ocean, sound waves, ripples in the sand, concentric orbits of planets, etc. Biologists and mathematicians have surmised a number of theories about how organisms might generate patterns with relatively few elements interacting in simple ways. Biological/mathematical models are a powerful and underutilized tool in embryology. Models can be used to generate hypotheses. Making predictions from a biological theory and testing those predictions is the basis for good science. Reaction-diffusion mechanisms Alan Turing was a mathematician who was instrumental in breaking the "enigma" code of the Germans during WWII. He turned his interest toward embryology and postulated what we now call "Turing mechanisms." Turing mechanisms (also called reaction diffusion mechanisms) involve two "morphogens." Others have hijacked the word imposing a strict definition upon it, but Turing intended it to be a vague term meaning "form producer." In many examples, the morphogens are diffusible molecules: Suppose that two chemicals exist, an activator (A) and an inhibitor (I) Rules: 1) "A" increases "A" and "I" a. this might be an increase in synthesis, protein activation, anything b. Synthesis/production of A and I will be proportional to A 2) "I" decreases "A" a. It might be destroyed, modified, removed, etc. b. A and I decrease in proportion to I 3) "I" diffuses faster than "B" a. "I" might have a smaller molecular weight, for example What will happen, if these 3 rules are obeyed?? Peaks and valleys of chemical concentration will form: Check out http://www.eb.mpg.de/dept4/meinhardt/periodic.html for some other neat patterns formed by simple models [Also see Dr. Harris' notes on Turing from previous years' courses. ] OK, so interesting patterns can be made by quite simple morphogen interactions. The connection comes from what the embryo does with the morphogens. What does a spot or stripe of high concentration mean for the cells of the embryo? One hypothesis is that high concentrations may activate phenotypic expression of a pigment whereas lower concentrations do not. (High concentrations could just as easily inhibit a pigment if that's the way the system is set up). http://plus.maths.org/issue30/features/dartnell/index.html The peaks of A are surrounded by higher levels of I. This has been shown to account for patterns in sea shells. Interestingly, this chemical phenomenon is not restricted to cellular levels. They can take place at an organismal/behavioral level. Ants pile up their deceased in a manner similar to this reaction diffusion mechanism. Ants laid around the circumference of a dish are piled in groups such that as a pile grows bigger, smaller piles nearby are eventually placed on the larger pile. The result: a positive feedback/activator behavior where the presence of a pile of ants causes more ants to be added to it. A trough (low concentration) of dead ants surrounds it. There's a zebrafish mutant that has the potential to be a player in a Turing mechanism. Zebrafish have 5 parallel stripes. Different mutations of a gene called 'leopard' causes various patterns to arise in concordance with computer simulation: http://www.ncbi.nlm.nih.gov/books/bv.fcgi?call=bv.View..ShowSection&rid=dbio.figgrp.88 Play with this one: http://texturegarden.com/java/rd/plugin/index.html All it takes is to change the rate of diffusion and you can create entirely different patterns. Some stable, others not. Positional Information One hypothesis about how embryos pattern themselves was proposed by Lewis Wolpert. This idea states that chemical diffusion gradients provide 'positional information' to the cells within the range of the gradients. The cells then respond by differentiating according to the information provided to it. In a pre-pattern (such as reaction diffusion models) morphogens might for example, turn on genes involved in pigmentation. Such information about whether pigment would be on or off provides a pre-pattern that is imposed on the cells which respond to the information, thereby creating the pattern. This is fundamentally different from positional information in that the information imparted is quite different in nature. Many developmental biologists will call anything positional information, something causes an organ or body part to form where it does. It may certainly impart information, but it may not necessarily be positional in nature. Molecular Clocks Vertebrate Segmentation: One example of developmental patterning is the formation of somites: somitogenesis. Somites are produced over time, in an anterior to posterior direction. A "wavefront" of an inducing substance continually regresses with the continual lengthening of the embryo. As pre-somitic mesoderm cells reach the end of the wavefront where concentrations are lowest, they begin to adopt a somite fate. The second part of the equation (where the segmentation comes in), involves a molecular clock. The somite-inducing signal is turned off and on at regular intervals so that only some of the pre-somite cells will organize into somites. This oscillating induction system combines two simple components to produces segmentation in vertebrates. Disclaimer: there are other, overlapping signals that reinforce this process. In theory, it could work on its own. How does the molecular clock work? Two genes are transcribed, whose proteins inhibit their own transcription. Negative feedback. So the activating signal is on for a period of time, then off. This Ôswitch' in turn, activates the somite inducing signals periodically. What's the difference between positional information and pre-patterning? What are some types of patterns? What properties are needed for a reaction diffusion system to work? Where else would a molecular clock be found during development?

 

back to syllabus

back to index page