One of the most intriguing problems facing modern science is to understand the genesis of organismic form. The question of how the information coded in linear DNA molecule becomes translated into a three-dimensional form has to date resisted a satisfactory answer. The point of view that we adopt following Frederick Cummings (1994) is to ask "How do we imagine it could have happened?", rather then "How did it happen?". We share an assumption that at some stage of evolution, a common strategy was evolved for the formation of multicellular organisms, which can be described as a set of generative rules or algorithms. These rules are reasonably simple when viewed on a certain higher level, and which have been elaborated extensively over geologic times.
It is more appropriate perhaps to think of this approach as constructing a "Game of Morphogenesis", but a game in which the thinly veiled motivation is to approach as closely as possible to the strategies employed by actual living organisms (Cf. Cummings, 1994). The level of formal description will be supra molecular, and even supra cellular. The extremely complicated cellular molecular machinery will be viewed as constituted to carry out the higher level mission described in this approach. Even the cell itself will be viewed as macromolecular "robot" devoted to the collective task of the highly coordinated morphogenetc movements.
Side by side with "classic" theoretical biology approach there is a growing branch of Artificial Life approach to modeling of mechanisms of development (Mjolsness et al., 1991; deBoer et al., 1992; Kitano, 1994). There are already works aimed at including simple morphogenetic models in evoluionary simulations (Mjolsness et all., 1991; Cangelosi et al., 1994; Dellaert and Beer, 1994a;b; 1996; Jakobi, 1995). The idea has been to encode rules that will themselves self-organize to produce a phenotype, as is the case of biological morphogenesis. Really, there is no direct mapping between genotype and phenotype in developing organisms. Rather, biological morphogenesis is the result of a co-operative self-regulating process that is controlled by the genome. Intrinsic properties of these developmental processes, when used in conjunction with Genetic Algorithms, is believed to lead to much more complex morphologies than those achievable with direct mapping (Dellaert and Beer, 1994a; Rocha, 1995).
Considering A-Life approach as elaboration of systems that could teach us something important and general about the natural phenomena that they model (See Mitchell & Forrest, 1995), I will concentrate attention on conclusions arising from the pattern-form interplay models. I really understand that apparently all models of biological morphogenesis are crude approximations of biological reality. However, ten year history of the approach (Since Cummings, 1985; Spirov & Boinovich, 1985; Harrison & Kolar, 1988) yields anyway some generalizations reviewed below.
There has been very much effort devoted in the past to questions of "pattern formation" and "morphogenesis". Often the geometry of a (usually flat) surface has been prescribed, and a "Turing-like" non-linear system of reacting and diffusing biochemicals have been investigating (For citations See Harrison, 1987). Most applications of reaction-diffusion approach to morphogenesis have assumed, throughout, a pattern-forming region of fixed size and shape with fixed boundary conditions. Hence, the change of shape and form (the morphogenesis proper!) which is thought to be a consequence of these primary prepatterns, is not considered. However other approaches have considered the geometric changes due to morphogenetic movements of epithelial surfaces (Gierer, 1977; Odell et al., 1981; Mittenthal and Mazo, 1983; Belintsev et al., 1987; Beloussov, 1991).
Morphogenesis, as distinct self-organizing processes, requires effective non-linear feedback between its dynamic components (Beloussov et al., 1994). This feedback should, on one hand, rapidly and precisely trace the deformation of embryonic layers caused by active mechano-chemical processes and, on the other hand, affect these very processes. Recently I (Spirov, 1992; 1993a,b) has analyzed a system of Turing-like equations coupled to geometric changes, where these geometric changes in turn couple back to alter Turing diffusion coefficients. Development of sea urchin was analyzed using this approach. It is the approach of coupling a "morphogen(s)" to geometrical changes, and having the geometrical changes in turn give rise to a change in the morphogen pattern, and so on, which is employed here, as an extension of the same approach of previous works (Spirov, 1992; 1993a,b; Cummings, 1990; 1994).
Equally with self-organization, the size regulation as the ability to adjust pattern to size over orders of magnitude is considered as a fundamental aspect of living organisms (Cummings, 1994). Multicellular-layer morphogenesis models, reviewed here, provide natural emergence of the regulations as intrinsic property of the modeling rather than an "add-on" feature. Regulations is turned out to be achieved by a "game" of pattern-formation and morphogenetic movements (Cummings, 1985; 1994). I hope the explorations as those reviewed here will contribute to regulative morphogenesis understanding, both in its own rights and as an approach to artificial morphogenesis (such as development of Dellaert & Beer's autonomous agents).
Get your own Free Homepage