Information

One widespread view about how pattern formation takes place is positional information (Wolpert 1989): cells in a territory acquire information (positional information) about their fate (or differentiation state) according to the specific concentration of a diffusible signal they receive from a source at the border of a territory. In that sense two diffusible signals (morphogens) can establish a two-dimensional coordinate system according to which cells determine their positional values. These morphogens determine the genetic program that a cell undertakes. From this view the exact spatial distribution of the morphogens is not important (provided that each cell or group of cells receives different concentrations). What is important is how this positional information is interpreted inside the cells. Originally it was even proposed that the differences in morphology among species would simply arise from different interpretation of a universal coordinate system (this idea has since been discarded). This perspective could explain how the fate of cells in a territory is determined, but not how the form of the territory can change. Morphogenetic movements are proposed to occur later as part of the interpretation of positional information (Wolpert 1989). How positional information is interpreted has not been described by these authors. Then the whole concept simply states that cells make developmental decisions according to signal concentration differences. The rest of pattern formation is relegated to the as yet unexplained interpretation of these differences. In practice, most developmental biologists use positional information just to state that cells in different places have different fates because they receive different signals or combinations of signals.

Nowadays it is well known that cells often communicate while they are engaged in morphogenetic movements. In addition, communication between cells is often reciprocal. Cells respond to received signals by sending other signals, expressing signal receptors, changing their adhesive properties (and then moving or changing shape), proliferating, dying, secreting extracellular matrix, etc. This results in a constant complex dynamic change in the position of cells and patterns of gene expression. This gives limited applicability to the positional information metaphor, even for the systems where it was first proposed: see the chick limb (Hinchliffe and Horder 1993) and Drosophila segmentation (Jaeger et al. 2004). In spite of that, this metaphor can be used as an extreme from which to exemplify how the variational approach can be used to compare developmental mechanisms. Precise signal concentration sensing can be done, as in early dorso-ventral patterning in Drosophila (Markstein et al. 2004), by having enhancers of different affinity (or different number of enhancers) for the same tran-scriptional factor in the promoter regions of different genes (if there is a molecular pathway transducing signal concentration into active transcriptional factor concentration). At least one distinct enhancer is required for any different fate choice based on morphogen concentration. By having two non-spatially overlapping sources of different morphogens, any distribution of cell types can be produced in a given two-dimensional spatial distribution of cells (three sources of different morphogens are required in three dimensions). This is because in this condition every cell receives a unique combination of concentrations of each morphogen (Figure 2.1) and then by appropriate combinations of enhancers of differential affinity any pattern can be produced. The number of different enhancers required increases by at least one with the number of cells having a state other than the default (not to be confused with the number of different cell states; Figure 2.1). Thus, a large genetic complexity is required to produce complex or large patterns. Parts in a pattern can change independently by genetic mutation only if they do not share an enhancer. In general, genetic variation is likely to produce relatively gradual pattern changes. In other words, the morphological variation produced and its relationship with genetic variation is more consistent with the neo-Darwinian view.

This mechanism does not make use of the form of the spatial distribution of signals (it is enough that each cell receives a different concentration). In contrast, there are other developmental mechanism that can use this spatial information opportunistically to produce relatively complex patterns without requiring many mutational changes and complex genetic properties. These involve simple gene networks where cells directly respond to signals by sending other signals (Figure 2.2). In that way, additional and more complex spatial distributions of signals are produced by combining the spatial distributions of several signals from sources in the previous pattern. If in addition cells are changing their location while signalling (in what has been called morphodynamic mechanisms) additional spatial distributions of signals or cells can be produced. This is because signals can be sent and received from groups of cells that have the forms possible by mor-phogenetic movements (rods, invaginations, condensation nodes, etc.; Newman and Müller 2000, Salazar-Ciudad 2006b). Essentially a larger

Figure 2.1 The diagram exemplifies the morphodynamic mechanism described in the text. In A and B the same field of cells is represented with the grey tones representing the concentrations of two signals. Signal 1 in A and signal 2 in B. Each matrix cell represents a cell. The spatial distribution of the signals is idealised (it should be more curved). In C and D a schema shows how different combinations of each signal concentration cause a cell to differentiate to black type. The interpretation in C gives rise to the pattern in E and the interpretation in D to the pattern in F.

Figure 2.1 The diagram exemplifies the morphodynamic mechanism described in the text. In A and B the same field of cells is represented with the grey tones representing the concentrations of two signals. Signal 1 in A and signal 2 in B. Each matrix cell represents a cell. The spatial distribution of the signals is idealised (it should be more curved). In C and D a schema shows how different combinations of each signal concentration cause a cell to differentiate to black type. The interpretation in C gives rise to the pattern in E and the interpretation in D to the pattern in F.

Figure 2.2 A and B as in Figure 2.1, and C as D in Figure 2.1. The interpretation in C of the signals in A and B gives rise to the pattern in E. In D it is shown how the same final pattern (in E) can be produced by gene 3 (expressed in black cells in E) being repressed where signal 1 or 2 has high concentration and active anywhere else.

Figure 2.2 A and B as in Figure 2.1, and C as D in Figure 2.1. The interpretation in C of the signals in A and B gives rise to the pattern in E. In D it is shown how the same final pattern (in E) can be produced by gene 3 (expressed in black cells in E) being repressed where signal 1 or 2 has high concentration and active anywhere else.

spectrum of forms of groups of cells and signals can be combined to produce a pattern. This does not allow the production of more patterns but the production of more complex and more diverse patterns from the same amount of genetic variation (Salazar-Ciudad and Jernvall 2004, Salazar-Ciudad 2006b). What is important is not only signal interpretation but also the spatial distributions of signals and the collective dynamic behaviour of groups of cells.

This positional information mechanism is an example of hierarchic (Salazar-Ciudad et al. 2000) morphostatic mechanism (in which morphogenetic movements happen after and because of signalling; Salazar-Ciudad et al 2003). Simulations of both kinds of mechanisms have shown that morphodynamic mechanisms do indeed produce patterns that are more complex, more distinct and related in more complex ways to genetic variation (Salazar-Ciudad and Jernvall 2004).

This has led to suggestions that morphodynamic mechanisms are more often involved in the formation of a pattern the first time it appears in evolution (Salazar-Ciudad and Jernvall 2004). This is because many more patterns easily appear by random mutation in morphodynamic mechanisms (while both mechanisms are equally likely to appear). Thus, evolution may often proceed opportunistically by first recruiting morphodynamic mechanisms. Over time, some of these mechanisms may be replaced by morphostatic mechanisms for the production of the same patterns because they allow more gradual variation and a simpler relationship between phenotype and genotype. This allows, depending on selective pressures, a faster and more efficient adaptation. Note that this is not opposed to the abovementioned baroque trends in the evolution of development because a morphostatic mechanism that can replace a morphodynamic mechanism necessarily involves many more genes and gene interactions. If new patterns are more likely to be added in later development, then morphodynamic mechanisms should be more frequent in later development (Salazar-Ciudad and Jernvall 2004).

concluding remarks

The previous examples show how evolutionary considerations (in the form of expectations about the likelihood of origin by genetic variation) can produce expectations about which competing hypothesis for developmental mechanisms is more likely for specific pattern transformations (for example the one in Figure 2.2). In the same way it has been explained how consideration of the genetic and variational properties of developmental mechanisms allows predictions about the evolution of morphology and development in different situations and stages. These inferences are modest in scope because relatively little is yet known about developmental mechanisms. This is likely to change soon. In general, this chapter advocates for the unique role of pattern formation and related concepts as a bridge to relate what is known in the different schools and in evolution and development in general. In that sense the concept of variational properties offers a definition of how hypotheses about how development functions should be tested and, at the same time, a basic conceptualisation of how different developmental mechanisms variously affect morphological evolution. The tandem genetic/variational properties, on the other hand, are used to infer ways in which development can evolve. These concepts help in explaining morphological change not only by selective arguments but also on the basis of which morphological variation is more likely to appear. Overall, these concepts aim to help give more explicit theoretical grounds for the gradual switch in evolutionary theory from selective explanations to explanations based on the interplay between selective forces and developmental capacities.

0 0

Post a comment