Dictyostelium Chemotaxis: Fascism Through the Back Door?

Dictyostelium discoideum cells undergo a major lifestyle change when they get hungry. They grow as unicellular amoebas, but when food starts to get short they aggregate into masses of up to a million cells, which in turn differentiate and sort out into mushroomshaped fruiting bodies. Aggregation and differentiation are rigorously controlled by a multitude of signalling pathways, which together cause cells to behave with amazing synchrony. Sydney Brenner is said to have remarked “Good God! Molecular Fascism!” on seeing thousands of aggregates simultaneously forming fruiting bodies [1]. Cell behaviour is regulated just as tightly during aggregation. Every few minutes, waves of movement propagate outwards from random centres. As the front of a wave passes, the cells in its vicinity lurch towards the source of the wave. Each wave thus pulls every cell it passes towards the centre, providing a simple mechanism to coordinate aggregation; the process is so efficient that only a few tens of waves are needed to convert a homogenous lawn of cells into discrete aggregates of as many as 106 cells. In Dictyostelium, the waves of movement are choreographed by cyclic AMP (cAMP). One cell at the centre emits a pulse of cAMP. Cells that receive a cAMP signal also emit cAMP themselves, which reinforces the signal and can allow a single wave to persist for several centimetres. The surrounding cells move by chemotaxis towards the source of the cAMP, generating the wave of movement. This highly sensitive, synchronized chemotaxis is a boon for experimental work — aggregating Dictyostelium cells are essentially specialized cAMP chemotaxis machines, which has lately made Dictyostelium the favourite organism of the chemotaxis field. About halfway through the process of aggregation, the cells’ behaviour alters — they elongate, join up end to end and form thick streams which allow yet more rapid and well-coordinated movement. Most researchers have hitherto believed that this behaviour is controlled by specialised cell–cell adhesions, but a recent study [2] suggests a new view. Kriebel et al. [2] have shown that adenylyl cyclase A (ACA), the enzyme which makes the cAMP used in aggregation, is enriched at the rear of polarized cells. This implies that the cAMP signals in streaming cells are far more spatially defined than previously thought. It appears that streaming cells are not responding to broad cAMP waves sweeping past them, as they see early in aggregation, but rather to a localised signal from the posterior end of the cell in front. These localised signals cannot be shown directly, unlike the earlier, broader cAMP waves, which were revealed in a gorgeous experiment using isotope dilution some years ago [3]. As well as being built on a very small scale, the localised signals are extremely short lived, as cells secrete large amounts of a very active phosphodiesterase [4]. Kriebel et al. [2], however, have demonstrated their presence in several ways. The clearest was by following chemotaxis in ACA null mutants. Lack of ACA does not seem to affect cells’ ability to do chemotaxis, but it does change the pattern: if a microneedle full of cAMP is held near a plate of ACA null cells, they all orient towards the stimulus (Figure 1). Wild-type cells are more subtle: they orient in a curved, fan-shaped pattern, with more distant cells pointing towards other cells nearer the needle, rather than the needle itself (Figure 1). Presumably in streams, where the cell density is far greater, each cell’s responses are almost exclusively towards the rear part of the cell in front. In other words, streaming cells are not armies marching to the decree of a few molecular fascists, but more like a long line of dogs sniffing one another’s bottoms as they walk in the park (as a crude scientist remarked at a recent meeting). The discovery of localised cAMP secretion is remarkable for a number of reasons. It emphasises one of the basic tenets of cell biology: cells are not homogenous blobs, and to understand how they work, we shall have to think architecturally. Secondly, it points to an unknown set of molecular signposts defining the back of a cell. The chemotaxis field is just starting to understand how the front of the cell is defined. Inositol lipids with 3-phosphates, in particular phosphatidylinositol 3,4,5-trisphosphate (PIP3), seem to play a fundamental role in specifying the leading edge [5]. The shape of polarized cells is maintained by a feedback loop which allows PIP3 to stimulate its own production, so leading edges tend to remain constant [6]. Localised sources of chemoattractant can cause cells to repolarize, apparently by creating a PIP3 signal which exceeds the capacity of the feedback loop to maintain itself, allowing cells to turn up chemoattractant gradients. But this offers no hints as to how the back is defined. There are two basic problems. Firstly, what is the signpost? It is hard to see how the absence of PIP3 could act as a signal to localise proteins. Secondly, how would ACA use the signpost, whatever it may be? Moving a protein with twelve transmembrane spans is not as straightforward as relocalizing a soluble protein via a PH domain. The first problem is perhaps the hardest to deal with. When Kriebel et al. [2] found that the positional localisation of ACA was important, they must have licked Dispatch Current Biology, Vol. 13, R353–R354, April 29, 2003, ©2003 Elsevier Science Ltd. All rights reserved. DOI 10.1016/S0960-9822(03)00274-4