Moving in the right direction: how eukaryotic cells migrate along chemical gradients

Abstract

Many cells have the ability to grow or migrate towards chemical cues. Oriented growth and movement require detection of the external chemical gradient, transduction of signals, and reorganization of the cytoskeleton. Recent studies in Dictyostelium discoideum and mammalian neutrophils have revealed a complex signaling network that enables cells to migrate in chemical gradients.

Fig. 1

Dictyostelium and neutrophils migrate along chemoattractant gradients. Left panel: Dictyostelium cells were placed on a chambered cover glass and exposed to a cAMP gradient generated by a micropipette. Right panel: Neutrophil-like HL-60 cells were placed on fibronectin-coated surface and exposed to an fMLP gradient.

Fig. 2

Chemotaxis is composed of motility, directional sensing, and polarity. Motility involves the periodic extension and retraction of pseudopodia that drive cellular movement in the absence of chemoattractant gradients. Directional sensing is demonstrated by the gradient-mediated relocalization of proteins in cells that have been immobilized by treatment with inhibitors of actin polymerization. Polarity is evident in an elongate cell shape as well as the asymmetric distribution of the cytoskeletal components and signaling molecules. These processes have overlapping, but distinct, properties as described in the text.

Fig. 3

Fluorescent images depicting the localization of signaling components. Top panel: In migrating Dictyostelium cells, the cAMP receptor cAR1 and Gβ subunit distribute evenly along the plasma membrane. Middle panel: PI3K is recruited to the leading edge whereas PTEN localizes to the lateral sides and the back. The reciprocal distribution of the two enzymes creates an internal PIP3 gradient as indicated by the localized accumulation of the PH domain from CRAC. Bottom panel: In some cases, although the proteins have a uniform distribution on the plasma membrane or in the cytosol, their activities are spatially restricted. Ras proteins and PKBR1 are activated at the cell's leading edge, as reflected by the localization of RBD and the phosphorylated form of PKBR1. Actin polymerization also enriches at the leading edge as indicated by the localization of the actin binding protein coronin.

Fig. 4

Signaling networks control chemotaxis. Depicted is a network of signaling events triggered by cAMP to control front projection and back contraction in Dictyostelium. At the front, binding of cAMP to GPCR leads to the activation of RasG and RasC, which in turn stimulate the activities of PI3K and TORC2, respectively. PI3K produces PIP3, which recruits PH-domain containing proteins including PKBA, CRAC and PhdA. PDK [86] and TORC2, composed of PiaA, Rip3, Lst8, and the Tor kinase, mediate the phosphorylation and activation of PKBA and PKBR1. PLA2 acts in parallel with the PIP3 pathway to regulate actin polymerization. Front signals also inhibit myosin II activity through the activation of the myosin heavy chain kinase (MHCKA) [87, 88]. At the back, PTEN is responsible for the degradation of PIP3. Myosin II is assembled into contractile filaments that suppress pseudopod formation and promote back retraction. The cGMP binding protein GbpC promotes the assembly and activity of myosin II. Positive links between components are indicated by (→) or dotted arrows (less defined steps) and inhibitory links are indicted by (⊥).

Fig. 5

Chemotactic responses and the LEGI-BEN model. (A) Top: Schematic of a biphasic response triggered by a global stimulation. Bottom: A biphasic response of PIP3 production in cells expressing GFP-labeled PH_crac (B) In the LEGI model, chemotactic stimuli generate a local excitor (LE), which rises faster than the global inhibitor (GI). The excitor reflects the local receptor occupancy, whereas the inhibitor depends on the averaged receptor occupancy. When a gradient is applied, there is an initial response (red line). However, when it reaches a steady state, chemotactic response persists at the front of the cell (red line pointed by the two arrows) because excitation (grey lines) exceeds inhibition (blue line) at the front while it is lower at the back. (C) In the LEGI-BEN model, the LEGI module controls a response regulator, RR, which is a positive driver of an excitable network triggered by stochastic noise. (D) Kymograph of a one-dimensional simulation produced by implementation of the LEGI-BEN hypothesis. The kymograph is a plot of Y. Before stimulation (from 0 s to 180 s), patches of activity are at random sites around the cell perimeter. When a gradient stimulus is applied at ∼180 s, activities appear all along the perimeter producing an initial uniform response that quickly “shuts-off” and becomes localized towards the high side of the gradient. When the gradient is repositioned at ∼660 s, a second uniform response is produced followed by a persistent directional response in the new direction. It should be noted that the excitable network is triggered by stochastic noise. If noise crosses threshold, it will trigger a response. Therefore signals sometimes appear at different points on the cell perimeter at the same time. For the same reason, even when most responses are biased to occur on the side of the cell facing the gradient, some events can take place by chance away from the gradient. These simulations are consistent with observations of real cells.