PerspectiveCell Migration

Steering in Quadruplet: The Complex Signaling Pathways Directing Chemotaxis

See allHide authors and affiliations

Science Signaling  03 Jun 2008:
Vol. 1, Issue 22, pp. pe26
DOI: 10.1126/scisignal.122pe26


Studies in the social amoeba Dictyostelium discoideum reveal that signaling cascades coordinating chemotactic directional sensing and migration are complex, with redundant pathways emerging as cells differentiate. Lack of accumulation of the leading-edge marker phosphatidylinositol-3,4,5-trisphosphate can be compensated by a pathway containing phospholipase A2 (PLA2) in early developed cells and guanylyl cyclase (GC) in later developed, polarized cells. Because numerous signaling networks operational during Dictyostelium chemotaxis are conserved in mammalian cells, PLA2 and GC pathways may also be effective in higher eukaryotes, providing avenues for future research.

Chemotaxis, the directed migration of cells along external chemical gradients, is a vital component of numerous biological processes, including immune responses, neuronal growth, and embryogenesis. It is characterized by the ability of cells to transduce shallow external gradients of chemoattractants into highly polarized intracellular events where polymerized actin (F-actin) is primarily enriched at the front of the cell for propulsion and myosin II is assembled at the back for contraction and retraction. Scientists have long searched for the biochemical signaling pathways that act as an internal compass for chemotaxing cells. A persistent candidate has been a pathway mediated by phosphatidylinositol-3,4,5-trisphosphate (PIP3), which is highly enriched at the leading edge (front) of chemotaxing Dictyostelium discoideum and mammalian neutrophils and mediates the recruitment of a subset of pleckstrin homology (PH) domain–containing proteins to the plasma membrane (1). These proteins in turn act as nucleation factors to spatially activate various effectors and to polarize responses. However, research now suggests that PIP3 is not the sole steering wheel in cells—under certain conditions, both Dictyostelium and neutrophils migrate directionally in the absence of PIP3 production (2). Indeed, a pathway mediated by phospholipase A2 (PLA2) has been suggested as a parallel compass in Dictyostelium (3, 4).

Veltman et al. now report that multiple compasses are available to Dictyostelium depending on the degree of cell differentiation and polarity (5). Chemotaxis in Dictyostelium is often studied under starvation conditions that initiate a developmental program that causes solitary cells to migrate toward chemoattractants to form tight aggregates that mature into a multicellular structure (6). Starved cells that contain phosphatidylinositol 3-kinase (PI3K) or PLA2 move up an external chemical gradient with normal directionality. However, when both the PI3K product PIP3 and PLA2 are absent from early developed cells (after ~5 hours of starvation), the directional response of the cells is strongly impaired, suggesting that they lack an internal compass (3). Veltman et al. show that if cells deficient in PI3K and PLA2 activities are allowed to progress further in their developmental program, until they adopt an intrinsic and stable polarization (after ~7 hours of starvation), an additional compass component emerges and normal chemotaxis is recovered.

To identify other pathways involved in controlling chemotaxis in polarized cells, the authors screened a series of mutants for chemotaxis defects in the context of the inhibition of both PI3K and PLA2. Under these conditions, removing the soluble guanylyl cyclase (sGC) blocked the chemotactic ability of late developed cells. The primary role of sGC is to synthesize the second messenger guanosine 3′,5′-monophosphate (cGMP, also known as cyclic GMP), which promotes the formation of myosin filaments in the posterior of Dictyostelium (7). sGC also showed an intriguing colocalization with newly polymerized actin (F-actin) at the front of the cell. Interestingly, although cells expressing catalytically deficient sGC mutants migrated slowly and had more lateral pseudopods compared to wild-type cells, they showed normal directionality. Conversely, cells that expressed a mutant sGC that does not colocalize with F-actin showed impaired directionality. These findings suggest that sGC contributes to directional sensing in a cGMP-independent manner. With this work, Veltman et al. have now identified a third pathway contributing to Dictyostelium’s compass. In addition, because the chemotaxis defects exhibited were dependent on the extent of polarization, this study also identifies cell polarity as a contributing factor to directed cell migration.

Signals That Work in Parallel with PIP3

Whereas the signals leading to the activation of PI3K have been extensively studied and are now established, how PLA2 and GC transduce chemical gradients is much less understood. Phospholipases are enzymes that cleave phospholipids (PLs) into fatty acids and lipophilic compounds. PLA2 specifically cleaves the second acyl chain of PLs to produce fatty acids, predominantly arachidonic acid (AA) and lysoPL (8). In mammalian cells, AA is converted into the eicosanoids leukotrienes and thromboxanes through the action of lipoxygenases and cyclooxygenases, respectively (9). LysoPL is a precursor of various mediators, such as lysophosphatidic acid (LPA) and platelet activating factor (PAF) (10). Eicosanoids, LPA, and PAF mediate their biological effects by binding to specific heterotrimeric GTP-binding protein (G protein)–coupled receptors (GPCRs). In addition, it has been suggested that AA regulates store-operated Ca2+ channels in mammalian cells (11). Mammalian PLA2’s are classified into five distinct categories, depending on their distribution (secreted, cytosolic, or lysosomal), sensitivity to Ca2+, or activity against PAF derivatives (8). The Dictyostelium PLA2 that is involved in regulating chemotaxis is part of the Ca2+-insensitive family of PLA2’s and is one of 14 such enzymes in the genome of this organism. No secreted or cytosolic PLA2’s have been identified in Dictyostelium (3). Chen and colleagues have shown that chemoattractants stimulate the rapid and transient production of AA derivatives in Dictyostelium, presumably independently of Ca2+, but the pathways leading to PLA2 activation have yet to be determined (3). Although eicosanoids have not been identified in Dictyostelium, PAF and AA have been shown to affect Ca2+ influx in this organism (12, 13). Remarkably, studies performed in monocytes have shown that the Ca2+-insensitive PLA2 (iPLA2β) and the cytosoloic PLA2 (cPLA2α) are distributed to distinct cellular compartments and selectively affect directionality and speed, respectively (14). It appears that PLA2, much like PI3K, may represent a conserved pathway to regulate eukaryotic chemotaxis.

GCs catalyze the conversion of guanosine triphosphate (GTP) into cGMP. There are two forms of GCs: membrane-bound (mGC) and soluble (sGC). In mammalian cells, heterodimeric sGCs mainly operate as nitric oxide (NO) receptors to mediate vasodilation. The seven mGCs identified in the human genome all consist of a single transmembrane domain, with an N-terminal extracellular domain and a C-terminal cytosolic domain that harbors the catalytic domain—although they function as homodimers (15). Although some mGCs are activated by external stimuli, their activity is also modulated intracellularly by calcium and adenine nucleotides. In Dictyostelium, two GCs are expressed: a 12–transmembrane domain GCA and a sGC (16). Interestingly, the sGC in Dictyostelium shares homology with mammalian soluble adenylyl cyclases, containing two cyclase catalytic domains and a long C-terminal domain (17). In Dictyostelium and neutrophils, addition of chemoattractants leads to the synthesis of cGMP, which, at least in Dictyostelium, occurs downstream of G proteins—although the precise pathway that leads to GC activation remains undetermined in both systems (16, 18). As presented by Veltman and colleagues, sGC is the key GC that regulates chemotaxis in Dictyostelium. Whereas some of the effects of sGC were apparently cGMP-independent, the cGMP-dependent regulation of myosin II assembly at the back of cells was mediated through the cGMP-binding protein GbpC, which harbors two cGMP-binding domains as well as Ras, mitogen-activated protein kinase kinase kinase (MAPKKK), and Ras guanine nucleotide exchange factor (GEF) domains (7); no cGMP-dependent protein kinases have been found in the Dictyostelium genome. The link between GCs and neutrophil chemotaxis remains controversial, and was suggested by studies using NO (19). However, neutrophils isolated from mice lacking the type I cGMP-dependent protein kinase (cGK1) are now known to exhibit enhanced chemotaxis (19)—a defect that could be linked to effects on vasodilator-stimulated phosphoprotein (VASP), which is regulated by cGK1 (20). Interestingly, a novel Rac to cGMP signaling pathway involving p21-activated kinase (PAK) as a direct activator of mGCs has been identified in migrating fibroblasts and is involved in platelet-derived growth factor (PDGF)–induced migration and the formation of lamellipodia (21). Clearly, more work is required to fully understand the ramifications of these complex signaling pathways.

Polarity in the Absence of PIP3 Production?

A stably polarized cell has distinct advantages during directed cell migration. First, cell elongation in a chemical gradient enhances the differences in receptor occupancy between the cell front and back, thereby increasing the detected signal-to-noise ratio (22). Second, Andrew and Insall showed that new actin-rich extensions, or pseudopods, most often emerge from existing pseudopods (23). This biased formation of pseudopods enables the persistent localization of proteins, including sGC (24) and the small G proteins Ras and Rac (25, 26), to the front of the cell. Interestingly, and in support of experimental data, a theoretical investigation aimed at understanding the way chemotactic information is processed suggests that a strategy involving a preexisting cell bias is more effective than one in which sensitivity is equally distributed in all directions (27). However, our understanding of cell polarity during chemotaxis is still mostly descriptive (Fig. 1).

Fig. 1.

Adaptive versus persistent polarity during chemotaxis. Early developed Dictyostelium cells respond quickly to a changing chemoattractant gradient (from top left to bottom center) by retracting the existing leading edge and initiating a new pseudopod pointing toward the new gradient (left). Conversely, later developed cells (right) execute a U-turn, maintaining the same pseudopod throughout the motion (from top right to bottom center). Dictyostelium’s early compass allows for a fast response to changing gradients but is sensitive to the inhibition of PI3K and PLA2. The later compass responds more slowly to changing chemoattractant conditions but is more robust to perturbations in the signaling networks (33). The biochemical markers that facilitate the onset of persistent cell polarity remain unknown; however, disruption of actin or myosin assembly returns a persistent polarized cell to a less polar state (29, 34).

Polarity requires an intact cytoskeleton. Polarized cells are more elongated than are nonpolar cells, the posterior of the polarized cell is desensitized to external gradients, and there is a tendency for anterior- and posterior-associated proteins and structures to remain at their respective poles (22). However, little is known about the proteins that govern the onset of stable cell polarity, and the mechanisms that allow cells to sustain differences between the anterior and posterior of the polarized cell are incompletely understood. In addition, the cascades known to be associated with cell polarity are PI3K-dependent. In both Dictyostelium and neutrophils, PIP3 production leads to the localization of front-defining proteins such as Rac and, through second messengers, the establishment of posterior contractile myosin, albeit by distinct pathways: cGMP (7) and PAKa (28) in Dictyostelium and RhoA in neutrophils (29). Importantly, we do not know how a cell achieves polarity in the absence of PIP3 production.

A polarized cell requires efficient communication between its anterior and posterior regions because regulators and their downstream targets are often spatially separated. For instance, protein kinase B (PKB), which is localized at the leading edge of Dictyostelium, phosphorylates and activates PAKa, which is enriched at the back of cells (28). In neutrophils, RhoA activation at the posterior requires inputs from anteriorly localized Cdc42 and the heterotrimeric G proteins G12 and G13 (30). Signal-relay pathways also require long-range regulation. For instance, the cytosolic regulator of adenylyl cyclase (CRAC) translocates to the leading edge of Dictyostelium and is required to activate adenylyl cyclase, which is highly enriched at the posterior (31). Small molecules rapidly diffuse through the cytosol such that the size of the cell is not a huge barrier to its localized function. However, large molecules or membrane-associated proteins require additional mechanisms, such as active transport, to mediate anterior-to-posterior communication. Interestingly, PIP3 and its targets are strongly associated with organizing the actin cytoskeleton (32). As more is discovered about how PLA2 and sGC regulate chemotaxis, it will be interesting to compare how each component of the cell’s compass arranges the cytoskeletal network, establishes cell polarity, and allows for anterior-to-posterior communication.

Evolution has yielded a robust and complex steering wheel for migrating eukaryotic cells. Multiple signaling pathways act in parallel to transduce external chemical gradients into internal signals that span the length of the cell. With the discovery of additional compass components, we are left with new questions: To what extent do these different signaling pathways regulate chemotaxis? What additional advantages does a cell gain from using parallel signaling pathways? How do signaling mechanisms depend on the extent of cell differentiation and external environment, and vice versa? Although the answers to these and other questions will surely reveal more complexity, they certainly represent an exciting, yet challenging direction for future research.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
View Abstract

Navigate This Article