Perspective

β-Catenin, Cancer, and G Proteins: Not Just for Frizzleds Anymore

See allHide authors and affiliations

Science's STKE  12 Jul 2005:
Vol. 2005, Issue 292, pp. pe35
DOI: 10.1126/stke.2922005pe35

Abstract

The lipid metabolite lysophosphatidic acid (LPA) mediates an impressive set of responses that includes morphogenesis, cell proliferation, cell survival, cell adhesion, and cell migration. LPA exerts its downstream signaling by binding to the LPA1, LPA2, and LPA3 (formerly Edg-2, -4, and -7) family of seven-transmembrane, segmented, heterotrimeric guanine nucleotide–binding protein (G protein)–coupled receptors. LPA actions of therapeutic interest include effects on wound healing, atherogenesis, thrombogenesis, and, of course, cancer. LPA has been implicated in the progression of human breast, ovarian, prostate, head and neck, and colon malignancies. In view of these earlier observations, a recent report that LPA stimulates the proliferation of colon cancer–derived cell lines was greeted with great anticipation for its possible contribution to the unraveling of details of cancer signaling downstream of LPA. LPA was shown to stimulate nuclear accumulation of β-catenin in a manner that depended on activation of Gαq by LPA2,3, activation of phospholipase Cβ, activation of a conventional protein kinase C, and phosphorylation and inhibition of glycogen synthase kinase 3-β. The phosphorylation of β-catenin by this kinase marks the protein for intracellular degradation; LPA suppresses this degradation and stimulates β-catenin accumulation. β-catenin is a pivotal molecule in the control of cell cycle progression and gene expression, activating both processes in combination with lymphoid-enhancing factor/T cell–factor–sensitive transcription and inhibiting both processes in combination with FOXO transcription factors. The ability of LPA to increase the cytoplasmic and nuclear accumulation of β-catenin provides a new dimension of knowledge linking lipid mediators to the dysregulation of β-catenin signaling in cancer.

The lipid metabolite lysophosphatidic acid (LPA) mediates an impressive set of responses that includes morphogenesis, cell proliferation, cell survival, adhesion, and migration. LPA exerts its downstream signaling by way of the LPA1, LPA2, and LPA3 (formerly Edg-2, -4, and -7) family of seven-transmembrane, segmented, heterotrimeric guanine nucleotide–binding protein (G protein)–coupled receptors (GPCRs). From a medical perspective, LPA actions of keen interest include effects on wound healing, atherogenesis, thrombogenesis, and, of course, cancer. LPA action has been implicated in the progression of malignancies of the human breast, ovaries, prostate, head and neck, and colon. In view of these earlier observations, the recent report of Yang et al. that LPA stimulates the proliferation of colon cancer–derived cell lines was greeted with great anticipation for its possible contribution to the unraveling of details of cancer signaling downstream of LPA (1). Their discovery that LPA stimulated nuclear accumulation of β-catenin, which controls lymphoid-enhancing factor/T cell–factor family (Lef/Tcf)–sensitive gene expression, gives a new dimension to our understanding of the role of LPA in cancer, linking lipid mediators to the dysregulation of β-catenin signaling in cancer.

To appreciate the full impact of Yang et al.’s work on how we think about LPA, signaling, and cancer, we must first take a step back, into the area of early development and another family of seven-transmembrane segmented receptors known as Frizzleds (2, 3). These cell surface receptors for the secreted glycoprotein Wnt ligands derived their name from the decidedly "frizzled" appearance of the wings of fruit flies deficient in the product of one of the frizzled genes (4). Essential for normal development and for the planar cell polarity controlling differentiation of the wing, eye, and other epithelia in the fly (and humans), Frizzled-1 (and its downstream signaling pathway) were subjected to intense genetic scrutiny that mapped most of the central players—from the Wnt ligand to the phosphoprotein β-catenin that controls gene expression—in the "canonical" or Wnt/β-catenin pathway (Fig. 1) (5). Research into the prominent role of Wnt/β-catenin signaling in cell proliferation, fate, and migration led to the discovery of the many ways in which dysregulation of this pathway can lead to cancer, with β-catenin playing a principal role (6). Upstream of β-catenin is a degradation complex [consisting of Axin, the adenomatous polyposis coli gene product (APC), and glycogen synthase kinase 3β (GSK3β), among other proteins] devoted to the degradation of cytoplasmic phosphorylated β-catenin. Upstream of the degradation complex, in turn, is Dishevelled (Dvl), a phosphoprotein that inhibits phosphorylation of β-catenin by GSK3β in response to the activation of Frizzled-1 by Wnt. The path between Frizzled and Dvl is unclear, although Frizzled-1 displays a PDZ ligand in its C-terminal cytoplasmic tail, and Dvl harbors a PDZ domain (7).

Fig. 1.

Schematic of G protein–coupled signaling of LPA and Wnts. LPA activates two receptors (LPA2 and LPA3) that can activate G proteins: Gq [linked to the activation of conventional PKC (cPKC) through the activation of phospholipase Cβ (PLCβ)] and G12/13 [likely linked to activation of the JNK cascade through activation of p115RhoGEF]. LPA activation of cPKC, like the activation of Dvl by Wnt binding to Frizzled-1 (Fz1), leads to suppression of GSK3β, which phosphorylates β-catenin, targeting the phospho–β-catenin for destruction by the degradation complex. The activation of either GPCR, LPA2,3 (by LPA), or Frizzled-1 (by Wnt) leads to the accumulation of β-catenin in the nucleus. In combination with FOXO transcription factors, nuclear β-catenin can suppress cell cycle progression; in combination with the Lef/Tcf–sensitive transcription factors, nuclear β-catenin can stimulate cell cycle progression and gene expression. The Wnt/Ca2+/cGMP pathway is also G protein–coupled and signals through Go and Gt2, leading to the activation of various transcription factors, including nuclear factor of activated T cells (14). The features common to these important signaling pathways are seven-transmembrane-segmented GPCRs, coupled through heterotrimeric G proteins to downstream cascades that regulate transcription factors and gene expression. Signaling elements for GPCR-based activation of β-catenin–mediated regulation of transcription are shaded for emphasis. PM, plasma membrane.

Several factors slowed progress in elucidating Wnt/β-catenin signaling proximal to Frizzled-1: There were no data to implicate the seven-transmembrane segmented Frizzleds as members of the superfamily of GPCRs; biologically active purified Wnts were not available to foster biochemical study of Frizzled signaling; and when compared to the well-known downstream effectors of heterotrimeric G proteins such as adenylylcyclases, phosphodiesterases, and ion channels that yield rapid transient responses, β-catenin–sensitive gene expression seemed like an unlikely effector for GPCRs. Study of the noncanonical Wnt/Ca2+ pathway provided the first data indicating that heterotrimeric G proteins may be mediating Frizzleds’ effects (8). The creation of functional chimeric receptors in which the cytoplasmic domains of Frizzled-2 were spliced onto the transmembrane and extracellular domains of a GPCR demonstrated agonist-induced noncanonical Wnt signaling to Ca2+ that required the participation of Go and Gt G proteins (9). Purification of active Wnt ligands became a reality (10) and enabled the confirmation and exploitation of early observations on Frizzled signaling derived from studies with the chimeric receptors. Finally, the readout of Lef/Tcf–sensitive gene expression by β-catenin did not seem quite so farfetched when GPCRs were discovered that regulated cell cycle progression (G2A) (11) and life span (Methuselah in Drosophila) (12). It was not long before both the noncanonical Wnt pathway involving Ca2+ and cyclic guanosine monophosphate (the Wnt/Ca2+/cGMP pathway) (8, 9, 13, 14) and the canonical Wnt/β-catenin pathways (15, 16) were populated with G proteins in several model systems in development, the holdout being the fruit fly. In 2005, Drosophila joined the larger group, which includes Xenopus, zebrafish, and mouse F9 teratocarcinoma embryonic stem cells, that shows Frizzleds to be GPCRs; the fly G protein Go is essential for the canonical Wnt signaling pathway as well as for the enigmatic signaling pathway that controls planar cell polarity (17).

The evolution of our understanding of lipid mediators (such as LPA and sphingosine-1 phosphate) taxed our ability to view lipid metabolites as possible ligands for cell surface receptors. Appreciation of LPA as a key regulator of many essential cellular functions (including proliferation, survival, and motility), as well as its possible roles in development, came about slowly (18), hastened by the identification of the seven-transmembrane segmented receptors that mediate LPA effects (19). With Wnt and Frizzled signaling, we were challenged to accept the signature seven-transmembrane segmented organization of Frizzleds as possible GPCRs, and the possibility of signaling through G proteins to produce biological responses more complex than simple changes in intracellular cyclic nucleotide or Ca2+ levels (20). For the LPA story, in contrast, we were challenged by the notion of LPA as a ligand, but anticipated that LPA receptors likely would be members of the GPCR superfamily and that G proteins would play some role in LPA signaling. LPA receptors, in fact, are known to signal through multiple G proteins: Gαi family and Gα12 family members for LPA1; Gα12 and Gαq family members for LPA2; and Gαq family members for LPA3 (21). Although G proteins have been implicated as "oncogenes" when mutated, these provocative roles appear, in retrospect, to be rare (22). Yang et al. showed, however, that LPA receptors can stimulate the proliferation of colon cancer–derived cell lines, especially the HCT116 and LS174T lines (1). It is how LPA mediates this progression of cancer cells as revealed in the work of Yang et al. that really caught the attention of cancer biologists and cell-signaling mavens.

LPA2 and LPA3, but not LPA1, were shown to mediate LPA-stimulated cell proliferation in colon cancer cells, and the suppression of both receptors in HCT116 cell xenografts was shown to attenuate tumor growth in nude mice (1). These data implicate signaling mediated by Gα12/13, Gαq, or both in LPA effects on tumor growth. The ability of the Wnt/β-catenin pathway to stimulate carcinogenesis when it is dysregulated after a mutation in any one of the many signaling path components (6), as well as the activation of the β-catenin pathway in colorectal cancer (23, 24), provided a provocative and testable hypothesis about LPA’s ability to stimulate colon cancer cell proliferation. Yang et al. tested this hypothesis at several key steps and confirmed the central tenet (1): that the LPA/β-catenin signaling pathway is a likely explanation for the ability of LPA to stimulate the growth of colorectal cancer cells. The report demonstrated that LPA is capable of (i) stimulating the phosphorylation of Ser9 of GSK3β, which inhibits the enzyme’s ability to phosphorylate β-catenin; (ii) increasing the nuclear accumulation of β-catenin; and (iii) inducing Lef/Tcf–sensitive gene activation. These observations provide a compelling though incomplete description of LPA’s effect (1). The pièce de résistance was uncovering the essential role of protein kinase C (PKC) in a screen of the effects of chemical inhibitors on the ability of LPA to stimulate phosphorylation of GSK3β. This observation focused attention on the possible role of Gq in this LPA response, a role further solidified by the inability of pertussis toxin (which inactivates Gi family members but not Gq) to block LPA-stimulated phosphorylation of GSK3β.

Although they are provocative and provide a new entry point into understanding the possible connections between lipid mediators and cancer, the observations of Yang et al. leave several key questions unanswered. Because many cancers, including colorectal cancer, display mutations elsewhere in the β-catenin pathway (25), is the LPA effect on β-catenin a common or prominent element in carcinogenesis? What role does or can LPA signal propagation through Gα12/13 play in the β-catenin pathway [or, more likely, in the c-Jun N-terminal kinase (JNK) cascade (26)], and is this role important in regulating cell proliferation? Recently, β-catenin has been shown to interact functionally with FOXO transcription factors in oxidative stress signaling to inhibit cell cycle progression (27). β-catenin appears to show potential as a dual regulator in cell proliferation: In combination with Lef/Tcf, it stimulates (28), whereas in combination with FOXO it can inhibit, cell cycle progression (27). Thus, G proteins, β-catenin, and cancer are a triad of terms no longer just for Frizzleds, but will likely extend to the broader family of GPCRs.

References

  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.
View Abstract

Stay Connected to Science Signaling

Navigate This Article