Elsevier

Cellular Signalling

Volume 26, Issue 2, February 2014, Pages 260-267
Cellular Signalling

β-Catenin-dependent pathway activation by both promiscuous “canonical” WNT3a–, and specific “noncanonical” WNT4– and WNT5a–FZD receptor combinations with strong differences in LRP5 and LRP6 dependency

https://doi.org/10.1016/j.cellsig.2013.11.021Get rights and content

Highlights

  • Six out of nine WNT3a/FZD chimeric proteins activate β-catenin-dependent signalling.

  • WNT3a/FZD induced signalling strongly dependent on LRP6 but not on LRP5 as co-receptor.

  • WNT4/FZD10 and WNT5a/FZD4 in combination with LRP6 show strong specific signal.

  • LRP5 acts as moderate co-receptor only for FZD10 in combination with WNT4 or WNT5a.

Abstract

The WNT/β-catenin signalling cascade is the best-investigated frizzled receptor (FZD) pathway, however, whether and how specific combinations of WNT/FZD and co-receptors LRP5 and LRP6 differentially affect this pathway are not well understood. This is mostly due to the fact that there are 19 WNTs, 10 FZDs and at least two co-receptors. In our attempt to identify the signalling capabilities of specific WNT/FZD/LRP combinations we made use of our previously reported TCF/LEF Gaussia luciferase reporter gene HEK293 cell line (Ring et al., 2011). Generation of WNT/FZD fusion constructs – but not their separate transfection – without or with additional isogenic overexpression of LRP5 and LRP6 in our reporter cells permitted the investigation of specific WNT/FZD/LRP combinations. The canonical WNT3a in fusion to almost all FZDs was able to induce β-catenin-dependent signalling with strong dependency on LRP6 but not LRP5. Interestingly, noncanonical WNT ligands, WNT4 and WNT5a, were also able to act “canonically” but only in fusion with specific FZDs and with selective dependence on LRP5 or LRP6. These data and extension of this experimental setup to the poorly characterized other WNTs should facilitate deeper insight into the complex WNT/FZD signalling system and its function.

Introduction

The WNT/FZD signalling pathways play pivotal roles in developmental processes such as cell proliferation, differentiation and migration [1]. The importance of this signalling becomes obvious when it is dysregulated, leading to many different, mostly severe pathologies, like osteoporosis, Alzheimer's disease or cancer [2], [3], [4]. In humans, there are 19 WNT ligands, ten frizzled receptors (FZDs) and at least two co-receptors, LRP5 and LRP6 (low-density lipoprotein receptor-related proteins 5 and 6), making the WNT/FZD signalling system highly complex. The so far best investigated WNT/FZD pathway depends on the regulation of the β-catenin level, whereby stabilized β-catenin accumulates in the cytosol, then translocates to the nucleus and binds to the T cell factor and lymphoid enhancer factor (TCF/LEF) family of transcriptional cofactors, thus activating the transcription of target genes, e.g. c-myc, cyclin D1, and axin2 [5], [6], [7], [8]. According to their biological activity WNTs have been historically classified into two different groups: the canonical WNTs, such as WNT1, WNT3a and WNT8, that are able to induce the β-catenin-dependent pathway and are mostly essential for embryonic development, and the noncanonical, i.e. supposedly β-catenin independently signalling WNTs, such as WNT4, WNT5a and WNT11, that participate in developmental processes such as planar cell polarity and convergent extension movements during gastrulation [9], [10], [11]. The most accepted current model is that canonical WNT ligands nucleate the formation of a physical complex with FZDs and LRPs that activates the β-catenin-dependent pathway, and that WNTs are the component most important for the signalling outcome [12], [13]. The ten FZDs form their own separate family of G protein-coupled receptors [14] although there have been only a few reports on their competence to activation of G proteins [15], [16], [17]. They share a conserved extracellular cysteine-rich domain (CRD), which has been identified as the binding site for WNT ligands [18], [19]. The most sequential diversity among FZDs is located in their intracellular C-termini that harbour, however, a KTxxxW domain, which presumably is the most important transducer of the WNT/β-catenin pathway [20], [21]. But to which extent and how the secreted WNT glycoproteins that also contain a highly conserved cysteine-rich domain [22] discriminate between these receptors are still not well understood. Furthermore the relatively unknown functions and specificities of LRP5 and LRP6 or of ROR1 and ROR2 as co-receptors contribute to the challenge to identify the role of specific WNT/FZD/co-receptor interactions in the induction of the β-catenin signalling cascade [23], [24]. Considerable work has been done that resulted in reports on the function of certain canonical or noncanonical WNTs with selected FZDs with regard to β-catenin-dependent and independent signalling, investigating in part also the dependency on the co-receptors LRP5 and LRP6 [25], [26], [27], [28]. Still, as these results were often obtained with completely different experimental background, little directly comparing information about the specificities of WNT/FZD/co-receptor combinations is available, although, for example the prototype of noncanonical WNTs, WNT5a is suggested to transduce not only β-catenin-independent but also β-catenin-dependent pathways relying on the (co-)receptor context [29], [30]. A systematic study that might also be transferred to the many less well-characterized members of the WNT family should therefore be of great general interest. In order to perform such systematic investigations we have generated a TCF/LEF Gaussia luciferase reporter gene HEK293 cell line that also allowed the regulatable isogenic overexpression of the co-receptors LRP5 and LRP6 through the use of the Flp-In TRex-system [31]. In the present study we determined the effects of the combinations of all FZDs (with exception of FZD3) with three prototypically canonical or noncanonical WNTs (WNT3a; WNT4 & WNT5a) – either expressed separately or as fusion proteins – on β-catenin-dependent signalling. We also investigated the effect of an overexpression of the co-receptors LRP5 and LRP6 on the signalling of these combinations.

This experimental setup allowed us to identify specific WNT/FZD combinations with strong differences in their dependency on the co-receptor LRP5 and LRP6 in activating the WNT/β-catenin-dependent pathway. The generation of WNT/FZD fusion proteins as compared to the separate transfection of respective WNTs and FZDs, proved the best approach to characterize the properties of specific combinations. With efficient tools like the available reporter gene cell lines and WNT/FZD fusion proteins it is possible to study further specific WNT/FZD combinations, especially by extending this approach also to the less-well characterized WNT ligands, for a better understanding of their role and therapeutic potential in the complex WNT/FZD/co-receptor system.

Section snippets

Reagents

Flp-In T-Rex HEK-293 cells were obtained from Life Technologies (Carlsbad, CA, USA). Cell culture media and additions were from PAA Laboratories (Coelbe, Germany). EcoTransfect was purchased from OZ Biosciences (Marseille, France). Poly-d-lysine and EZview™ Red anti-myc affinity matrix were purchased from Sigma-Aldrich (Taufkirchen, Germany). Fugene HD reagent and EDTA-free protease inhibitor cocktail were from Roche (Mannheim, Germany). G418-BC sulphate was obtained from Biochrom AG (Berlin,

Results

WNT-induced inhibition of constitutive β-catenin degradation results in its cytosolic accumulation followed by translocation into the nucleus where it acts as a transcription factor promoting TCF/LEF dependent transcription of respective target genes [2], [5]. Recently, we demonstrated the functionality of TCF/LEF reporter HEK293 cells (later addressed as reporter cells) in combination with the Flp-In-TRex expression system (Life Technologies) for stable, but regulatable protein expression as

Discussion

In this study we set out to systematically investigate the specificity of the interaction of WNT ligands with FZD receptors with regard to β-catenin-dependent signalling. In addition, we wanted to assess in direct comparison the contributions of the co-receptors LRP5 and LRP6 that in general are not dealt with separately and therefore often only one or the other has been considered in respective studies [37], [38], [39]. We took all 10 human FZDs (with exception of FZD3, that unfortunately was

Conclusion

To summarize, the results of our study using three different WNT ligands and nine FZD receptors suggest that canonical WNT ligands are much more promiscuous in their ability to activate the β-catenin-dependent pathway via different FZD receptors, an action in which they are strongly supported by LRP6 but not LRP5 as co-receptor. Noncanonical WNTs can also induce β-catenin-dependent signalling, require, however, a much more selected setting with regard to FZD subtype and LRP5 or LRP6. Our

Acknowledgements

The study was supported by the European Research Council (ERC AdG °249929 to C.W.) and Deutsche Forschungsgemeinschaft (FOR809 to C.W.).

The authors thank C. Seidl for excellent technical assistance.

References (53)

  • H. Clevers

    Cell

    (2006)
  • H.Q. Wang et al.

    Biochem. Biophys. Res. Commun.

    (2012)
  • B.T. MacDonald et al.

    Dev. Cell

    (2009)
  • S.L. Holmen et al.

    Biochem. Biophys. Res. Commun.

    (2005)
  • V.L. Katanaev et al.

    Cell

    (2005)
  • A.S. Nichols et al.

    Cell. Signal.

    (2013)
  • H.Y. Wang et al.

    Cell. Signal.

    (2006)
  • E. Bourhis et al.

    J. Biol. Chem.

    (2010)
  • F. Verkaar et al.

    Cell. Signal.

    (2009)
  • I. Perobner et al.

    Int. J. Biochem. Cell Biol.

    (2012)
  • P. Bernard et al.

    Int. J. Biochem. Cell Biol.

    (2007)
  • J. Mao et al.

    Mol. Cell

    (2001)
  • T. Golan et al.

    J. Biol. Chem.

    (2004)
  • A. Mikels et al.

    J. Biol. Chem.

    (2009)
  • W. Lu et al.

    Cell

    (2004)
  • S.L. Holmen et al.

    J. Biol. Chem.

    (2002)
  • H. Yamamoto et al.

    Dev. Cell

    (2006)
  • V. Bryja et al.

    Cell. Signal.

    (2007)
  • T.D. King et al.

    Cell. Signal.

    (2012)
  • C.Y. Logan et al.

    Annu. Rev. Cell Dev. Biol.

    (2004)
  • A. Gurney et al.

    Proc. Natl. Acad. Sci. U. S. A.

    (2012)
  • B. Lustig et al.

    Mol. Cell. Biol.

    (2002)
  • T.C. He et al.

    Science

    (1998)
  • A. Novak et al.

    Cell. Mol. Life Sci.

    (1999)
  • G.T. Wong et al.

    Mol. Cell. Biol.

    (1994)
  • S.J. Du et al.

    Mol. Cell. Biol.

    (1995)
  • Cited by (0)

    View full text