The evolutionary transition from simple chordate body plans to complex vertebrate body plans was driven by the acquisition of the Neural Crest, a unique stem cell population with broad, multi-germ layer developmental potential (Le Douarin and Kalcheim, 1999; Hall, 2000; Bronner and LeDouarin, 2012; Prasad et al., 2012). At gastrula and neurula stages, Neural Crest cells are found within the presumptive ectoderm at the neural plate border, and they will ultimately contribute to ectodermal derivatives, including much of the peripheral nervous system. However, despite their site of origin, neural crest cells also contribute many mesodermal cell types to the body plan, including cartilage, bone, and smooth muscle, and also make contributions to otherwise endodermal organs such as the thyroid (Le Douarin and Kalcheim, 1999). Until recently, models for how neural crest cells acquire their remarkably broad potential proposed that inductive interactions orchestrated by BMP, FGF, and Wnt signals endowed these cells with greater potency than the cells they were derived from developmentally or evolutionarily (Huang and Saint-Jeannet, 2004; Taylor and LaBonne, 2007; Prasad et al., 2012; Rogers et al., 2012; Stuhlmiller and García-Castro, 2012a). Such a mechanism conflicts, however, with the generalized view of embryonic development as a progressive restriction of developmental potential.

We recently demonstrated that much of the regulatory network which controls the pluripotency of blastula inner cell mass/animal cap cells is shared with neural crest cells, shedding new light on the origins of the neural crest cells and the evolution of vertebrates (Buitrago-Delgado et al., 2015). Indeed, we found that factors that have long been considered neural crest potency factors, such as Snail1 (Taylor and LaBonne, 2007) and Sox5 (Nordin and LaBonne, 2014), are expressed earlier, in blastula animal pole cells, and are required for their pluripotency. Together, these findings suggest that neural crest cells arise through retention of the transcriptional circuitry that controls the pluripotency of the blastula cells they are derived from, avoiding the lineage restriction that characterizes neighboring cells (Buitrago-Delgado et al., 2015; Hoppler and Wheeler, 2015; Le Douarin and Dupin, 2016). This revised model raises fundamental questions about how the cells that will become the neural crest escape lineage restriction in order to maintain broad developmental potential, and how this relates to signals that have previously been implicated in the genesis of these stem cells. For example, BMP signaling has been found to play an essential role in the pluripotency of both blastula stem cells and neural crest cells (Ying et al., 2003; Kléber et al., 2005; Nordin and LaBonne, 2014), together with Sox5, which directs the target specificity of BMP R-Smads in both cell types (Nordin and LaBonne, 2014). It will be important to build on these insights and further delineate the roles of other signaling pathways in the retention of pluripotency.

FGF signaling is used reiteratively throughout embryonic development to pattern multiple tissue types and germ layers. While FGF signaling has a well established role in the formation of mesoderm, it has also been linked to the formation of the neuroectoderm/neural plate, as well as to anterio-posterior patterning of the CNS (Slack et al., 1987; Amaya et al., 1991; Xu et al., 1997; Hongo et al., 1999; Hardcastle et al., 2000; Ribisi et al., 2000; Fletcher et al., 2006; Dorey and Amaya, 2010; Wills et al., 2010). While anterior neural induction mediated by BMP antagonists can occur independent of FGF signaling, FGFs clearly play a role in posterior neural development (Wills et al., 2010). Importantly, FGF signaling has also been implicated in the establishment of both neural crest stem cells and the neural plate border region more generally (Mayor et al., 1997; LaBonne and Bronner-Fraser, 1998; Monsoro-Burq et al., 2003; Hong et al., 2008; Stuhlmiller and García-Castro, 2012b; Yardley and García-Castro, 2012). Although the precise role of FGF signaling in the establishment of these cell populations relative to BMP or Wnt signals is not fully understood, at least some enhancers for neural plate border genes have been shown to require FGF signaling (Garnett et al., 2012). Intriguingly, these same genes are also expressed in pluripotent blastula cells, making it important to re-examine the role of FGF-mediated signals at earlier times in Xenopus development, with a focus on understanding their role in the retention of blastula stage pluripotency proposed to underlie genesis of Neural Crest cells. Such studies might also help shed light on the highly context dependent role played by FGF signaling in the regulation of pluripotency in cultured embryonic stem (ES) cells (Lanner and Rossant, 2010). Activation of FGF signaling in naïve mouse embryonic stem cells (mESCs) promotes lineage restriction of these cells (Kunath et al., 2007), whereas FGF activity maintains primed embryonic stem cells, also known as epiblast stem cells (EpiSCs) in a pluripotent/undifferentiated state (Brons et al., 2007; Tesar et al., 2007). Given our model that neural crest cells arise via retention of the attributes of pluripotent blastula cells, we wondered if FGF signaling might play a role in preventing premature lineage restriction of these cells.

In this study, we investigate the requirement for FGF signaling in the transient pluripotency of blastula animal pole cells, and the subsequent establishment of the neural crest state. We find that FGF signaling is essential for normal gene expression in pluripotent blastula cells, and for the capacity of these cells to respond properly to lineage restriction cues. We investigate which FGF-dependent signaling cascades mediate these effects, and find a striking switch in cascade utilization as cells transit from a pluripotent state to a lineage restricted state. We show that pluripotent blastula cells exhibit high Map Kinase signaling, whereas cells undergoing lineage restriction are characterized by increased PI3 Kinase/Akt signaling. Finally, we provide evidence that the balance of FGF-directed Map Kinase and PI3 Kinase/Akt signaling activity plays a role in the retention of blastula stage potential in neural crest cells.

Early embryonic cell fate decisions result from the interplay of a relatively small number of signaling pathways. Because these signals must be used reiteratively to direct a diverse array of outcomes, their output must be highly context-specific. Yet, in many cases, the mechanisms by which this is accomplished remain poorly understood. In this study, we uncover a striking switch in FGFR effector pathway utilization as cells transit from a pluripotent to a lineage-restricted state, adding important new insights into how FGF signaling regulates developmental potential. FGF-mediated Map Kinase activation is prominent in blastula stem cells prior to their exit from the pluripotent state, but then decreases as cells become lineage restricted. By contrast, FGF-mediated PI3 Kinase/Akt signaling is low in pluripotent cells, but increases dramatically as cells undergo lineage restriction. Importantly, both signaling cascades are blocked by dnFGFR4, showing that they are FGF mediated signals. However, although FGFR4 is the predominant FGF receptor expressed in blastula animal pole cells, it is not the only FGF receptor expressed there. It is possible that dnFGFR4 is also blocking the activity of other FGF receptors, such as FGFR1, as dominant negative proteins can sometimes inhibit the activity of related factors. We therefore interpret our findings using this receptor as demonstrating a role for FGF signaling in general, and not a specific role for FGFR4. Interesting, however, dnFGFR1 does not interfere with expression of neural crest markers in the manner that dnFGFR4 does (Figure 1—figure supplement 1).

Given the striking switch in cascade activation as cells move from pluripotency to lineage restriction, we chose to focus on the roles played by these signaling cascades. We show that Map Kinase signaling is essential for the pluripotency of blastula stem cells, whereas PI3 Kinase/Akt signaling appears to play a role in the ability of these cells to adopt a subset of lineage restricted states. Crosstalk between the Map Kinase and PI3 Kinase/Akt pathways is known to be highly context dependent, and can be cooperative or antagonistic (Aksamitiene et al., 2012). Indeed, antagonism between these pathways has been proposed to control the decision by angioblast progenitors to adopt artery versus venous fates (Hong et al., 2006). It will therefore be of great interest to investigate the mechanisms via which the observed switch in pathway utilization is achieved in blastula stem cells. Our preliminary analysis of the gene expression changes that occur as cells progress from a pluripotent to a lineage restricted state suggests that this might be mediated, at least in part, by a change in expression of intracellular adaptor proteins that scaffold the Map kinase vs. PI3 Kinase cascades (Geary and LaBonne, unpublished). We cannot rule out, however, that there could be a switch in the utilization of different FGF receptors that may contribute to the change in pathway utilization. Further studies may shed light on the very different effects that FGF signaling has on the pluripotency of naïve vs. primed embryonic stem cell cultures (Brons et al., 2007; Kunath et al., 2007; Tesar et al., 2007; Hanna et al., 2010; Lanner and Rossant, 2010).

We interpret the findings reported here as evidence that in this system, FGF-mediated Map Kinase activity is required for the pluripotency of blastula stem cells. An alternative interpretation might be that these signals are required for exit from pluripotency. We favor the first interpretation for a number of reasons including the observation that pMap Kinase is high when these cells are pluripotent and down-regulated as they exit pluripotency, and that activating MEK prolongs the expression of the pluripotency marker Sox3. Moreover, blocking Map Kinase signaling does not prevent transit to a neuronal progenitor state, which it might if these signals controlled exit from pluripotency.

In Xenopus, FGF signaling had previously been implicated in the adoption of both mesodermal and neural states, and our current work lends new mechanistic insights into these studies. For example, FGF signaling has been shown to be required for activin-mediated mesoderm formation, but its contributions to this process remained unclear (LaBonne and Whitman, 1994). Our findings that FGF-mediated Map Kinase activation is required for the pluripotency of blastula animal pole cells provides a long sought explanation for this requirement. Our work also sheds light on the role of FGF signaling in neural induction mediated by BMP-antagonists (Launay et al., 1996; Sasai et al., 1996). It has been shown that blocking FGF signaling with the tyrosine kinase inhibitor SU5402 does not prevent adoption of a definitive neural state in response to noggin, even though it can alter Sox2/3 expression (Wills et al., 2010). We find similar effects when blocking Map Kinase signaling, with Chordin-mediated Sox2 expression blocked, but not Nrp1 (Figure 4). Conversely, we find that blocking the PI3 Kinase/Akt cascade prevents transit to both a neural progenitor and definitive neural state., The latter finding confirms what has been described previously and attributed to PI3 Kinase/Akt regulating GSK3 and Wnt signaling (Peng et al., 2004). Although we cannot rule out the possibility of cross-talk between these two signaling pathways, we report an earlier role for PI3 Kinase/Akt in mediating establishment of the neural progenitor state, prompting further exploration into a potential mechanism for its function during early ectodermal patterning. This could be linked to a recently described role for Akt in controlling progenitor cell progression (Pegoraro et al., 2015). The later role for PI3 Kinase/AKT signaling is consistent with the findings of Hongo and colleagues that a dominant negative FGFR4 inhibits commitment of animal cap cells to a neuronal state in response to neural-inducing cues (Hongo et al., 1999).

FGF signaling has also been previously implicated in the genesis of the neural crest (LaBonne and Bronner-Fraser, 1998; Monsoro-Burq et al., 2003; Hong et al., 2008; Nichane et al., 2010; Garnett et al., 2012), although its role relative to BMP and Wnt signaling has remained unclear. Neural Crest stem cells are of central importance to the development and evolution of vertebrates (Groves and LaBonne, 2014), and thus understanding the signals controlling their remarkable developmental potential is essential. Our recent work provides evidence that neural crest cells arise through partial retention of the regulatory network controlling the pluripotency of blastula ancestors (Buitrago-Delgado et al., 2015). Thus, it is crucial to re-examine the roles of signals, such as FGF, which had previously been hypothesized to ‘induce’ developmental potential in these cells, and ask if they might instead be acting earlier in development to control whether blastula animal pole cells become lineage restricted or retain pluripotency. Our findings that FGF-mediated Map Kinase signaling is required for the pluripotency of blastula animal pole cells supports such a model.

These findings led us to ask whether the striking change in FGF effector pathway utilization, from Map Kinase to PI3 Kinase/Akt, as cells transit from a pluripotent to a lineage restricted state could be important for the establishment of the Neural Crest population. Specifically, we wanted to know if the relative levels of Map Kinase and PI3 Kinase/Akt signals displayed by cells could predict or instruct the Neural Crest state. Strikingly, we found that cells that had been reprogrammed to a Neural Crest state by expression of Pax3 and Zic1 retained high levels of Map Kinase activity and low levels of Akt activity even at neurula stages, when control explants had become lineage restricted and transitioned to a high Akt, low Map Kinase activity state (Figure 7a). We found similar results when using an alternative regimen (Snail2 + Wnt signaling) for reprograming to a neural crest state (not shown) indicating that this signaling cascade signature correlates with the retention of pluripotency.

This correlation suggested that retaining a signature of high Map Kinase and low PI3 Kinase/Akt activity might be essential to establishing the Neural Crest stem cell population. We therefore asked if blocking Map Kinase activation and/or prematurely activating PI3 Kinase/Akt signals would interfere with reprograming animal pole explants to a Neural Crest state. Importantly, we found that the expression of FoxD3 and Sox9, characteristic of the Neural Crest state established by Pax3/Zic1, was blocked by either inhibiting Map Kinase activity or prematurely activating Akt using a constitutively active PI3 Kinase subunit (Figure 7b,c). These data support a model in which context dependent control of effector pathways activated by FGF signaling in blastula animal pole cells controls not only the timing of the progression from pluripotency to lineage restriction of these cells, but also the retention of pluripotency and protection from lineage restriction in the cells that will become the Neural Crest. Our findings further suggest that the retention of FGF-mediated Map Kinase signaling in a subset of pluripotent blastula cells may have been an important step in the acquisition of Neural Crest cells, and thus in the evolution of vertebrates.

1. 1. Aksamitiene E
   2. Kiyatkin A
   3. Kholodenko BN

   (2012) Cross-talk between mitogenic Ras/MAPK and survival PI3K/Akt pathways: a fine balance

   Biochemical Society Transactions 40:139–146.

   https://doi.org/10.1042/BST20110609

   • PubMed
   • Google Scholar
2. 1. Amaya E
   2. Musci TJ
   3. Kirschner MW

   (1991) Expression of a dominant negative mutant of the FGF receptor disrupts mesoderm formation in Xenopus embryos

   Cell 66:257–270.

   https://doi.org/10.1016/0092-8674(91)90616-7

   • PubMed
   • Google Scholar
3. 1. Bronner ME
   2. LeDouarin NM

   (2012) Development and evolution of the neural crest: an overview

   Developmental Biology 366:2–9.

   https://doi.org/10.1016/j.ydbio.2011.12.042

   • PubMed
   • Google Scholar
4. 1. Brons IG
   2. Smithers LE
   3. Trotter MW
   4. Rugg-Gunn P
   5. Sun B
   6. Chuva de Sousa Lopes SM
   7. Howlett SK
   8. Clarkson A
   9. Ahrlund-Richter L
   10. Pedersen RA
   11. Vallier L

   (2007) Derivation of pluripotent epiblast stem cells from mammalian embryos

   Nature 448:191–195.

   https://doi.org/10.1038/nature05950

   • PubMed
   • Google Scholar
5. 1. Buitrago-Delgado E
   2. Nordin K
   3. Rao A
   4. Geary L
   5. LaBonne C

   (2015) Neurodevelopment. Shared regulatory programs suggest retention of blastula-stage potential in neural crest cells

   Science 348:1332–1335.

   https://doi.org/10.1126/science.aaa3655

   • PubMed
   • Google Scholar
6. 1. Carballada R
   2. Yasuo H
   3. Lemaire P

   (2001)

   Phosphatidylinositol-3 kinase acts in parallel to the ERK MAP kinase in the FGF pathway during Xenopus mesoderm induction

   Development 128:35–44.

   • PubMed
   • Google Scholar
7. 1. Dorey K
   2. Amaya E

   (2010) FGF signalling: diverse roles during early vertebrate embryogenesis

   Development 137:3731–3742.

   https://doi.org/10.1242/dev.037689

   • PubMed
   • Google Scholar
8. 1. Fletcher RB
   2. Baker JC
   3. Harland RM

   (2006) FGF8 spliceforms mediate early mesoderm and posterior neural tissue formation in Xenopus

   Development 133:1703–1714.

   https://doi.org/10.1242/dev.02342

   • PubMed
   • Google Scholar
9. 1. Fukuda M
   2. Gotoh I
   3. Adachi M
   4. Gotoh Y
   5. Nishida E

   (1997) A novel regulatory mechanism in the mitogen-activated protein (MAP) kinase cascade. Role of nuclear export signal of MAP kinase kinase

   The Journal of Biological Chemistry 272:32642–32648.

   https://doi.org/10.1074/jbc.272.51.32642

   • PubMed
   • Google Scholar
10. 1. Garnett AT
    2. Square TA
    3. Medeiros DM

    (2012) BMP, Wnt and FGF signals are integrated through evolutionarily conserved enhancers to achieve robust expression of Pax3 and Zic genes at the zebrafish neural plate border

    Development 139:4220–4231.

    https://doi.org/10.1242/dev.081497

    • PubMed
    • Google Scholar
11. 1. Golub R
    2. Adelman Z
    3. Clementi J
    4. Weiss R
    5. Bonasera J
    6. Servetnick M

    (2000) Evolutionarily conserved and divergent expression of members of the FGF receptor family among vertebrate embryos, as revealed by FGFR expression patterns in Xenopus

    Development Genes and Evolution 210:345–357.

    https://doi.org/10.1007/s004270000076

    • PubMed
    • Google Scholar
12. 1. Green JB
    2. Smith JC

    (1990) Graded changes in dose of a Xenopus activin A homologue elicit stepwise transitions in embryonic cell fate

    Nature 347:391–394.

    https://doi.org/10.1038/347391a0

    • PubMed
    • Google Scholar
13. 1. Groves AK
    2. LaBonne C

    (2014) Setting appropriate boundaries: fate, patterning and competence at the neural plate border

    Developmental Biology 389:2–12.

    https://doi.org/10.1016/j.ydbio.2013.11.027

    • PubMed
    • Google Scholar
14. 1. Hall BK

    (2000) The neural crest as a fourth germ layer and vertebrates as quadroblastic not triploblastic

    Evolution and Development 2:3–5.

    https://doi.org/10.1046/j.1525-142x.2000.00032.x

    • PubMed
    • Google Scholar
15. 1. Hanna J
    2. Cheng AW
    3. Saha K
    4. Kim J
    5. Lengner CJ
    6. Soldner F
    7. Cassady JP
    8. Muffat J
    9. Carey BW
    10. Jaenisch R

    (2010) Human embryonic stem cells with biological and epigenetic characteristics similar to those of mouse ESCs

    PNAS 107:9222–9227.

    https://doi.org/10.1073/pnas.1004584107

    • PubMed
    • Google Scholar
16. 1. Hardcastle Z
    2. Chalmers AD
    3. Papalopulu N

    (2000) FGF-8 stimulates neuronal differentiation through FGFR-4a and interferes with mesoderm induction in Xenopus embryos

    Current Biology 10:1511–1514.

    https://doi.org/10.1016/S0960-9822(00)00825-3

    • PubMed
    • Google Scholar
17. 1. Hong CC
    2. Peterson QP
    3. Hong JY
    4. Peterson RT

    (2006) Artery/vein specification is governed by opposing phosphatidylinositol-3 kinase and MAP kinase/ERK signaling

    Current Biology 16:1366–1372.

    https://doi.org/10.1016/j.cub.2006.05.046

    • PubMed
    • Google Scholar
18. 1. Hong CS
    2. Park BY
    3. Saint-Jeannet JP

    (2008) Fgf8a induces neural crest indirectly through the activation of Wnt8 in the paraxial mesoderm

    Development 135:3903–3910.

    https://doi.org/10.1242/dev.026229

    • PubMed
    • Google Scholar
19. 1. Hong CS
    2. Saint-Jeannet JP

    (2007) The activity of Pax3 and Zic1 regulates three distinct cell fates at the neural plate border

    Molecular Biology of the Cell 18:2192–2202.

    https://doi.org/10.1091/mbc.E06-11-1047

    • PubMed
    • Google Scholar
20. 1. Hongo I
    2. Kengaku M
    3. Okamoto H

    (1999) FGF signaling and the anterior neural induction in Xenopus

    Developmental Biology 216:561–581.

    https://doi.org/10.1006/dbio.1999.9515

    • PubMed
    • Google Scholar
21. 1. Hoppler S
    2. Wheeler GN

    (2015) It's about time for neural crest

    Science 348:1316–1317.

    https://doi.org/10.1126/science.aab2719

    • Google Scholar
22. 1. Huang X
    2. Saint-Jeannet JP

    (2004) Induction of the neural crest and the opportunities of life on the edge

    Developmental Biology 275:1–11.

    https://doi.org/10.1016/j.ydbio.2004.07.033

    • PubMed
    • Google Scholar
23. 1. Hudson C
    2. Clements D
    3. Friday RV
    4. Stott D
    5. Woodland HR

    (1997) Xsox17alpha and -beta mediate endoderm formation in Xenopus

    Cell 91:397–405.

    https://doi.org/10.1016/S0092-8674(00)80423-7

    • PubMed
    • Google Scholar
24. 1. Iverson C
    2. Larson G
    3. Lai C
    4. Yeh LT
    5. Dadson C
    6. Weingarten P
    7. Appleby T
    8. Vo T
    9. Maderna A
    10. Vernier JM
    11. Hamatake R
    12. Miner JN
    13. Quart B

    (2009) RDEA119/BAY 869766: a potent, selective, allosteric inhibitor of MEK1/2 for the treatment of cancer

    Cancer Research 69:6839–6847.

    https://doi.org/10.1158/0008-5472.CAN-09-0679

    • PubMed
    • Google Scholar
25. 1. Kléber M
    2. Lee HY
    3. Wurdak H
    4. Buchstaller J
    5. Riccomagno MM
    6. Ittner LM
    7. Suter U
    8. Epstein DJ
    9. Sommer L

    (2005) Neural crest stem cell maintenance by combinatorial Wnt and BMP signaling

    The Journal of Cell Biology 169:309–320.

    https://doi.org/10.1083/jcb.200411095

    • PubMed
    • Google Scholar
26. 1. Kunath T
    2. Saba-El-Leil MK
    3. Almousailleakh M
    4. Wray J
    5. Meloche S
    6. Smith A

    (2007) FGF stimulation of the Erk1/2 signalling cascade triggers transition of pluripotent embryonic stem cells from self-renewal to lineage commitment

    Development 134:2895–2902.

    https://doi.org/10.1242/dev.02880

    • PubMed
    • Google Scholar
27. 1. LaBonne C
    2. Bronner-Fraser M

    (1998)

    Neural crest induction in Xenopus: evidence for a two-signal model

    Development 125:2403–2414.

    • PubMed
    • Google Scholar
28. 1. LaBonne C
    2. Burke B
    3. Whitman M

    (1995)

    Role of MAP kinase in mesoderm induction and axial patterning during Xenopus development

    Development 121:1475–1486.

    • PubMed
    • Google Scholar
29. 1. LaBonne C
    2. Whitman M

    (1994) Mesoderm induction by activin requires FGF-mediated intracellular signals

    Trends in Genetics 120:150–472.

    https://doi.org/10.1016/0168-9525(94)90085-X

    • PubMed
    • Google Scholar
30. 1. Lanner F
    2. Rossant J

    (2010) The role of FGF/Erk signaling in pluripotent cells

    Development 137:3351–3360.

    https://doi.org/10.1242/dev.050146

    • PubMed
    • Google Scholar
31. 1. Launay C
    2. Fromentoux V
    3. Shi DL
    4. Boucaut JC

    (1996)

    A truncated FGF receptor blocks neural induction by endogenous Xenopus inducers

    Development 122:869–880.

    • PubMed
    • Google Scholar
32. Book

    1. Le Douarin N
    2. Kalcheim C

    (1999)

    The Neural Crest

    Cambridge University Press.

    • Google Scholar
33. 1. Le Douarin NM
    2. Dupin E

    (2016) The Pluripotency of Neural Crest Cells and Their Role in Brain Development

    Current Topics in Developmental Biology 116:659–678.

    https://doi.org/10.1016/bs.ctdb.2015.10.008

    • PubMed
    • Google Scholar
34. 1. Lea R
    2. Papalopulu N
    3. Amaya E
    4. Dorey K

    (2009) Temporal and spatial expression of FGF ligands and receptors during Xenopus development

    Developmental Dynamics 238:1467–1479.

    https://doi.org/10.1002/dvdy.21913

    • PubMed
    • Google Scholar
35. 1. Lee PC
    2. Taylor-Jaffe KM
    3. Nordin KM
    4. Prasad MS
    5. Lander RM
    6. LaBonne C

    (2012) SUMOylated SoxE factors recruit Grg4 and function as transcriptional repressors in the neural crest

    The Journal of Cell Biology 198:799–813.

    https://doi.org/10.1083/jcb.201204161

    • PubMed
    • Google Scholar
36. 1. Mayor R
    2. Guerrero N
    3. Martínez C

    (1997) Role of FGF and noggin in neural crest induction

    Developmental Biology 189:1–12.

    https://doi.org/10.1006/dbio.1997.8634

    • PubMed
    • Google Scholar
37. 1. Monsoro-Burq AH
    2. Fletcher RB
    3. Harland RM

    (2003) Neural crest induction by paraxial mesoderm in Xenopus embryos requires FGF signals

    Development 130:3111–3124.

    https://doi.org/10.1242/dev.00531

    • PubMed
    • Google Scholar
38. 1. Monsoro-Burq AH
    2. Wang E
    3. Harland R

    (2005) Msx1 and Pax3 cooperate to mediate FGF8 and WNT signals during Xenopus neural crest induction

    Developmental Cell 8:167–178.

    https://doi.org/10.1016/j.devcel.2004.12.017

    • PubMed
    • Google Scholar
39. 1. Nichane M
    2. Ren X
    3. Bellefroid EJ

    (2010) Self-regulation of Stat3 activity coordinates cell-cycle progression and neural crest specification

    The EMBO Journal 29:55–67.

    https://doi.org/10.1038/emboj.2009.313

    • PubMed
    • Google Scholar
40. 1. Nie S
    2. Chang C

    (2007) PI3K and Erk MAPK mediate ErbB signaling in Xenopus gastrulation

    Mechanisms of Development 124:657–667.

    https://doi.org/10.1016/j.mod.2007.07.005

    • PubMed
    • Google Scholar
41. 1. Nordin K
    2. LaBonne C

    (2014) Sox5 Is a DNA-binding cofactor for BMP R-Smads that directs target specificity during patterning of the early ectoderm

    Developmental Cell 31:374–382.

    https://doi.org/10.1016/j.devcel.2014.10.003

    • PubMed
    • Google Scholar
42. 1. Pegoraro C
    2. Figueiredo AL
    3. Maczkowiak F
    4. Pouponnot C
    5. Eychène A
    6. Monsoro-Burq AH

    (2015) PFKFB4 controls embryonic patterning via Akt signalling independently of glycolysis

    Nature Communications 6:5953.

    https://doi.org/10.1038/ncomms6953

    • PubMed
    • Google Scholar
43. 1. Peng Y
    2. Jiang BH
    3. Yang PH
    4. Cao Z
    5. Shi X
    6. Lin MC
    7. He ML
    8. Kung HF

    (2004) Phosphatidylinositol 3-kinase signaling is involved in neurogenesis during Xenopus embryonic development

    Journal of Biological Chemistry 279:28509–28514.

    https://doi.org/10.1074/jbc.M402294200

    • PubMed
    • Google Scholar
44. 1. Prasad MS
    2. Sauka-Spengler T
    3. LaBonne C

    (2012) Induction of the neural crest state: control of stem cell attributes by gene regulatory, post-transcriptional and epigenetic interactions

    Developmental Biology 366:10–21.

    https://doi.org/10.1016/j.ydbio.2012.03.014

    • PubMed
    • Google Scholar
45. 1. Ribisi S
    2. Mariani FV
    3. Aamar E
    4. Lamb TM
    5. Frank D
    6. Harland RM

    (2000) Ras-mediated FGF signaling is required for the formation of posterior but not anterior neural tissue in Xenopus laevis

    Developmental Biology 227:183–196.

    https://doi.org/10.1006/dbio.2000.9889

    • PubMed
    • Google Scholar
46. 1. Rogers CD
    2. Jayasena CS
    3. Nie S
    4. Bronner ME

    (2012) Neural crest specification: tissues, signals, and transcription factors

    Wiley Interdisciplinary Reviews: Developmental Biology 1:52–68.

    https://doi.org/10.1002/wdev.8

    • PubMed
    • Google Scholar
47. 1. Sasai Y
    2. Lu B
    3. Piccolo S
    4. De Robertis EM

    (1996)

    Endoderm induction by the organizer-secreted factors chordin and noggin in Xenopus animal caps

    The EMBO Journal 15:4547–4555.

    • PubMed
    • Google Scholar
48. 1. Sasai Y
    2. Lu B
    3. Steinbeisser H
    4. De Robertis EM

    (1995) Regulation of neural induction by the Chd and Bmp-4 antagonistic patterning signals in Xenopus

    Nature 376:333–336.

    https://doi.org/10.1038/376333a0

    • PubMed
    • Google Scholar
49. 1. Slack JM
    2. Darlington BG
    3. Heath JK
    4. Godsave SF

    (1987) Mesoderm induction in early Xenopus embryos by heparin-binding growth factors

    Nature 326:197–200.

    https://doi.org/10.1038/326197a0

    • PubMed
    • Google Scholar
50. 1. Stuhlmiller TJ
    2. García-Castro MI

    (2012a) Current perspectives of the signaling pathways directing neural crest induction

    Cellular and Molecular Life Sciences 69:3715–3737.

    https://doi.org/10.1007/s00018-012-0991-8

    • PubMed
    • Google Scholar
51. 1. Stuhlmiller TJ
    2. García-Castro MI

    (2012b) FGF/MAPK signaling is required in the gastrula epiblast for avian neural crest induction

    Development 139:289–300.

    https://doi.org/10.1242/dev.070276

    • PubMed
    • Google Scholar
52. 1. Taylor KM
    2. LaBonne C

    (2007) Modulating the activity of neural crest regulatory factors

    Current Opinion in Genetics & Development 17:326–331.

    https://doi.org/10.1016/j.gde.2007.05.012

    • PubMed
    • Google Scholar
53. 1. Tesar PJ
    2. Chenoweth JG
    3. Brook FA
    4. Davies TJ
    5. Evans EP
    6. Mack DL
    7. Gardner RL
    8. McKay RD

    (2007) New cell lines from mouse epiblast share defining features with human embryonic stem cells

    Nature 448:196–199.

    https://doi.org/10.1038/nature05972

    • PubMed
    • Google Scholar
54. 1. Thomsen G
    2. Woolf T
    3. Whitman M
    4. Sokol S
    5. Vaughan J
    6. Vale W
    7. Melton DA

    (1990) Activins are expressed early in Xenopus embryogenesis and can induce axial mesoderm and anterior structures

    Cell 63:485–493.

    https://doi.org/10.1016/0092-8674(90)90445-K

    • PubMed
    • Google Scholar
55. 1. Umbhauer M
    2. Marshall CJ
    3. Mason CS
    4. Old RW
    5. Smith JC

    (1995) Mesoderm induction in Xenopus caused by activation of MAP kinase

    Nature 376:58–62.

    https://doi.org/10.1038/376058a0

    • PubMed
    • Google Scholar
56. 1. Wills AE
    2. Choi VM
    3. Bennett MJ
    4. Khokha MK
    5. Harland RM

    (2010) BMP antagonists and FGF signaling contribute to different domains of the neural plate in Xenopus

    Developmental Biology 337:335–350.

    https://doi.org/10.1016/j.ydbio.2009.11.008

    • PubMed
    • Google Scholar
57. 1. Xu RH
    2. Kim J
    3. Taira M
    4. Sredni D
    5. Kung H

    (1997)

    Studies on the role of fibroblast growth factor signaling in neurogenesis using conjugated/aged animal caps and dorsal ectoderm-grafted embryos

    Journal of Neuroscience 17:6892–6898.

    • PubMed
    • Google Scholar
58. 1. Yardley N
    2. García-Castro MI

    (2012) FGF signaling transforms non-neural ectoderm into neural crest

    Developmental Biology 372:166–177.

    https://doi.org/10.1016/j.ydbio.2012.09.006

    • PubMed
    • Google Scholar
59. 1. Ying QL
    2. Nichols J
    3. Chambers I
    4. Smith A

    (2003) BMP induction of Id proteins suppresses differentiation and sustains embryonic stem cell self-renewal in collaboration with STAT3

    Cell 115:281–292.

    https://doi.org/10.1016/S0092-8674(03)00847-X

    • PubMed
    • Google Scholar
60. 1. Zimmerman LB
    2. De Jesús-Escobar JM
    3. Harland RM

    (1996) The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4

    Cell 86:599–606.

    https://doi.org/10.1016/S0092-8674(00)80133-6

    • PubMed
    • Google Scholar

Editors’ note: a previous version of this study was rejected after peer review, but the authors submitted for reconsideration. The first decision letter after peer review is shown below.]

Thank you for choosing to send your work, "FGF mediated MAPK and PI3K/Akt Signals make distinct contributions to pluripotency and the establishment of Neural Crest", for consideration at eLife.

Your submission has been assessed by a Senior and Reviewing Editor, Marianne Bronner, and three reviewers who have elected to remain anonymous. The reviewers found the paper to be potentially interesting and well-written. However, they also felt the authors need to strengthen biochemical characterization of the pathway and confirm that this proposed "molecular switch" is downstream of the receptor. To accomplish this the authors would have to collect persuasive evidence linking Akt to FGF by demonstrating dnFGFR4 specificity, using other means of interfering with pathways downstream of FGF, and using more markers to prove their point. In addition, they should better discuss previous literature on this topic as well as alternative interpretations of their data. In general, it was felt that the present manuscript is a description of a phenomenon rather than a molecular characterization of a new mechanism. Please see the individual reviews below for further details.

Given that these revisions are likely to take a significant amount of time and the policy of eLife is to reject papers that require more than two months for modifications, we have no choice but to decline the present version of the paper. That said, if you feel you can make the revisions, we will make every effort to return a new version of the manuscript to the same reviewers.

Reviewer #1:

Geary and LaBonne describe differential activation of the downstream effector pathways of FGF-signaling in pluripotent and lineage-restricted Xenopus cells. They utilize animal cap assays to show that blastula stem cells have high levels of MAP Kinase signaling, while cells that are differentiating have prevalence of PI3 Kinase/Akt signals – both downstream of FGFs. The authors suggest that the transition between these two effector pathways, which takes place during gastrulation, is required for the establishment of distinct cell lineages. Furthermore, reprogramming experiments are employed to show that maintenance of high MAP Kinase signaling is required for the establishment of the neural crest population. In line with their previous work, the authors suggest that maintenance of FGF/MAPK signaling from blastula stages is necessary for specification of the neural crest and might underlie its multipotency.

The manuscript is well written and the results are presented clearly. The switch in FGF effector deployment is an exciting phenomenon that holds important implications for the mode of action of signaling pathways during early development. Whereas the authors convincingly tie this molecular switch to cellular properties such as multipotency, the fact that virtually all experiments are performed in animal cap assays raises substantial concerns about the validity of the authors' findings in vivo. Thus, evidence of the presence of the MAPK-PI3K switch in developing embryos would be an important addition to the work.

Major points:

1) Lack of in vivo experiments: The authors employ animal cap assays for both loss and gain of function experiments, and to test questions pertaining to cell potential of blastula and neural crest cells. Validation of the FGF effector switch in vivo would greatly enhance the work. In particular, I was puzzled by the reprogramming experiments – why not examine neural crest formation as it takes place in a developing embryo instead of artificially generating neural crest-like cells through overexpression?

2) On the central role ascribed to FGFR4: Throughout the text, the authors state that activation of the MAP Kinase and the PI3 Kinase/Akt pathways are downstream of FGFR4. This claim is based on the expression levels of FGF receptors in the early Xenopus embryo (data not shown), and the use of a FGFR4 dominant negative construct, which causes loss of both Erk and Akt phosphorylation. Given the promiscuity of dnFGFR constructs and the presence of multiple FGF receptors in the blastula/neural crest, I am not convinced that FGFR4 is the main trigger of the MAPK and PI3 effector pathways. Instead, I am tempted to speculate that the switch occurs due to activation of a different FGFR type or isoform. The central role of FGFR4 in this process can only be confirmed by specific knockdown of this transcript.

3) Crosstalk between FGF effector pathways: Since intracellular effector pathways of signaling systems are very complex and known to intersect, how does activation/inhibition of PI3K affect activity of MAPK and vice versa? Are the effector pathways inhibiting each other as one might speculate from Figure 3 and Table 1?

4) Possible mechanisms of the molecular switch: While characterization of the molecular mechanism controlling the switch between effector pathways may be beyond the scope of the work, as I reader I expected to find this point addressed in the Discussion. What would be the factors that could tilt the balance between MAPK and PI3K, specifically during gastrula stages?

Reviewer #2:

This manuscript makes advances in the understanding of pluripotency of the "animal cap" which here is referred to as pluripotent ectoderm, in line with the hypothesis that Neural crest is the retention of a pluripotent state. This case was made in a previous paper, and the current work builds on that.

This manuscript makes the case that FGF acting through FGFR4, and MAPkinase maintains pluripotency. In contrast, Akt/PI3Kinase terminate pluripotency.

The manuscript builds a circumstantial case for the importance of FGFR4 in maintaining pluripotency, but the results are based on a dominant negative receptor, whose specificity has not been proven. Therefore discussion regarding FGFR4 dominant negatives, and whether this manipulation implicated FGFR4 specifically in signaling needs to be tightened up or experimentally verified.

The results with FGFR4 appear to be very different from those from Hongo/Okamoto, who essentially argued the opposite, that DN FGFR4 blocks anterior neural induction. The precise experiments seem different, but it would be very useful to arrive at a coherent explanation if possible.

The most interesting aspect of the work is the distinction between Akt/PI3K activity in not being required for pluripotency, and Mapkinase activity requirement for pluripotency. One would expect that this would be more thoroughly documented, using independent inhibitors of the pathways, and assaying additional consequences of pluripotency, in addition to Epidermal Keratin. I don't think this would be an unusual request, but is in line with the general view that pharmacological manipulations, where possible should be backed up with additional drug or molecular manipulation.

With respect to the inhibition of progression of the cells to neural states with the PI3K inhibitor, it would be essential to know what the cells are doing instead (presumably retaining pluripotency markers?). It is important to rule out the possibility that this does not simply cause cell death, or general ill health of the cells, which would also prevent maturation of the cells, which some have argued accounts for the failure of neutralization when FGF is inhibited (Wills, Harland).

With respect to the requirement for both MapKinase and PI3Kinase pathways being required for mesoderm induction, the manuscript should refer to previous work on this, such as Umbhauer and Smith, and Carballada, Lemaire. Their results appear to be similar, so perhaps this is not new.

Figure 6A is not very compelling – the premature downregulation of Sox3 by p110-Caax and the retention of Sox3 by MAPK should be quantified more convincingly.

Figure 7C and D are not very compelling; again some quantification would be useful.

Reviewer #3:

In this manuscript, the authors investigate the signalling requirements for Xenopus animal caps to remain in a multipotent state or to differentiate into different ectodermal derivatives (epidermis/neural) or other germ layers (endoderm/mesoderm). FGF signalling is well known to be important for the regulation of pluripotency in mouse and other species, hence they focus on FGF.

The authors show that FGF signalling via FGFR4 is required for early blastula animal caps to become epidermis, and that as animal caps age they switch from high MAPK to high Akt signalling, while losing MAPK activity. While the acquisition of epidermal fate depends on MAPK signalling, the acquisition of neural identity after Chordin injection requires Akt, thus, both pathways are required differentially. They then go on to show that the same is true for the acquisition of endodermal (MAPK dependent) and mesoderm (Akt dependent) fates.

Next, they aim to assess whether prolonged MAPK or premature Akt activation alters animal cap 'pluripotency' (only assessed by Sox3) and suggest that this indeed changes the retention of pluripotent character.

Finally, the authors have previously shown that neural crest cells have retained pluripotency rather than undergoing lineage restriction. They now assess whether induction of neural crest cells by a combination of transcription factors alters MAPK or Akt signalling and show that this indeed activates MAPK, but not Akt, and that MAPK inhibition or Akt overactivation prevents neural crest induction by these factors.

They conclude that MAPK is essential for the pluripotency state, while Akt is important for animal caps to adopt some lineages.

I am somewhat concerned about the novelty of their findings and about the final interpretation of the results with a central conclusion relying on a single marker. Modulation of FGF signalling during pluripotency and the loss thereof is well established, and in this sense the study adds detail to a well-established principle rather than revealing novel concepts. Of course it is interesting how different downstream pathways act differentially, however this in itself is not a new principle.

A central part of their conclusion is that MAPK mediated FGF signalling is required for cells to maintain pluripotency (e.g. Discussion, second paragraph). However, they never show that pluripotency makers depend on MAPK – they only ever examine Oct60 and Sox3 with Sox3 also being expressed in neural tissue. This is not sufficient to make a general conclusion about pluripotency. In Figure 6 they show that prolonged MAPK activation leads to prolonged Sox3 expression, while activation of Akt does the opposite. They equate this to retaining pluripotency. However, as they themselves show in Figure 2, they consider Sox3 as neural marker and argue that this is indeed neural tissue because it does not express Oct60. They must demonstrate that in the experiment in Figure 6 Oct60 and other 'pluripotency makers' are affected and that MAPK inhibition leads to loss of pluripotency genes. For this to be convincing they have to demonstrate a panel of genes rather than just one.

The authors argue that retaining MAPK activity is key for retaining pluripotency in neural crest cells. However they do not show this; they show that neural crest induction by Pax3/Zic1 elevates FGFR4 expression and MAPK suggesting that these changes in signalling activity is a consequence of neural crest cell formation, but not the driver.

In Figure 2 the authors show that dnFGFR4 enhances neural markers (sox3, nrp1). They then go on to show that Akt signalling is required for cells to become neutralised when injected with chordin (Figure 4). These results are contradictory, and the first result is inconsistent with findings that FGF is required for neural induction. Does Akt or MAPK inhibition enhance neural gene expression similar to what is shown in Figure 2B?

In summary, while the experiments are generally well conducted and controlled the results appear to be overinterpreted often depending on a single marker and the conclusions are rather forceful. The involvement of FGF in pluripotency is well established, although a potential differential role if this can be proven may be interesting.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for submitting your article "FGF mediated MAPK and PI3K/Akt Signals make distinct contributions to pluripotency and the establishment of Neural Crest" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by Marianne Bronner as the Senior and Reviewing Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

The reviewers find the manuscript improved. However, they think that some of the conclusions overreach the data and that parts of the manuscript should be rewritten and softened, as described below:

Essential revisions:

1) Since the additional experiments examining the role of FGFR4 were not definitive, we ask that you rewrite this part of the manuscript to make it clear that a possible role for FGFR4 is speculative. While it may have a role in the process, the evidence is not definitive and the mechanism by which the DNFGFR1 construct has effects different from DNFGFR1 is unresolved.

2) It's unclear if the experiment with the DNFGFR1 is of any significance and feels like an afterthought. This would also impact the Discussion since it is possible that the switch between MAPK and PI3K involves activation of distinct receptors.

[Editors’ note: the author responses to the first round of peer review follow.]

Reviewer #1:

[…] 1) Lack of in vivo experiments: The authors employ animal cap assays for both loss and gain of function experiments, and to test questions pertaining to cell potential of blastula and neural crest cells. Validation of the FGF effector switch in vivo would greatly enhance the work. In particular, I was puzzled by the reprogramming experiments – why not examine neural crest formation as it takes place in a developing embryo instead of artificially generating neural crest-like cells through overexpression?

These experiments were done in explants because in those you can monitor the progression from pluripotency to lineage restricted without influence from other tissues and signals, allowing the experiments to be more tightly controlled. This also allows separation from the confounded role of FGF signaling in posteriorization of the embryonic axes. In this work we were focused on how pluripotent blastula (ES) cells retain their pluripotency, and this was best explored by isolating them. Also, because pharmacological inhibitors lack spatial control, you cannot be certain that the phenotypic effects are direct on the tissue of interest when working with intact embryos. Nonetheless, in the revised manuscript we have now included whole embryo experiments as supplements to Figures 1 and 7 confirming the effects of these pathways on neural crest endogenously.

2) On the central role ascribed to FGFR4: Throughout the text, the authors state that activation of the MAP Kinase and the PI3 Kinase/Akt pathways are downstream of FGFR4. This claim is based on the expression levels of FGF receptors in the early Xenopus embryo (data not shown), and the use of a FGFR4 dominant negative construct, which causes loss of both Erk and Akt phosphorylation. Given the promiscuity of dnFGFR constructs and the presence of multiple FGF receptors in the blastula/neural crest, I am not convinced that FGFR4 is the main trigger of the MAPK and PI3 effector pathways. Instead, I am tempted to speculate that the switch occurs due to activation of a different FGFR type or isoform. The central role of FGFR4 in this process can only be confirmed by specific knockdown of this transcript.

We agree that we have not ruled out the possibility that the dnFGFR4 does not block the activity of other FGFRs and we now discuss this in the text, and make clear that we are blocking FGF signaling, but not necessarily only FGFR4 signaling. We also include data showing that dnFGFR1 does not have the same effects on neural crest that dnFGFR4 does. We note that in response to the reviews we did attempt to morpholino deplete FGFR4 vs FGFR1 – we purchased four published MOs for these receptors listed on the Xenbase site, but we found they tended to crash out of solution and were not useful for loss of function assays. We asked Gene Tools to look at the sequences, and opined that they were poorly designed and predicted to be insoluble. Rather than spend more money on morpholinos, we have more recently been trying to use G0-CRiSPR mutagenesis to deplete these receptors. Unfortunately we have not been able to get early and uniform enough mutagenesis to do these experiments (plus the receptors are maternally provided) and we have delayed publication of this work as long as we can. We had never meant to emphasize the role of one particular receptor in this work (we used FGFR4 because this receptor is the predominant one expressed at blastula stages) – our focus has always been on the downstream effector pathways of FGF signaling and their functions and we make that clear in the text.

3) Crosstalk between FGF effector pathways: Since intracellular effector pathways of signaling systems are very complex and known to intersect, how does activation/inhibition of PI3K affect activity of MAPK and vice versa? Are the effector pathways inhibiting each other as one might speculate from Figure 3 and Table 1?

Figure 3—figure supplement 1 now shows the specificity of the Mek and PI3K inhibitors for their own pathways.

4) Possible mechanisms of the molecular switch: While characterization of the molecular mechanism controlling the switch between effector pathways may be beyond the scope of the work, as I reader I expected to find this point addressed in the Discussion. What would be the factors that could tilt the balance between MAPK and PI3K, specifically during gastrula stages?

We now address this point in the Discussion. We believe that the context/competency of the cells changes over time, and that changes in the expression of intracellular FGF pathway adaptors/modulators over developmental time likely underlies the switch in effector pathway utilization. We agree that experiments on this are beyond the scope of the current work.

Reviewer #2:

[…] The manuscript builds a circumstantial case for the importance of FGFR4 in maintaining pluripotency, but the results are based on a dominant negative receptor, whose specificity has not been proven. Therefore discussion regarding FGFR4 dominant negatives, and whether this manipulation implicated FGFR4 specifically in signaling needs to be tightened up or experimentally verified.

The results with FGFR4 appear to be very different from those from Hongo/Okamoto, who essentially argued the opposite, that DN FGFR4 blocks anterior neural induction. The precise experiments seem different, but it would be very useful to arrive at a coherent explanation if possible.

The reviewer refers to a 1999 DB paper that focused on a role for FGF signaling in anterior neural induction that used dnFGFR1 and dnFGFR 4 constructs. The reviewer is correct that the experiments in that paper are done quite differently than ours: explants were isolated at gastrula stages (stage 10), a number of the experiments were done with dissociated cells, and changes in gene expression were largely analyzed by RT-PCR assays at stage 25 with only two genes in common to our study. With so many differences it’s how to come up with a single explanation for why the results differ. We can only report the results of our experiments, which focused on earlier stages and which were all carried out numerous times with highly reproducible results.

The most interesting aspect of the work is the distinction between Akt/PI3K activity in not being required for pluripotency, and Mapkinase activity requirement for pluripotency. One would expect that this would be more thoroughly documented, using independent inhibitors of the pathways, and assaying additional consequences of pluripotency, in addition to Epidermal Keratin. I don't think this would be an unusual request, but is in line with the general view that pharmacological manipulations, where possible should be backed up with additional drug or molecular manipulation.

In the revised manuscript we have repeated all of the experiments with at least one additional way of inhibiting each pathway. We now use dnRaf, in addition to the Mek inhibitors, for Mapk inhibition. We use Wortmanin and Ly294, in addition to dnPI3K, for Akt inhibition. We also use multiple markers for each cell fate examined (EpK and Trim29 for epidermis; Xbra and MyoD for mesoderm; Sox17 and endodermin for endoderm; Sox2/3 and Nrp1 for neural etc.). We examine pluripotency both by marker expression and by functional tests of pluripotency. We also now include qPCR data quantifying gene expression changes in response to inhibition of each pathway.

With respect to the inhibition of progression of the cells to neural states with the PI3K inhibitor, it would be essential to know what the cells are doing instead (presumably retaining pluripotency markers?). It is important to rule out the possibility that this does not simply cause cell death, or general ill health of the cells, which would also prevent maturation of the cells, which some have argued accounts for the failure of neutralization when FGF is inhibited (Wills, Harland).

The revised manuscript includes qPCR data examining the changes in gene expression in explants inhibited for each pathway. In the case of the PI3K inhibitor there is an interesting up-regulation of a broad set of lineage markers, suggesting that the cells are poised for lineage decisions but impaired in their decision making. We were also very careful to use low doses of inhibitors such as LY294 as they can have off target effects at higher doses. For our studies we titrated our doses to the lowest level sufficient to completely block Akt activation, which is 3-10 fold lower than doses used in some studies and which can generate nonspecific effects, so we are confident that these changes in gene expression are specific to this pathway.

With respect to the requirement for both MapKinase and PI3Kinase pathways being required for mesoderm induction, the manuscript should refer to previous work on this, such as Umbhauer and Smith, and Carballada, Lemaire. Their results appear to be similar, so perhaps this is not new.

These papers are now cited/discussed. These pathways have indeed been implicated in mesoderm induction previously. What differs here is how that finding is interpreted. Those studies, and those of LaBonne and Whitman, focused on a single lineage decision. Our current work is focused on pluripotency and lineage decisions more broadly. We thus provide a new interpretation of those prior findings and put them in a broader context. For example, with respect to Mapk, we propose a general effect on pluripotency rather than a specific effect on mesoderm.

Figure 6A is not very compelling – the premature downregulation of Sox3 by p110-Caax and the retention of Sox3 by MAPK should be quantified more convincingly.

We have both replaced the images and included qPCR data here.

Figure 7C and D are not very compelling; again some quantification would be useful.

We have provided better images for these data. Other typos and references errors noted in the review have been corrected and we thank the reviewer for prompting these changes, which have improved the manuscript.

Reviewer #3:

[…] I am somewhat concerned about the novelty of their findings and about the final interpretation of the results with a central conclusion relying on a single marker. Modulation of FGF signalling during pluripotency and the loss thereof is well established, and in this sense the study adds detail to a well-established principle rather than revealing novel concepts. Of course it is interesting how different downstream pathways act differentially, however this in itself is not a new principle.

We respectfully disagree about the novelty of these findings. First, a role for FGF in pluripotencyhad not been shown in this system, which is the system in which a lot of prior work on germ layer formation had been carried out. Indeed FGF had been ascribed very different roles in the Xenopus system, which did not integrate very well with findings in mouse and human ES cells. We feel therefore that our results advance the field by showing the central role this signaling is playing in pluripotency is evolutionarily conserved. Our work also further advances the field by adding exciting and very novel data on a switch in effector pathways. This finding is not only important in its own right, but also may help explain differences in what FGF signaling does in naive vs. primed ES (and human vs. mouse) cells. Finally, we show a completely novel role for FGF dependent Map kinase signaling in the retention of pluripotency that led to neural crest formation, providing a significant mechanistic advance over our prior Science paper.

A central part of their conclusion is that MAPK mediated FGF signalling is required for cells to maintain pluripotency (e.g. Discussion, second paragraph). However, they never show that pluripotency makers depend on MAPK – they only ever examine Oct60 and Sox3 with Sox3 also being expressed in neural tissue. This is not sufficient to make a general conclusion about pluripotency. In Figure 6 they show that prolonged MAPK activation leads to prolonged Sox3 expression, while activation of Akt does the opposite. They equate this to retaining pluripotency. However, as they themselves show in Figure 2, they consider Sox3 as neural marker and argue that this is indeed neural tissue because it does not express Oct60. They must demonstrate that in the experiment in Figure 6 Oct60 and other 'pluripotency makers' are affected and that MAPK inhibition leads to loss of pluripotency genes. For this to be convincing they have to demonstrate a panel of genes rather than just one.

The strongest and most relevant data in assessing pluripotency comes from functional studies that challenge cells to adopt different fates/states. Those pluripotency experiments are shown in Figures 3, 4 and 5, and each fate/state is evaluated with more than one marker. We believe these data collective provide compelling evidence for a loss of pluripotency. In addition we examine changes in the expression of a panel of genes in response to inhibitor treatment in Figure 5—figure supplement 2.

The authors argue that retaining MAPK activity is key for retaining pluripotency in neural crest cells. However they do not show this; they show that neural crest induction by Pax3/Zic1 elevates FGFR4 expression and MAPK suggesting that these changes in signalling activity is a consequence of neural crest cell formation, but not the driver.

We do show that establishing a neural crest state using Pax3/zip retains pMapK (and FGFR4 expression), however we also more directly show in Figure 7 that blocking Map kinase activation blocks expression of neural crest markers Sox9 and Foxd3 in explants. In addition, in Figure 7—figure supplement 2, we show that dnRaf blocks neural crest formation in whole embryos.

In Figure 2 the authors show that dnFGFR4 enhances neural markers (sox3, nrp1). They then go on to show that Akt signalling is required for cells to become neutralised when injected with chordin (Figure 4). These results are contradictory, and the first result is inconsistent with findings that FGF is required for neural induction. Does Akt or MAPK inhibition enhance neural gene expression similar to what is shown in Figure 2B?

The level of Sox3/Nrp1 expression found in dnFGFR4 explants is much weaker than observed in chordin-neuralized explants (now shown in Figure 2). We think this represents a bias toward a neuronal progenitor state but not neural induction. Blocking Mapk activation does not affect chordin-mediated neural induction, but blocking Akt activation does, as the reviewer notes. Interestingly blocking Akt also leads to weak upregulation of Sox2 and Sox3 along with other lineage markers (see Figure 4—figure supplement 2), which we interpret as evidence that the cells are poised for lineage decisions, but impaired in their decision making. We don’t find these data contradictory but rather that it reflects the complexity of signaling systems, and blocking any one downstream signaling cascade may not fully recapitulate blocking all downstream signals. It should be noted that FGF signaling is not required for neural induction/anterior neural development in Xenopus (BMP antagonism is sufficient for this, see Wills et al. 2010 for example) but has been shown to be essential for posterior neural development.

[Editors' note: the author responses to the re-review follow.]

Essential revisions:

1) Since the additional experiments examining the role of FGFR4 were not definitive, we ask that you rewrite this part of the manuscript to make it clear that a possible role for FGFR4 is speculative. While it may have a role in the process, the evidence is not definitive and the mechanism by which the DNFGFR1 construct has effects different from DNFGFR1 is unresolved.

We have rewritten these parts of the manuscript to reflect this.

2) It's unclear if the experiment with the DNFGFR1 is of any significance and feels like an afterthought. This would also impact the Discussion since it is possible that the switch between MAPK and PI3K involves activation of distinct receptors.

We have added this to the Discussion.

Author details

1. Lauren Geary

   Department of Molecular Biosciences, Northwestern University, Evanston, United States

   Contribution

   Data curation, Formal analysis, Investigation, Writing—original draft, Writing—review and editing

   Competing interests

   No competing interests declared

2. Carole LaBonne

   1. Department of Molecular Biosciences, Northwestern University, Evanston, United States
   2. Robert H Lurie Comprehensive Cancer Center, Northwestern University, Evanston, United States

   Contribution

   Conceptualization, Formal analysis, Supervision, Funding acquisition, Writing—review and editing

   For correspondence

   clabonne@northwestern.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-6001-7179

• 2,504

  Page views

• 301

  Downloads

• 24

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.