Vertebrate Cell Differentiation, Evolution, and Diseases: The Vertebrate-Specific Developmental Potential Guardians VENTX/NANOG and POU5/OCT4 Enter the Stage

During vertebrate development, embryonic cells pass through a continuum of transitory pluripotent states that precede multi-lineage commitment and morphogenesis. Such states are referred to as “refractory/naïve” and “competent/formative” pluripotency. The molecular mechanisms maintaining refractory pluripotency or driving the transition to competent pluripotency, as well as the cues regulating multi-lineage commitment, are evolutionarily conserved. Vertebrate-specific “Developmental Potential Guardians” (vsDPGs; i.e., VENTX/NANOG, POU5/OCT4), together with MEK1 (MAP2K1), coordinate the pluripotency continuum, competence for multi-lineage commitment and morphogenesis in vivo. During neurulation, vsDPGs empower ectodermal cells of the neuro-epithelial border (NEB) with multipotency and ectomesenchyme potential through an “endogenous reprogramming” process, giving rise to the neural crest cells (NCCs). Furthermore, vsDPGs are expressed in undifferentiated-bipotent neuro-mesodermal progenitor cells (NMPs), which participate in posterior axis elongation and growth. Finally, vsDPGs are involved in carcinogenesis, whereby they confer selective advantage to cancer stem cells (CSCs) and therapeutic resistance. Intriguingly, the heterogenous distribution of vsDPGs in these cell types impact on cellular potential and features. Here, we summarize the findings about the role of vsDPGs during vertebrate development and their selective advantage in evolution. Our aim to present a holistic view regarding vsDPGs as facilitators of both cell plasticity/adaptability and morphological innovation/variation. Moreover, vsDPGs may also be at the heart of carcinogenesis by allowing malignant cells to escape from physiological constraints and surveillance mechanisms.

Since VENTX/NANOG and POU5/OCT4 are vertebrate-specific innovations that arose in the last common ancestor of extant vertebrates ( Figure 1B) [15], we propose to call them vertebrate-specific "Developmental Potential Guardians" (vsDPGs hereafter). Interestingly, whilst cell transplantation of vertebrate donor blastula/epiblast cells (expressing vsDPGs) revealed their capacity to give rise to all somatic cell types of the host organism [2,6,49], comparable studies in invertebrate chordates (i.e., hemichordates, cephalochordates, and urochordates) showed that blastula cells keep a memory of their previous spatial localization while interpreting external stimuli and thus lack stricto sensu pluripotency features [6,[50][51][52]. This suggest that the molecular origin of vsDPGs may be intimately linked to the rise of a "new and broader competence potential" in cells during vertebrate development.
The imperative transition from refractory to competent pluripotency in development lies at the core of complex molecular interaction between vsDPGs and FGF/MAPK signaling pathway. Several studies propose that MEK1 (MAP2K1 kinase, MAPK pathway) is the "universal competence factor" in vertebrate PSCs transitioning from a refractory to a competent state. MEK1 mediates VENTX/NANOG protein degradation by the proteasome through a PEST destruction motif (an amino acid sequence enriched with Proline-P, Glutamate-E, Serine-S and Threonine-T in the VENTX/NANOG N-terminus) thus allowing PSCs to responds to differentiating cues [9,14,[59][60][61][62].    Studies in Xenopus showed that VENTX protein degradation occurs during PSCs mitosis (anaphase) in a non-polarized asymmetric manner ( Figure 2B) [14]. Once PSC enter into mitosis, VENTX localizes on the chromosomes until metaphase, when it detaches from DNA in a MEK1-independent manner. During anaphase, chromosomes of one daughter forming PSC inherit VENTX protein, which re-localizes on DNA, whereas the other daughter forming PSC does not inherit VENTX. Such asymmetric inheritance/distribution of VENTX at anaphase is under the control of MEK1, which regulates unpolarized asymmetric degradation/clearance of VENTX through PEST destruction motif of VENTX at anaphase [14]. The SCF-β-TRCP ubiquitin-mediated global clearance of VENTX allows for germ layer determination at the onset of morphogenesis (i.e., gastrulation) [62]. Thus, asymmetric cell division (ACD) of PSCs in vivo results in a heterogeneous population of PSCs with high-VENTX (refractory) and low-VENTX (competent) activity. Functional analyses further demonstrated that either PEST-mutant VENTX (undegradable) gain-of-function (GOF) or MEK1 loss-of-function (LOF) result in symmetric distribution of VENTX in mitotic PSCs and prolonged maintenance of refractory and undifferentiated state in vivo [14]. Thus, MEK1-mediated asymmetric cell distribution of VENTX and the resulting PSCs heterogeneity is mandatory for the pluripotency continuum in vivo ( Figure 2B,C), as well as SCF-β-TRCP ubiquitin-mediate global clearance of VENTX in committing embryonic cells at the onset of morphogenesis (i.e., gastrulation) in vivo [14,62].
Studies in hESCs strengthen the role of asymmetric cell division (ACD) in conferring diverging fates to mitotic pluripotent stem cells (PSCs), whereby one daughter cell maintains refractory/naive state (High-NANOG) whilst the other daughter becomes competent (Low-NANOG) to respond to pro-differentiation cues [63,64]. This would explain why inhibition of MEK1 activity, or undegradable forms of VENTX/NANOG, can lock vertebrate PSCs in a naïve/refractory pluripotent state [1,14,24,[59][60][61]. Furthermore, PSCs ACD suggests that the decision to commit is largely determined before the pro-differentiation cue is transcriptionally effective and can be predicted by a cell's pre-existing VENTX/NANOG protein distribution ( Figure 2C).
Recently, several studies point out the fundamental function of chromatin organization and poised enhancers, which represent distal regulatory elements that control the expression of major developmental genes, during the process of embryonic cell commitment and the establishment of stable epigenetic memory, which signs the exit from pluripotency state at the onset of gastrulation [68][69][70][92][93][94].
Altogether, these data suggest that PSCs transition from refractory/naïve to competent/formative state does not occurs in a synchronous and homogeneous manner, but rather PSCs represent an heterogenous population with biased competences to execute cell commitment depending on the levels of VENTX/NANOG protein and their physical interaction with POU5/OCT4. The significance of the asynchronous exit from refractory/naïve pluripotency and the heterogeneous competence of PSCs to commit need more attention in order to gain better insight on embryonic cell diversification and commitment.

Reprogramming Capacity of Developmental Potential Guardians Shapes Neural Crest Multipotency and Vertebrate Evolution
The neural crest (NC) is a vertebrate-specific cell population that is specified at the neuro-epithelial border (NEB) and that is empowered with a broader developmental potential than the ectodermal lineage of origin ( Figure 3). Together with the three primary germ layers (i.e., ectoderm, mesoderm, endoderm), NC is often referred as the "fourth germ layer" due to its multipotency [95]. The NEB cells are located into the ectodermal layer and are specified by inductive signals (e.g., FGFs, WNTs) from the overlaying mesoderm at gastrulation. The neuro-epithelial border cells (NEBs) form a transient embryonic cell population. Investigations on the transcriptional state of vertebrate NEBs by scRNA-seq analyses in several vertebrate models (i.e., zebrafish, Xenopus, chicken) have shown that NEBs express the NEB-specific gene regulatory network (GRN: PAX3/7, ZIC1, MSX1/2, TFAP2, ZNF703) ( Figure 3A) together with ectodermal-lineage specifiers (SOX2/3, GMNN, LHX5) [37,96,97]. These studies thus propose that NEBs initially transit through an ectodermal state. Consistent with this idea, Xenopus NEBs spontaneously differentiate into sensory neurons when isolated or transplanted [98], whereas mammalian (i.e., mouse and human) NEB stem cells are transcriptionally biased toward a neuroectodermal state (PAX3, ZIC,  At the end of gastrulation, vertebrate NEBs segregate in two distinct cell populations: (1) neurogenic placodal ectoderm, an unipotent neurogenic cell population of the anterior ectoderm that express SIX1 and EYA1 and give rise to sensory organs in the head, and (2) multipotent neural crest cells (NCCs), which express NC-specific GRN SNAI1/2, FOXD3, SOX8/9/10 (SOX-E family) ( Figure 3A) [95,99]. The transition from NEB-to-NC is dependent on WNT signaling pathway, as observed in vertebrate embryos and mammalian NEBs [77][78][79][80][81][82]100] During neurulation, NCCs loss neuroepithelial features through an epithelialmesenchymal transition (EMT) process, delaminate from the closing neural tube, migrate and then differentiate into a large spectrum of differentiated cell types: craniofacial bones, cartilages, muscles and the heart outflow mesenchyme, thyroid cells, secretory cells of the adrenal medulla, peripheral nervous system (PNS) and melanocytes ( Figure 3A) [95,99].
The loss of ectomesenchyme potential and craniofacial development after ventx2 LOF can be rescued by mammalian Nanog. Furthermore, ventx2 is sufficient and necessary to reprogram, together with NEB specifiers (i.e., PAX3, ZIC1), refractory differentiated epithelial ectodermal cells to immature and undifferentiated NCCs in vivo and in vitro [15]. Thus, when compared to the unipotent NEBs of invertebrates, VENTX/NANOG would empower the ectodermal NEBs of the proto-vertebrate ancestor with a new and broader competence to acquire alternative cell fates (ectomesenchyme) but permissive to the ectodermal (neurogenic) potential, thereby promoting the rise of multipotent NEBs/ NCCs, the "new head" and the vertebrate subphylum. [15] ( Figure 3D). Gradual acquisition of new and axial-specific regulatory sub-circuits further allowed evolution of gnathostome NCCs derivatives [104,112], in which reprogramming processes still link at the heart of the broad developmental potential of NNCs and their competence to acquire alternative cell fates during evolution [15,[112][113][114]. Consistent with these findings, POU5/OCT4 acts similarly to VENTX/NANOG in mouse developing NCCs [46]. Therefore, the rise of vsDPGs and multipotent NCCs should be considered a major step in the rise and evolution of vertebrates, as loss of vsDPGs in vertebrate multipotent NEBs can be interpreted as a regression to a primitive/atavistic unipotent condition that is functionally comparable to invertebrate NEBs.
They are mechanistically linked to endogenous in vivo reprogramming of neuroepithelial border cells towards the neural crest cells potential [15,46]. As a result, the epigenetic memory of vertebrate ancestor NEBs/NCCs became permissive for the activation of a pre-existing ectomesenchyme genetic program in parallel with ectoderm genetic program, whilst the epigenetic memory of invertebrate NEBs maintains a bias toward neurogenic program and ectomesenchyme remains exclusive of mesendoderm derivatives [105][106][107][108][109]. These data help to explain the origin and the biological relevance of vsDPGs as reprogramming factors [18,19,[29][30][31][32], which could arise, be selected and shaped by their function in NCCs. Further addition of new genetic sub-circuits during vertebrate evolution allowed the rise of new and alternative cell types from NCCs [104,[112][113][114][115][116] with a direct impact on the survival, adaptability and fitness of the organisms. Interestingly, other factors involved in control of early pluripotency are reactivated in NCCs (e.g., lin28A) that regulate NCCs multipotency downstream to the WNT signaling pathway, which is secreted from the dorsal neural tube and acts as "positional information" cue [100].
Due to their exploratory behavior and multipotency, NCCs have been recognized as a key contributor to phenotypic plasticity and evolvability [11,12,15,[117][118][119][120][121][122]. It will be interesting to understand whether changes in VENTX/NANOG and POU5/OCT4 spatiotemporal expression or protein activity (e.g., stability/degradation, physical interactions with co-factors and/or DNA) and distribution during mitoses (asymmetric cell division) within NCCs may contribute to cellular heterogeneity and fate choice during development, as observed in PSCs [15].

Developmental Potential Guardians Role in Neuro-Mesodermal Progenitors and Vertebrate Axial Length
As demonstrated by in situ hybridization (ISH) and scRNA-seq analyses in Xenopus and zebrafish embryos, vertebrate-specific developmental potential guardians (vsDPGs) are expressed in the posterior neuro-mesodermal progenitors (NMPs), an undifferentiated and multipotent cell population that participate in neural tube, trunk NC, somites and notochord formation during posterior axis elongation [16,22,39,48,131] (Figure 1C). A conserved feature of vertebrate NMPs is the co-expression of lineage-specific mesodermal (T/BRA) and neural (SOX2) regulators that, in concerto with differentiating cues (e.g., FGFs, WNTs, Retinoic Acid/RA, TGFβ) coordinate NMPs differentiation into neural/NC or mesodermal lineages [39,[74][75][76][131][132][133]. Functional analyses in Xenopus and zebrafish suggested that vsDPGs may participate in the maintenance of undifferentiated and quiescent NMPs since vsDPGs LOF lead to posterior axis truncation [17,21,22,134]. Similarly, mammal and chicken NMPs contribute to both neural, neural crest and mesodermal posterior derivatives and coexpress T/BRA and SOX2 [132,135,136]. Whereas zebrafish NMPs are multipotent up to bud stage/early somitogenesis and then a quiescent population of NMPs is maintained up to the end of somitogenesis [131,133,137], amniotes NMPs are maintained multipotent throughout somitogenesis [132,135,136] and this is likely due to the difference in proliferation and volumetric growth and proliferative phase(s) observed among amniotes versus non-amniotes [131]. Thus, vertebrate NMPs are competent to generate both neural/NC or mesodermal cell fates, and this potential is linked to the global proliferation rate, the stage of development, volumetric growth and, ultimately, the mode of development (namely, fast developmental rate in amniotes versus slow developmental rate in non-amniotes) [131,133,137].
Conditional manipulation of vsDPGs activity in gastrulating mouse embryo impact on posterior axis development through HOX expression [138,139]. Strikingly, modulation of POU5/OCT4 activity in mouse trunk NMPs impact on the growth and length of the posterior axis and induces a more posterior shift of HOX gene-expression boundaries in the extended trunk [138]. This lead to hypothesize that exacerbated posterior axis length observed in snakes (or eels) requires a sustained and prolonged maintenance of NMPs pool by vsDPGs in order to ensure a cellular source for forming neural tube, PNS, notochord and somites.
Therefore, it is tempting to speculate that vsDPGs might confer variability and evolvability to the organism by modulating the harmonious growth and shape of the posterior axis, as observed in a snake's body length. Since the timing of NMPs fate-restriction, cell rearrangement and volumetric growth of the posterior body axis varies among vertebrate species [131,133] the contribution of vsDPGs to the posterior axis development in vertebrates and their evolution, a consequence of their function in early pluripotency, need more accurate analysis in vivo. Comparative molecular-cellular analyses in closely related species, together with genomic data and mathematic/morphometric modelling, may ultimately tackle the question about the mechanism(s) by which evolution operate [11,12,115,116,[131][132][133][135][136][137][140][141][142][143][144].
Altogether, the roles of vsDPGs in NCCs/Head and NMPs/Trunk-Tail suggest an exciting and yet poorly explored contribution of vsDPGs to the global architecture of a living organism acting through tissue growth/shape to phenotype and evolution. Several studies have investigated the mechanism(s) at the heart of vertebrate phenotypical/morphological variability and evolution [117][118][119][120][121][122]131]. It will be interesting to understand whether the evolutionary dynamics of vsDPGs (i.e., duplications and loss), the variations in functional sites modulating their stability and their physical interactions [65,66], as well as modifications occurring in DNA sequences (e.g., promoters, enhancers) and the epigenetic modifications controlling their expression may help to infer causal mechanism(s) driving phenotypical variations among species. Since VENTX/NANOG and POU5/OCT4 evolutionary history show an intriguingly high degree of complexity [14,15,22,23] (Figure 1B), it will be important to understand whether their evolutive dynamics correlates with vertebrate phenotypical variations among species.
Whereas NANOG function has been extensively characterized in cancers, mainly for its expression and activity in cancer stem cells (CSCs), less is known about VENTX. Thus, the BioGRID Open Repository of CRISPR screens (BioGRID-ORCS database) [164] shows that VENTX is involved in growth/proliferation/resistance of several cancer cell lines (e.g., brain, pancreatic, renal and ovarian cancers) (Table 1). Intriguingly, VENTX is important for the proliferation of neural stem cells (NSCs), Glioma, and Glioblastoma (Table 1), thus suggesting an important role for VENTX both in normal/physiological and abnormal conditions of the human brain. Accordingly, Gene Expression Profiling Interactive Analysis (GEPIA2 database) based on primary tumors and normal samples from the TCGA and the GTEx databases [165] shows that VENTX is highly expressed in both low-grade-glioma (LGG) and glioblastoma (GBM) when compared to normal samples ( Figure 4A) and scRNA-seq data (available on Broad Institute Single Cell Portal database) further shows that VENTX is expressed in malignant GBM cancer cells ( Figure 4B) [166]. Survival rate based on VENTX expression levels ( Figure 4C) suggests that VENTX may be used in the prognostic of brain cancer development and patient survival as well. Thus, it would be relevant to better understand the function of VENTX in human brain physiology and carcinogenesis.
Survival rate based on VENTX expression levels ( Figure 4C) suggests that VENTX may be used in the prognostic of brain cancer development and patient survival as well. Thus, it would be relevant to better understand the function of VENTX in human brain physiology and carcinogenesis.
Altogether, vsDPGs play a role in carcinogenesis and therapeutic resistance, suggesting that they may present crucial targets to counteract cancer development, aggressiveness and relapse. It is intriguingly to note that vsDPGs are not Mut-driver-genes but, due to their aberrant expression levels, mainly represent Epigenetic-driver-genes (Epidriver-genes) [145]. How and when vsDPGs are reactivated in cancer cells is still poorly understood, as well as their relevance in establishing and maintaining cancer heterogeneity. Answering these questions may shed new light on the hierarchy of events leading to carcinogenesis and lead to more precise prophylactic and therapeutic approaches [147]. Due to their tumorigenic and reprogramming potential, it will be crucial to understand whether vsDPGs exclusively participate in cancer progression and resistance, or if they physiologically act earlier than expected during malignant transformation of a normal cell to a cancer cell of origin in vivo [175], thus prior to the appearance of a tumor mass and intertumoral cell heterogeneity. This should improve strategies for future preventive therapeutic approaches targeting cancer cell(s) at their early stages.

Conclusions and Perspectives
Vertebrate-specific developmental potential guardians (vsDPGs) have allowed us to delineate the molecular and cellular bases of the embryonic developmental potential in vivo. VsDPGs control the pluripotency continuum and competence for multilineage commitment in vertebrate pluripotent stem cells (PSCs), as well as later reacquisition of multipotency in neural crest cells (NCCs) through an "endogenous reprogramming" process. Furthermore, vsDPGs activity in vertebrate neuro-mesodermal progenitors (NMPs) participates in posterior axis growth and elongation. Altogether, these findings suggest that vsDPGs are key players of embryogenesis, but also suggest a less explored function in vertebrate evolution.
Since cell types expressing vsDPGs have been described as heterogeneous populations, it would be interesting to understand whether inter-specific variations in vsDPGs activity may impact on the growth, morphology and shape of the organism. In fact, most of these studies display technical limitations since the function of vsDPGs has been analyzed with limited spatial and/or temporal control. As an example, conditional vsDPGs GOF and/or LOF have been done in the whole organisms or cell types, with catastrophic effect on the organism or tissue(s). This may be useful for developmental and morphological analyses, but clearly this approach prevents an understanding of vsDPGs role at single cell level. For example, specific vsDPGs LOF in vertebrate NCCs allows one to elucidate the global endogenous reprogramming process conferring multipotency to vertebrate NEBs/NCCs. However fine spatial-temporal variations in vsDPGs activity in single NCCs should avoid the global catastrophic effects previously described and may be useful in elucidating their role in NCCs stemness and in shedding light on how the dynamics of undifferentiated state may impact on cell fate, tissue growth, and morphogenesis. It is interesting to note that WNT signaling pathway controls the early undifferentiated state of pre-migratory NCCs in chicken, as well as morphological and phenotypical variation of skull in fishes [100,118]. It has been thus suggested that variations in the activity (and downstream targets) of signaling pathways (e.g., WNTs, BMPs, SHH) may contribute in vertebrate morphological and phenotypical variation, adaptation and evolution [117][118][119][120][121][122]. Since vsDPGs can physically interact with downstream effectors (CTNNB1, SMADs) of such pathways (e.g., WNTs, BMPs) and are expressed in embryonic cells (i.e., NCCs and NMPs) contributing to morphological and phenotypical variation among vertebrates, it is reasonable to assume that variations in vsDPGs activity may contribute to vertebrate phenotypical variations and evolution, beyond their role in PSCs. Therefore, the development of new approaches allowing precise spatiotemporal single cell manipulation through LOF/GOF of target genes, such as vsDPGs, may help to understand the dynamics of tissue formation, morphogenesis, stemness during ontogenesis, and to infer phenotypical changes throughout evolution.
Recently, versatile optogenetic approaches have been developed to control gene expression and protein activity in a live animal at single cell level and with temporal resolution of a few seconds [129,176]. Among them, an original and fast optogenetic approach is based on a conditional ERT/caged Cyclofen-OH (cCYC) induction system in vivo, allowing the activation of specific genes either permanently (by using Cre-ERT/loxP system) or transiently (by using a Gal4-ERT/UAS system) [125][126][127][128][129][130]. Such an optogenetic approach allows precise spatiotemporal control of gene expression and protein activity at single cell level or in few cells. This strategy is compatible with the photoactivation of a wide variety of proteins. Therefore, optical methods open opportunities for the local spatiotemporal investigation of developmental processes, identification (and manipulation) of stem cells, and the study of carcinogenesis at single cell level in a live organism [127,176]. This strategy may provide crucial information about vsDPGs activity in development and differentiation/reprogramming in vivo with unprecedented spatiotemporal resolution.
Can vsDPGs affect the evolvability of the system acting on the unit of evolution (i.e., the cell) and how? Studies in vertebrates suggested that vsDPGs can confer cell plasticity/adaptability during development. Hence, precise manipulation of the spatiotemporal activity of vsDPGs at single cell level by optogenetic approaches may ultimately answer the question about the rise and changes of shapes in evolution. Quantitative and biophysical approaches at multiscale resolution (from single cell to tissue/organ), together with mathematical modelling and imaging, can ultimately help to elucidate how the fundamental and common laws governing morphogenesis during ontogeny and phenotypical variations throughout evolution intersect with vsDPGs activity in cells [11,12,115,116,124,[131][132][133]135,136,[140][141][142][143][144].
Furthermore, manipulating vsDPGs at single cell level may help to better understand the process of carcinogenesis, the malignant transformation of a normal cell to a cancer cell of origin, how heterogeneity arise among cancer cells and, ultimately, be useful to develop new and patient specific therapeutic approaches targeting each cell type via specific anticancer agents [145][146][147]175]. Therefore, the development of new tools allowing for genetic modifications (e.g., gene editing) with high spatiotemporal resolution would have a great impact on the whole scientific community and should revolutionize our knowledge of the rules governing animal development, morphogenesis, shape and evolution, as well as the origin of human diseases.