The evolutionally-conserved function of group B1 Sox family members confers the unique role of Sox2 in mouse ES cells
- 1.3k Downloads
In mouse ES cells, the function of Sox2 is essential for the maintenance of pluripotency. Since the Sox-family of transcription factors are well conserved in the animal kingdom, addressing the evolutionary origin of Sox2 function in pluripotent stem cells is intriguing from the perspective of understanding the origin of pluripotency.
Here we approach this question using a functional complementation assay in inducible Sox2-null ES cells. Assaying mouse Sox proteins from different Groups, we found that only Group B1 and Group G proteins were able to support pluripotency. Interestingly, invertebrate homologs of mammalian Group B1 Sox proteins were able to replace the pluripotency-associated function of mouse Sox2. Moreover, the mouse ES cells rescued by the Drosophila SoxNeuro protein are able to contribute to chimeric embryos.
These data indicate that the function of mouse Sox2 supporting pluripotency is based on an evolutionally conserved activity of the Group B1 Sox family. Since pluripotent stem cell population in developmental process could be regarded as the evolutional novelty in vertebrates, it could be regarded as a co-optional use of their evolutionally conserved function.
KeywordsPluripotent stem cells Sox2 Evolution Co-option
Pluripotency is a unique feature of the cells found in early vertebrate embryos. Pluripotent stem cells give rise to all cell types of the organism, including germ cells, but, unlike zygotes, they do not have the ability to give rise organisms autonomously . The pluripotent phenotype is primarily determined by the expression of a set of pluripotency-associated transcription factors, as demonstrated by the induction of pluripotency in somatic cells transfected with four transcription factors to give rise to induced pluripotent stem (iPS) cells . Of these four transcription factors, Oct3/4 (encoded by Pou5f1) and Sox2 are known to be essential for maintaining pluripotency in mouse embryonic stem (ES) cells [3, 4]. In contrast, the functions of Klf4 and Myc are dispensable for pluripotency, but primarily support self-renewal in the absence of the cytokine leukemia inhibitory factor (LIF) [5, 6, 7, 8].
Pluripotent stem cell populations have been definitively identified in mammalian embryos, but their presence in other vertebrate embryos remains unclear, with no pluripotent stem cell lines yet isolated from other taxa. Frog animal cap cells behave similarly to pluripotent cells, but have never been shown to yield stem cells capable of propagating in vitro . The absence of pluripotent stem cells is evident in ascidian embryos, since blastomeres exhibit mosaic behavior upon artificial separation . Likewise there is no evidence of pluripotent stem cells in well studied invertebrates, including fly and nematode. It therefore seems that the presence of a pluripotent stem cell population in the early embryo is a novelty exclusive to higher vertebrates.
Addressing the evolutionary origins of transcription factor functions coupled with pluripotency is an interesting challenge, since it may provide insights into the origins of the pluripotency-associated transcription factor network. Oct3/4 belongs to the POU family and its evolutionary history has been addressed in several studies [11, 12, 13]. The POU family of transcription factors are categorized into 6 classes and Oct3/4 (Pou5f1) is a member of class V. Although homologs of the other classes of POU family members can be found in the genomes of invertebrates, such as Caenorabditis elegans and Drosophila melanogaster, there is no class V POU family member in any invertebrate genome studied to date, strongly suggesting that Oct3/4 is a genetic novelty in the vertebrate genome. In fish genomes, Pou2 (Pou5f3) is the evolutionarily oldest member of the class V POU family, but is not syntenic with Oct3/4 in mammals, and the functional complementation assays using fish Pou2 in mouse ES cells revealed only a weak ability to substitute for the function of Oct3/4 in supporting pluripotency. In the case of amphibian, there are three class V POU family members in the genome of Xenopus tropicalis and these are found in tandem at a region syntenic with the Pou2 locus in the fish genome. One of these, Xlpou91, is known to rescue Oct3/4 function, providing a signature of molecular evolution. Monotreme Oct3/4 is a true ortholog that shares conserved synteny with other mammalian Oct3/4 genes, and functions to replace native Oct3/4 in mouse ES cells, although the homology of its POU domain at the amino acid level in comparison to the mouse ortholog is not markedly different from its homology to the POU domain of zebrafish Pou2. Interestingly, the monotreme genome also possesses the Pou2 ortholog with conserved synteny, indicating that Oct3/4 and Pou2 are paralogous. Recently, it was suggested that a lizard genome may include an ortholog of Oct3/4 in a conserved syntenic position, but to date no functional analysis has been reported. The above observations indicate that Oct3/4 is a relatively recent evolutionary acquisition whose ancestor evolved in vertebrates as Pou2, and subsequently underwent a duplication to generate Oct3/4 as a new class V POU family member.
Sox2 belongs to the Sry-related high mobility group (HMG)-box (Sox) family of transcription factors, whose members are characterized by a conserved HMG box DNA binding domain related to the mammalian testis determining factor Sry. Members of the Sox family bind to consensus DNA sequences and act as either transcriptional activators or repressors . In the mouse genome, there are 20 members of the Sox family categorized into 8 groups. Sox2 belongs to Group B1, which has two other members, Sox1 and Sox3. The Group B1 Sox family is well conserved and has been identified in virtually all multi-cellular vertebrate and invertebrate animals, where they share conserved functions in neural development. In pluripotent stem cells, it is known that Sox2 interacts with Oct3/4 to activate the transcription of target genes [15, 16]. When and how Sox2 acquired its function as a partner of Oct3/4 is an interesting but as yet unanswered question, although it has been suggested that this function is unique to Group B1 Sox family members. Interestingly, it has been shown that in Drosophila the Group B protein Dichaete interacts with the Class III POU protein Vvl during neural development and in this role the mouse Sox2 gene is able to functionally substitute for the fly protein , suggesting that Sox-POU interactions are ancient.
In the present study, we sought to reveal the molecular basis of the function of Sox2 in pluripotent stem cells and shed light on its evolutionary origins. We employed a functional complementation assay in mouse ES cells in which endogenous Sox2 alleles were disrupted by gene targeting, while a tetracycline-regulatable Sox2 transgene supports pluripotency . We tested mouse Sox family members from different groups along with Group B1 Sox family members from various organisms for their ability to substitute for mouse Sox2 in maintaining pluripotency. We find that a single evolutionarily-conserved amino acid is important for the function of SoxB1 proteins in maintaining mouse ES cell self-renewal. Finally, we report that the Drosophila SoxNeuro protein can substitute for the functions of mouse Sox2 required to maintain pluripotency.
Sox1, Sox3, and Sox15 can replace the function of Sox2 in mouse ES cells
The Sox expression vectors were then introduced into 2TS22C mouse ES cells, these lack endogenous Sox2 and are maintained by transgenic Sox2 expression regulated by the tet-OFF system . These ES cells maintain self-renewal in the absence of tetracycline (Tc) but cease self-renewal and undergo differentiation to trophectoderm upon addition of Tc. We introduced the Sox expression vectors using the PiggyBac system, that results in efficient integration of transgenes into the mouse genome , and individual lines were cultured with or without Tc. If an introduced Sox family member is capable of replacing the function of Sox2, the cells will be able to continue self-renewal over several passages in the presence of Tc (Fig. 1b). In line with this expectation, we found that expression vectors containing Sox1, Sox3, Sox4, Sox5, Sox8, Sox9, Sox11, Sox13, Sox14, Sox15, Sox18, and Sox21 yielded comparable numbers of stem cell colonies after culturing for seven days in the absence of Tc (Fig. 1c). In contrast, both Sox6 and Sox7 evidently produced differentiated cell colonies (Additional file 1: Figure S1B). We found that Sox6 expression resulted in many trophectoderm-like colonies, perhaps attributable to its strong function as a transcriptional repressor (Fig. 1a), which could compete with the function of Sox2. In contrast, Sox7 induced parietal endoderm-like cells, in line with our recent report . Stable cell lines carrying each Sox factor were dissociated and 1 × 104 cells were seeded in the presence of Tc, allowing selective propagation of the rescued stem cells. At the first passage, most of the Sox vector containing cells showed obviously differentiated phenotypes as did the cells containing the empty vector control, whereas Sox2 transfectants maintained an undifferentiated phenotype. At this period, only Sox1 and Sox3 supported stem cell colony formation as efficiently as Sox2. After the third passage in this condition, the rescuing ability of each Sox factor was evaluated by stem cell colony formation. Among the 15 Sox factors tested, only three (Sox1, Sox3 and Sox15) showed the ability to restore ES cell self-renewal although Sox15 supported very few stem cell colonies at the first passage (Fig. 1c).
Absence of Sox2 in these rescued ES cells was confirmed by immunostaining with anti-Sox2 antibody (Fig. 1d) and the expression of Sox1, Sox3 or Sox15 from the transgenes were confirmed by immunostaining with the relevant specific Sox antibody (Additional file 1: Figure S1C-E). We found that the rescued cells maintained expression of the pluripotency-associated transcription factors Oct3/4, Nanog and Klf4 (Fig. 1d), indicating they behave as self-renewing ES cells. These data indicate that the function of Sox2 in supporting mouse ES cell self-renewal is highly specific and shared with few other Sox family proteins.
A single conserved amino acid in the HMG box is responsible for the unique function of Sox2
Self-renewal in mouse ES cells is evolutionarily conserved in Group B1 Sox proteins
We then assessed the ability of the three Group B1 proteins to functionally substitute for mouse Sox2 in ES cells. We performed the rescue assay using 2TC22C ES cells described above and found that amphioxus SoxB1, ascidian Ci-SoxB1 and fly SoxN possess the ability to support self-renewal of Sox2-null ES cells (Fig. 3e). The immunostaining analysis of these ES cells confirmed the maintenance of Oct3/4, Nanog and Klf4 expression of with loss of endogenous mouse Sox2 (Fig. 3f). Therefore, the evolutionally-conserved function of the SoxB1 homologs is sufficient to maintain mouse ES cells self-renewal.
Fly SoxN maintains pluripotency of mouse ES cells
These ES cells, designated as Hae ES cells, grow normally and maintain the expression of Oct3/4 (Fig. 4b). We confirmed the presence of SoxN and absence of Sox2 by immunostaining with specific antibodies (Fig. 4b). When we injected these cells into blastocysts followed by uterine transfer to pseudo-pregnant females, we obtained chimeric embryos with obvious systemic distribution of DsRed-positive cells (Fig. 4c). These results indicate that Hae ES cells retain pluripotency and are able to contribute to many lineages in the developing mouse embryo.
Tissue-specific transcription factors act as the primary determinants of cell phenotypes in multi-cellular organisms. Genetic evidence from model organisms suggests that multiple transcription factors cooperate through direct and indirect interactions to determine a single phenotype. Several evolutionary novelties arose during the evolution of animals to generate the complex cellular architectures that characterize these organisms. Such events were presumably coupled with the evolution of new transcription factor functions. The presence of a pluripotent stem cell population during early developmental stages is a unique feature of higher vertebrates that facilitates flexible developmental processes. We and others previously showed that the Pou5f1 gene encoding Oct3/4, one of the factors essential for conferring the pluripotent phenotype, may have co-evolved with the acquisition of a pluripotent cell population during vertebrate evolution [11, 12, 13]. In contrast, in the case of its partner Sox2, not only do all vertebrates possess Sox2 orthologs, but SoxB1 genes sharing high sequence homology in the HMG box DNA binding domain are found across all metazoan genomes, indicating an very early origin during animal evolution . In the present study we show that invertebrate SoxB1 homologs possess the ability to functionally substitute for Sox2 and support the self-renewing state in mouse ES cells, indicating that the unique function of Sox2 in pluripotent stem cells is based on a conserved function of metazoan SoxB1 proteins.
We identified a single amino acid, K57 positioned in the third α-helix of the DNA binding domain, is responsible for conferring the unique function of Sox2 in supporting pluripotency, raising questions about the role of this amino acid. Interestingly, a previous report indicated that this amino acid is positioned in the interaction surface with Oct3/4 in ternary Sox-Oct complex formed on the Utf1 enhancer . That study also showed that two amino acid substitutions, including K57E, abolished the interaction with Oct3/4, which agrees well with our finding that mutated Sox2 proteins carrying the K57E substitution were no longer able to maintain pluripotency. Interestingly, this unique amino acid is conserved in all group B Sox family members, in mouse Group G proteins and in the invertebrate SoxB proteins we assayed. The evidence that Sox15 is able to replace the function of Sox2 supports the importance of this amino acid. Moreover, as reported previously, the introduction of this amino acid to Sox17 conferrs the ability to functionally substitute for Sox2 in iPS cell assays , supporting the view that this amino acid, at least in part, is able to confer unique pluripotency functions of Sox2. Clearly K57 is not a sufficient for Sox2-specific pluripotency functions in ES cells: for example, the Group B2 proteins Sox14 and Sox21 also possess this amino acid but do not support self-renewal of Sox2-null ES cells. However, this is not surprising since these two proteins are known transcriptional repressors , which we confirmed in our HeLa cell reporter assay. In contrast, all three invertebrate Group B1 proteins we tested act as transcriptional activators in the same assay. These data suggested that the HMG-box with K57 and a transactivation domain are minimal requirements for Sox2 activity supporting pluripotency. We previously reported the failure of hydra SoxB to replace the function of Sox2, which could be due to the lack of proper transactivation domain that can act in mouse ES cells since it also possess K57 .
The functional conservation of SoxB proteins in supporting ES cell self-renewal and pluripotency is remarkable but not unique, with several reports of conserved function between mouse and fly now published. In a famous example, expression of the mouse Pax6 gene can induce the formation of ectopic eye-like structures, mimicking the phenotypes elicited by the fly homolog eyeless . The mutant phenotype of Drosophila tinman was partially rescued by the transgenic expression of a mouse homologue Nkx2.5 . In the case of Sox2, we have shown that mouse Sox2 transgenes can rescue aspects of neural development in Dichaete null mutants . However, in all these cases the functional conservation is manifest in homologous developmental processes: Pax6 and eyeless normally function during eye development, Nkx2.5 and tinman participate in heart development, and both Sox2 and Dichaete have well documented roles in neural development. In contrast, here we show that Drosophila SoxN can replace Sox2 function in pluripotent stem cells, for which there is no homologous cells or tissues in Drosophila. To our knowledge a comparable case has only been reported with the mouse GroupE Sox gene Sox10 and its role in neural crest development. In this case the mouse gene was replaced by the Drosophila Group E gene Sox100B and, as we find here, although the fly lacks any conterpart to the vertebrate neural crest, the fly gene is able to provide substantial function in the absence of Sox10 . We believe these cases provide evidence that supports the idea that conserved functions of homologous genes can be integrated into new functions acquired during evolution that generate biological novelty.
How was the conserved Sox2 function integrated into the pluripotency-associated transcription factor network? Previous work showed that the expression of Sox2 in ES cells is supported by the ES-specific enhancers. SRR1 and SRR2, located at the 5′ and 3′ proximal regions of Sox2, were first identified . SRR1/N-2 is regulated by Oct3/4  and SRR2 is activated by Oct3/4 and Sox2 in ES cells . Comparative genome analysis of human, mouse and chicken Sox2 orthologs revealed several conserved elements that possess this enhancer activity . Of these, SRR2 is conserved in chicken but not in Xenopus and zebrafish, suggesting its evolutionally new origin . In addition, the recent reports demonstrated that the distal super-enhancer element possesses higher contribution to the transcriptional activation of Sox2 in ES cells although the evolutional origin of this element has not been characterized [34, 35]. We consider that the acquisition of novel regulatory elements would be necessary since Sox2 has evolutionarily conserved functions essential for neural development that restrict the flexibility to acquire new functions by modifying the protein sequence. Interestingly, Sox2 function is also important for self-renewal of mouse trophoblast stem (TS) cells. We previously reported that Sox2 has an alternative partner, Tfap2c rather than Oct3/4, and regulates different sets of target genes in ES and TS cells . Since the trophectoderm lineage gives rise to the placenta, which is an obvious evolutional novelty acquired by eutherians, it will be of interest to dissect the function of Sox2 in mouse TS cells using a approaches similar to those described here. Such an analysis should provide further insights into the molecular mechanisms used during evolution to generate new functions for tissue-specific transcription factors and regulatory networks without increasing the number of genes in the genome.
We demonstrate that the function of Sox2 to maintain self-renewal of mouse ES cells is specifically shared by Group B1 of Sox-family members. The invertebrate SoxB1 members also possess this activity although they lack pluripotent stem cell population in developmental process, indicating that the function of SoxB1 factor in pluripotent stem cells is a co-optional use of the evolutionally-conserved function.
All ORFs for mouse Sox family members and Group B Sox factors from other species were isolated by PCR with KOD-Fx (Toyobo) or Pfx polymerase (Invitrogen) using primer pairs with either mouse ES cDNA, FANTOM clones or genomic DNA as listed in Additional file 3: Table S1. Amplicons were subcloned into the PiggyBac expression vector pPBCAG-cHA-IB, which was made by insertion of CAG-IB into the PiggyBac transposon unit . Mutagenesis of mouse Sox2 and other Sox family members was performed by PCR with KOD-Fx with the primer pairs listed in Additional file 4: Table S2. The ORFs of all of vectors were sequenced and confirmed free of unexpected mutations.
ES cell culture and complementation assay
2TS22C ES cells were cultured in GMEM supplemented with 10 % Knockout Serum Replacement (KSR; Invitrogen), 1 % fetal calf serum (FCS), 1 × non-essential amino acids (Nakarai), 1 mM Sodium pyruvate (Nakarai), 10−4 M 2-mercaptoethanol and 103 U/ml of mouse LIF on gelatin coated surface. For the complementation assay, 3 × 104 2TS22C ES cells were seeded in wells of 48 well plates. The following day, cells were transfected with 1 μg of the PiggyBac Sox expression vector and 1 μg of pCAGGS-PBase  using Lipofectoamine 2000 (Invitrogen), and replated into four wells of a 12-well plate. Cells were selected with 10 μg/ml of Blasticidin S (Invivogen) from one to seven days after transfection. Three wells of cells were stained with Leischman stain to count the numbers of stem cell colonies. One well of cells was dissociated and 3 × 103 cells seeded into a well of a 12-well plate either in the presence or absence of 1 μg/ml of tetracycline. After six days, one-fifth of the dissociated cells were replated into a well of a 12-well plate followed by culture with tetracycline for 6 days. Stem cell colonies were scored by Leischman staining as well as re-seeding into either 12-well or 48-well plates for RNA preparation and immunostaining, respectively.
HeLa cells were cultured in GMEM with 10 % FCS. 104 HeLa cells were seeded into a well of a 96-well plate. The following day, three wells of cells were transfected with 0.5 μg of the PiggyBac Sox expression vector, 0.5 μg of either SOX (AACAAAG) × 7 (tandem repeat)-tk-luc or SAC (CCGCGGT) × 7 (tandem repeat)-tk-luc and 10 ng of pRL-SV (Promega) followed by the culture for 24 h. The cells were then tested for luciferase activity using the Dual luciferase assay kit (Promega) with Centro LB 960 luminometer (Berthold).
Cells were fixed with 4 % paraformaldehyde in PBS for 30 min at 4°C, followed by permealization with 0.2 % TritonX100 in PBS for 10 min at RT. These cells were incubated with the following primary antibodies overnight at 4°C; mouse monoclonal anti-Sox2 (R&D Systems, MAB2018), 1:1000; goat polyclonal anti-Sox2 (Neuromics, GT15098), 1:1000; mouse monoclonal anti-Oct3/4 (Santa Cruz, 5279), 1:1000; rat monoclonal anti-Nanog (e-Bioscience, MLC-51), 1:1000; rabbit polyclonal anti-Klf4 (Santa Cruz, 20691), 1:300; anti-SoxN antisera (Ferrero et al., 2014), 1:300; rabbit polyclonal anti-Sox1 (Chemicon, ab5768), 1:300; rabbit polyclonal anti-Sox3 (Santa Cruz 20089), 1:300; goat polyclonal anti-Sox15 (Santa Cruz, 17354), 1:300: goat polyclonal anti-Sox17 (R&D Systems, AF1924), 1:300. After washing, cells were incubated with appropriate secondary antibodies conjugated with Alexa-488 for 1 h at RT with Hoechst 33258 and fluorescent images were taken on an Olympus OX-71 equipped with a CCD camera.
Establishment of Hae ES cells
The Drosophila SoxN knock-in vector was generated by replacing loxP-mouse Sox2 ORF-IRES-Bsd-pA-loxP in a previously described knockout vector  with Drosophila SoxN ORF-IRES-neo-pA. The linealized plasmid DNA of this knock-in vector was transfected into 2CG2 ES cells by electroporation followed by selection with G418 and Gancyclovir. Surviving clones were screened by genomic DNA-PCR to identify knock-in cell lines. Sucessful knock-in ES cells were cultured with dexamethasone for activation of CreGR followed by the replating at clonal density with G418 and puromycin. The DeRed-positive clones were isolated and viability in culture with blasticidin S was tested. Blasticidin S-sensitive clones were screened by genomic DNA-PCR to obtain Sox2-null ES cell lines maintained by SoxN, we designate these as Hae ES cells.
Production of chimeric embryos
Dissociated Hae ES cells were introduced into a C57BL6 blastocyst by microinjection, which was then transferred to the uterus of a pseudopregnant female ICR mouse. Embryos were collected at 13.5 dpc and the contribution of the injected ES cells to the chimera evaluated by fluorescence microscopy. All animal experiments conformed to our Guidelines for the Care and Use of Laboratory animals and were approved by the Institutional Committee for Laboratory Animal Experimentation (RIKEN Kobe Institute).
We thank Dr Austin Smith and Ge Guo (University of Cambridge, UK) for providing us the PiggyBac vector system.
Availability of data and materials
The datasets and materials generated during the current study available from the corresponding author on reasonable request.
HN designed the study, performed cell culture experiments and wrote the manuscript. AN, MU, MS isolated Drosphila, amphioxus and ascidian SoxB1 homologs, respectively. SK performed genome analyses to identify invertebrate SoxB1 homologs. SR contributed to the immunostaining with anti-SoxN antisera and performed critical check of the manuscript. SO carried out embryo manipulation and analyzed chimeric embryos. All authors read and approved the final manuscript.
The authors declare that they have no conflict of interest.
Consent for publication
Ethics approval and consent to participate
All animal experiments conformed to our Guidelines for the Care and Use of Laboratory animals and were approved by the Institutional Committee for Laboratory Animal Experimentation (RIKEN Kobe Institute).
- 15.Sharov AA, Masui S, Sharova LV, Piao Y, Aiba K, Matoba R, Xin L, Niwa H, Ko MS. Identification of Pou5f1, Sox2, and Nanog downstream target genes with statistical confidence by applying a novel algorithm to time course microarray and genome-wide chromatin immunoprecipitation data. BMC Genomics. 2008;9:269.PubMedPubMedCentralCrossRefGoogle Scholar
- 27.Hemmrich G, Khalturin K, Boehm AM, Puchert M, Anton-Erxleben F, Wittlieb J, Klostermeier UC, Rosenstiel P, Oberg HH, Domazet-Loso T, et al. Molecular signatures of the three stem cell lineages in hydra and the emergence of stem cell function at the base of multicellularity. Mol Biol Evol. 2012;29:3267–80.PubMedCrossRefGoogle Scholar
- 30.Cossais F, Sock E, Hornig J, Schreiner S, Kellerer S, Bosl MR, Russell S, Wegner M. Replacement of mouse Sox10 by the Drosophila ortholog Sox100B provides evidence for co-option of SoxE proteins into vertebrate-specific gene-regulatory networks through altered expression. Dev Biol. 2010;341:267–81.PubMedCrossRefGoogle Scholar
- 33.Iwafuchi M, Yoshida Y, Onichtchouk D. M leichsenring, W Driever, T Takemoto, M Uchikawa, Y Kamachi, H Kondoh: The Pou5f1/Pou3f-dependent but SoxB-independent regulation of conserved enhancer N2 initiates Sox2 expression during epiblast to neural plate stages in vertebrates. Dev Biol. 2011;352:354–66.CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.