Stage-dependent differential gene expression profiles of cranial neural crest-like cells derived from mouse-induced pluripotent stem cells
- 11 Downloads
Cranial neural crest cells are multipotent cells that migrate into the pharyngeal arches of the vertebrate embryo and differentiate into various craniofacial organ derivatives. Therefore, migrating cranial neural crest cells are considered one of the most attractive candidate cell sources in regenerative medicine. We generated cranial neural crest like cell (cNCCs) using mouse-induced pluripotent stem cells cultured in neural crest-inducing medium for 14 days. Subsequently, we conducted RNA sequencing experiments to analyze gene expression profiles of cNCCs at different time points after induction. cNCCs expressed several neural crest specifier genes; however, some previously reported specifier genes such as paired box 3 and Forkhead box D3, which are essential for embryonic neural crest development, were not expressed. Moreover, ETS proto-oncogene 1, transcription factor and sex-determining region Y-box 10 were only expressed after 14 days of induction. Finally, cNCCs expressed multiple protocadherins and a disintegrin and metalloproteinase with thrombospondin motifs enzymes, which may be crucial for their migration.
KeywordsCranial neural crest Migratory neural crest iPS cells RNA sequencing Adamts
Stem cell-based tissue engineering is important in the field of oral science because it facilitates the regeneration of damaged tissues or organs [1, 2]. Various stem cell populations exhibiting regeneration potential in the craniofacial region have been identified. Of these, cranial neural crest cells (cNCCs) are considered one of the most important candidates owing to their role in craniofacial tissue organization . cNCCs comprise a multipotent population of migratory cells that are unique to the vertebrate embryo and can differentiate into various craniofacial organ derivatives [4, 5]. The neural crest (NC) can form teratoma when transplanted into immunocompromised animals . cNCC development involves three stages [7, 8, 9, 10]: the neural plate border stage, the premigratory stage, and the migratory stage. During the migratory stage, cNCCs delaminate from the posterior midbrain and individual rhombomeres in the hindbrain  and migrate into the pharyngeal arches to form skeletal elements of the face and teeth and contribute to formation of the pharyngeal glands (the thymus, thyroid, and parathyroid) . Therefore, cNCCs presumably represent a new treatment strategy for diseases of the craniofacial region .
Development from the premigratory to migratory stage proceeds swiftly ; thus, it is typically difficult to detect the precise time point of this transition . A recent transcriptome analysis of pure populations of migratory cNCCs cells expressing the sex-determining region Y-box 10 (Sox10) from chicks  has substantially improved our understanding of cNCC characteristics. However, whether these cells are in the migratory stage and how long it takes to promote embryonic stem (ES) cell-derived NCCs from the premigratory to migratory stage remains unclear. In recent years, the use of induced pluripotent stem (iPS) cells as a revolutionary approach to treat various medical conditions has garnered much attention [17, 18], and iPS cells as a cell source have shown several evident advantages over ES cells and primary cultured cNCCs in regenerative medicine . In addition, embryonic NC development depends on several environmental factors that influence the regulation of NC progenitors and timing of differentiation; therefore, it is important to elucidate the regulatory gene networks and expression profiles of mouse iPS (miPS) cell-derived cNCCs. Recent advances in next-generation RNA sequencing (RNA-seq) technologies have facilitated comprehensive analysis of gene expression profiles [19, 20, 21]. Therefore, in the present study, we used RNA-seq to investigate the gene expression landscape of cNCCs induced from miPS cells. We treated iPS-derived cells with cNCC induction medium for 14 days and performed RNA-seq experiments. Our results indicated that c-Myc; ETS proto-oncogene 1, transcription factor (Ets1); Sox10; a disintegrin and metalloproteinase domain metallopeptidase with thrombospondin motifs (Adamts) 2 and 8; protocadherin alpha (Pcdha) 2, 5, -7, -11, and -12; protocadherin alpha subfamily C,1 (Pcdhac1); and protocadherin gamma subfamily C,3 (Pcdhgc3) may be appropriate markers for migratory cNCCs induced from miPS cells.
Materials and methods
miPS cell culture
The miPS cells used in the present study (APS0001; iPS-MEF-Ng-20D-17 mouse-induced pluripotent stem cell line) were purchased from RIKEN BRC (Ibaraki, Japan) . The cells were incubated with inactivated murine embryonic fibroblast (MEF) feeder cells in Dulbecco’s Modified Eagle’s Medium (DMEM; Invitrogen, Carlsbad, CA, USA) supplemented with 15% KnockOut™ Serum Replacement (Invitrogen), 1% nonessential amino acids (Chemicon, Temecula, CA, USA), 1% l-glutamine (Chemicon), 1000 U/mL penicillin–streptomycin (P/S; Invitrogen), and 0.11 mM 2-mercaptoethanol (Wako Pure Chemical Industries Ltd., Osaka, Japan); 60-mm cell culture plates were used for passaging the cells at a density of 1 × 105 cells/plate. Cells were grown in 5% CO2 at 95% humidity, and the culture medium was changed each day.
Embryoid body (EB) formation and cNCC differentiation
O9-1 cell culture
O9-1 cells, a mouse cNCC line, were purchased from Millipore (Billerica, MA, USA) and cultured as a control, as previously described .
RNA extraction and quantitative reverse transcription polymerase chain reaction analysis (qRT-PCR)
Primers used for quantitative reverse transcription polymerase chain reaction (qRT-PCR)
Forward primer sequence
Reverse primer sequence
The cells were fixed with 4% paraformaldehyde (Wako Pure Chemical Industries Ltd.) for 15 min followed by methanol (Wako Pure Chemical Industries Ltd) for 5 min. After washing, the nonspecific binding of antibodies was blocked by adding 5% bovine serum albumin (BSA; Wako Pure Chemical Industries Ltd.) in a phosphate-buffered saline with 0.5% Triton X-100 (PBST) for 1 h. The cells were then incubated with the primary antibodies Snai1 1:50 for Rabbit polyclonal anti-Snai1 (26183-1-AP; Proteintech Group, Inc. Chicago, IL, USA.) and Sox10 1:500 for Mouse monoclonal anti-Sox10 (AMAb91297; Atlas Antibodies, Bromma, Sweden.) in PBST for 2 nights at 4 °C. We conducted that the positive control of Snai1 was O9-1 cells (cranial neural crest cells) and the positive control of Sox10 was DP cells (dental pulp cells). The negative control of Snai1 and Sox10 was SNL cells (fetus fibroblast cells) (Fig. S1).They were then incubated in the secondary antibodies fluorescein isothiocyanate Alexa Flour 488-conjugated affinity purified Goat anti-Rabbit IgG (H&L) (ab150077; Abcam, Cambridge, MA, USA) at a dilution of 1:500 for Snai1 and Alexa Flour 568-conjugated affinity purified Goat anti-Mouse IgG (H&L) (A-11004; Invitrogen) at a dilution of 1:500 for Sox10 in PBST for 1 h. Eventually, the cells were stained with 4,6-diamidino-2-phenylindole (DAPI; Sigma, Livonia, MI, USA) to visualize the nuclear DNA.
Total RNA from each sample was used to construct libraries with the Illumina TruSeq Stranded mRNA LT Sample Prep Kit (Illumina, San Diego, CA, USA), according to the manufacturer’s instructions. Polyadenylated mRNAs are commonly extracted using oligo-dT beads, following which the RNA is often fragmented to generate reads that cover the entire length of the transcripts. The standard Illumina approach relies on randomly primed double-stranded cDNA synthesis, followed by end-repair, dsDNA adapter ligation, and PCR amplification. The multiplexed libraries were sequenced as 125-bp paired-end reads using the Illumina Hiseq 2500 system (Illumina). Prior to performing any analysis, quality of the data was confirmed and read cleaning, such as adapter removal and simple quality filtering, was performed using Trimmomatic (ver. 0.32). Subsequently, the paired-end reads were mapped to the mouse genome reference sequence GRCm38 using the Burrows–Wheeler Aligner (ver. 0.7.10). The number of sequence reads mapped to each gene domain using SAM tools (ver. 0.1.19) was counted, and the reads per kilobase of transcript per 1 million mapped reads (RPKM) for known transcripts were calculated to normalize the expression level data to gene length and library size, thereby facilitating the comparison of different samples.
Gene expression profiles and immunohistochemistry of cNCCs derived from miPS cells
NC specifier transcription factors
Neural crest (NC) genes that have previously been examined in vivo
Neural crest (NC) transcription factors that have previously been examined in vivo
The transcription factor AP-2 alpha (Ap2) along with paired box 3 (Pax3) and zinc finger protein of the cerebellum 1 (Zic1), both of which are regulated by Ap2, were the most highly expressed genes in d7 cells (Fig. 3a). Pax6, which has been reported in human ES and iPS-derived NC cells (Tables 2, 3), was detected in both d7 and d14 cells, whereas Pax7, which has not previously been reported in the mouse NC, was detected in the d7 cells (Fig. 3a). In contrast, the homeobox genes gastrulation brain homeobox 2 (Gbx2), msh homeobox 1 (Msx1), distal-less homeobox 3 (Dlx3), Zic2, and Zic3 were not detected in d7 or d14 cells, and the homeobox genes Zic1 and Dlx5 were only expressed in the d7 cells, despite these having been reported in the NC of a range of species (Table 2); however, Meis homeobox 2 (Meis2) was expressed in both d7 and d14 cells.
Neural crest (NC) transcription factors that have previously been examined in vitro
The premigratory NC markers Ngfr, heart and neural crest derivatives expressed 2 (Hand2), Snai1, and Snai2 were only detected in the d7 cells; however, other premigratory NC markers, such as the platelet derived growth factor receptor, alpha polypeptide (Pdgfra); 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 4 (Pfkfb4); inhibitor of DNA binding 2 (Id2), Id3, and Id4; and nestin (Nes) were detected in both d7 and d14 cells (Fig. 3a).
Expression of migratory NC markers such as Sox5, -6, -8, -9, and -10, which encode members of the sex-determining region Y (SRY)-related high mobility group (HMG)-box family of transcription factors and are crucial in several aspects of NCCs, were detected in d7 or d14 cells. Sox10, a known marker for migratory cNCCs in various species (Table 2), was only detected in d14 cells similar to the other migratory NC markers. Twist family bHLH transcription factor 1 (Twist1), which is activated via various signal transduction pathways and is crucial for E-cadherin downregulation, as well as beta-1,3-glucuronyltransferase 1 (B3gat1/Hnk1), which plays a role in the formation of CD57 epitope, was detected in both d7 and d14 cells. In contrast, the expression of the trunk NC markers lit guidance ligand 1/2 (Slit1/2), which plays an important role in trunk NC cell migration toward ventral sites, was upregulated only in d7 cells (Fig. 3a).
Finally, expressions of tenascin C (Tnc), cadherin-6 (Cdh6), and ras homolog family member B (Rhob), all of which are related to cell adhesion and motility [81, 82, 83, 84, 85], were significantly increased in both d7 and d14 cells (Fig. 3b).
Metzincin superfamily zinc proteinase and protocadherin superfamily
Only Adam1a, -8, -10, and -12 were upregulated in both d7 and d14 cells (Fig. 4c, d); this is contrary to reports that the members of this family are important in NC migration and that Adam-10, -12, -15, -19, and -33 are expressed in the mouse NC . Moreover, various Adamts family genes, which are important for connective tissue organization and cell migration, were upregulated in either d7 or d14 cells (Fig. 4c, d). In particular, Adamts1 expression was markedly increased, whereas Adamts2 and -8 expressions, which are presumably important in cancer cell invasion [87, 88, 89], increased in the later stages of differentiation.
Expression of the protocadherin superfamily based on RNA sequencing data
In the present study, we derived cells from miPS which are closely migratory cNCCs genes. Previously, NCCs have been derived from ES or iPS cells using various approaches [91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110], and the protocol used in the present study was based on the methods outlined by Bajpai et al. ; however, few studies have investigated changes in the properties of cNCCs at different time points after induction.
In the present study, d7 and d14 cells expressed typical NC markers, such as Ngfr, Snai1, and Snai2. In contrast, O9-1 cells (controls) did not express Ngfr, suggesting that cNCCs derived from miPS cells are better than O9-1 cells for evaluating cNCC characteristics . Moreover, unlike O9-1 cells, d14 cells expressed markedly high levels of Sox10, which is considered a reliable marker for migratory cNCCs. Since cNCCs are involved in craniofacial tissue organization, several reports are available on their gene expression profiles; however, these reports show varying results with species and protocols. Moreover, cNCCs rapidly differentiate in the embryo ; thus, it is considerably difficult to synchronize the timing of isolation to a particular point during their development. Furthermore, migratory cNCCs intermingle with other cell types in the embryo, further complicating the isolation and characterization of a pure cell population. Consequently, there have been few reports on cNCC markers [16, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36]. Simões-Costa et al.  successfully isolated Sox10-positive cNCCs from chicken embryos and analyzed their gene profiles. Similarly, we detected Sox10 expression in d14 cNCCs. Reportedly, there are multiple NCC populations , and iPS cells can differentiate into numerous different NCC populations in the same culture. Therefore, this diversity in populations may explain the discrepancies in results; however, under the conditions used in the present study, c-Myc; Ets1; Sox10; Adamts2; Adamts8; Pcdha2, -5, -7, -11, and -12; Pcdhac1, and Pcdhgc3 may represent useful markers for migratory cNCCs. Furthermore, our results indicated that d7 cells were in the premigratory stage despite expressing numerous NC markers. Therefore, cNCCs derived from miPS cells required > 14 days to become migratory in vitro, and this duration is considerably longer than that observed in the mouse embryos in vivo under the same conditions .
The use of RNA-seq facilitates the normalization of expression levels of different genes, allowing comparisons between samples. In our triplicate experiments, none of the induced cNCCs expressed several homeobox genes considered to be expressed in the early stages of cNCC differentiation. In particular, we did not observe FoxD3 expression in either d7 or d14 cells, although it has been recognized as one of the key transcription factors in cNCCs . These contradictory results suggest that cNCCs derived from miPS cells express distinct gene regulatory networks. FoxD3, a pluripotent stem cell marker gene that plays an important role in maintaining pluripotency, is expressed at different time points in different cells, but its expression decreases in a time-dependent manner , indicating that FoxD3 may not be a key regulator in iPS-derived cNCCs. However, we speculate that iPS cells express sufficient levels of FoxD3 to differentiate into cNCCs.
Protocadherins belong to the cadherin superfamily and are involved in intercellular interactions , whereas metzincins are key proteinases that facilitate cell migration . Unfortunately, the abundances of members of these families hindered their analysis; however, because RNA-seq enabled us to comprehensively evaluate the gene expression profiles, we were able to focus on expressions of all procadherin and metazicin family members. As expected, we observed that several Adam and Adamts genes were upregulated, with most of the Admats genes showing significantly increased expression. The Adam genes with increased expression in cNCCs were membrane-bound, whereas Adamts genes which secreted proteinases, indicating that the expression of various Adamts may allow the matrix to be digested more efficiently and that each proteinase may be capable of digesting a different type of extracellular matrix protein . Therefore, the secretion of various Adamts and Pcdh proteins may play a crucial role in cNCC migration.
In summary, cNCCs derived from miPS exhibited RNA expression profiles that partly overlap with previously reported profiles. These cells may be useful for the regeneration of tissue formed by NCCs (osteoblast, melanocyte, and glial cells). We observed that although the resulting cNCCs exhibited several NC specifiers, they lacked some of the specifiers, indicating that a distinct molecular network may regulate gene expression in miPS-derived cNCCs. Moreover, our results indicated that c-Myc; Ets1; Sox10; Adamts2 and -8; Pcdha2, -5,-7, -11, and -12; Pcdhac1; and Pcdhgc3 may represent appropriate markers for migratory miPS-derived cNCCs. Finally, cNCCs expressed a wide spectrum of genes encoding Adamts family enzymes that may be crucial for their migration.
We are grateful to Professor T. Azuma, MD, PhD, Department of Biochemistry, and Professor T. Ichinohe, DDS, PhD, Department of Dental Anesthesiology, for their guidance. We also thank S. Onodera and A. Saito, Department of Biochemistry.
The funding was received by Ministry of Science and Technology (Grant nos. KIBANKENNKYU(B)18H03007 and 18K09753 KIBANKENNKYU(C)).
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflicts of interest.
- 6.McConnell AM, Mito JK, Ablain J, Dang M, Formichella L, Fisher DE, Zon LI (2018) Neural crest state activation in NRAS driven melanoma, but not in NRAS-driven melanocyte expansion. Dev Biol 449(17):30855–30862Google Scholar
- 13.Okuno H, Mihara FR, Ohta S, Fukuda K, Kurosawa K, Akamatsu W, Sanosaka TS, Kohyama J, Hayashi K, Nakajima K, Takahashi T, Wysocka J, Kosaki K, Okano H (2017) CHARGE syndrome modeling using patient-iPSCs reveals defective migration of neural crest cells harboring CHD7 mutations. eLife 6:e21114CrossRefPubMedPubMedCentralGoogle Scholar
- 25.Antonellis A, Bennett WR, Menheniott TR, Prasad AB, Lee-Lin SQ, NISC Comparative Sequencing Program, Green ED, Paisley D, Kelsh RN, Pavan WJ, Ward A (2006) Deletion of long-range sequences at Sox10 compromises developmental expression in a mouse model of Waardenburg-Shah (WS4) syndrome. Hum Mol Genet 15(2):259–271CrossRefPubMedGoogle Scholar
- 54.Alkobtawi M, Ray H, Barriga EH, Moreno M, Kerney R, Monsoro-Burq AH, Saint-Jeannet JP, Mayor R (2018) Characterization of Pax3 and Sox10 transgenic Xenopus laevis embryos as tools to study neural crest development. Dev Biol 444(17):30693–30700Google Scholar
- 75.Prasad MK, Reed X, Gorkin DU, Cronin JC, McAdow AR, Chain K, Hodonsky CJ, Jones EA, Svaren J, Antonellis A, Johnson SL, Loftus SK, Pavan WJ, McCallion AS (2011) SOX10 directly modulates ERBB3 transcription via an intronic neural crest enhancer. BMC Dev Biol 11:40CrossRefPubMedPubMedCentralGoogle Scholar
- 86.Tomczuk M, Takahashi Y, Huang J, Murase S, Mistretta M, Klaffky E, Sutherland A, Bolling L, Coonrod S, Marcinkiewicz C, Sheppard D, Stepp MA, White JM (2003) Role of multiple beta1 integrins in cell adhesion to the disintegrin domains of ADAMs 2 and 3. Exp Cell Res 290(1):68–81CrossRefPubMedGoogle Scholar
- 91.Okawa T, Kamiya H, Himeno T, Kato J, Seino Y, Fujiya A, Kondo M, Tsunekawa S, Naruse K, Hamada Y, Ozaki N, Cheng Z, Kito T, Suzuki H, Ito S, Oiso Y, Nakamura J, Isobe K (2013) Transplantation of neural crest-like cells derived from induced pluripotent stem cells improves diabetic polyneuropathy in mice. Cell Transplant 22(10):1767–1783CrossRefPubMedGoogle Scholar
- 92.Seki D, Takeshita N, Oyanagi T, Sasaki S, Takano I, Hasegawa M, Takano-Yamamoto T (2015) Differentiation of odontoblast-like cells from mouse induced pluripotent stem cells by Pax9 and Bmp4 transfection stem cells. Transl Med 4(9):993–997Google Scholar
- 93.Mizuseki K, Sakamoto T, Watanabe K, Muguruma K, Ikeya M, Nishiyama A, Arakawa A, Suemori H, Nakatsuji N, Kawasaki H, Murakami F, Sasai Y (2003) Generation of neural crest-derived peripheral neurons and floor plate cells from mouse and primate embryonic stem cells. Proc Natl Acad Sci USA 100(10):5828–5833CrossRefPubMedGoogle Scholar
- 96.Aihara Y, Hayashi Y, Hirata M, Ariki N, Shibata S, Nagoshi N, Nakanishi M, Ohnuma K, Warashina M, Michiue T, Uchiyama H, Okano H, Asashima M, Furue MK (2010) Induction of neural crest cells from mouse embryonic stem cells in a serum-free monolayer culture. Int J Dev Biol 54(8–9):1287–1294CrossRefPubMedGoogle Scholar
- 105.Liu Q, Spusta SC, Mi R, Lassiter RN, Stark MR, Höke A, Rao MS, Zeng X (2012) Human neural crest stem cells derived from human ESCs and induced pluripotent stem cells: induction, maintenance, and differentiation into functional schwann cells. Stem Cells Transl Med 1(4):266–278CrossRefPubMedPubMedCentralGoogle Scholar
- 107.Karbalaie K, Tanhaei S, Rabiei F, Kiani-Esfahani A, Masoudi NS, Nasr-Esfahani MH, Baharvand H (2015) Stem cells from human exfoliated deciduous tooth exhibit stromal-derived inducing activity and lead to generation of neural crest cells from human embryonic stem cells. Cell J 17(1):37–48PubMedPubMedCentralGoogle 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.