TISSUE‐SPECIFIC STEM CELLS
Alternative Generation of CNS Neural Stem Max‐Planck‐Institute for Brain Research, Research Group Deve‐ Cells and PNS Derivatives from Neural Crest lopmental Neurobiology, Frank‐ Derived Peripheral Stem Cells furt/M, Germany; bInnsbruck Med‐ ical University, Institute for Neu‐ roscience, Innsbruck, Austria; cIns‐ Marlen Webera, Galina Apostolovab, Darius Widerac, Michel titute of Cell Biology, University of Mittelbronnd, Georg Dechantb, Barbara Kaltschmidtc, e, Bielefeld, Bielefeld, Germany; dE‐ Hermann Rohrera dinger Institute, (Neurological In‐ stitute), Frankfurt/M, Germany; e Key words. neural crest • neural stem cell • PNS • CNS • reprogramming Molecular Neurobiology, Universi‐ ty of Bielefeld, Bielefeld, Germany ABSTRACT Corresponding author: Hermann Neural crest‐derived stem cells (NCSCs) from the embryonic PNS can be re‐ Rohrer, Max‐Planck‐Institute for programmed in neurosphere culture (NS) to rNCSCs that produce CNS prog‐ Brain Research, Research Group eny, including myelinating oligodendrocytes. Using global gene expression Developmental Neurobiology, analysis we now demonstrate that rNCSCs completely lose their previous Max‐von‐Laue‐Str. 4, 60438 Frank‐ PNS characteristics and acquire the identity of neural stem cells derived from furt/M, Germany, e‐mail: Her‐ embryonic spinal cord (SCSCs). Reprogramming proceeds rapidly and results
[email protected], in a homogenous population of Olig2‐, Sox3‐ and Lex‐positive CNS stem cells. Phone: +49‐69‐96769‐319, Fax: Low‐level expression of pluripotency inducing genes Oct4, Nanog and Klf4 +49‐69‐96769‐201 argues against a transient pluripotent state during reprogramming. The ac‐ quisition of CNS properties is prevented in the presence of BMP4 (BMP Received April 28, 2014; accepted NCSCs) as shown by marker gene expression and the potential to produce for publication September 06, PNS neurons and glia. In addition, genes characteristic for mesenchymal and 2014; perivascular progenitors are expressed, which suggests that BMP NCSCs are directed towards a pericyte progenitor/mesenchymal stem cell (MSC) fate. ©AlphaMed Press Adult NCSCs from mouse palate, an easily accessible source of adult NCSCs, 1066‐5099/2014/$30.00/0 display strikingly similar properties. They do not generate cells with CNS cha‐ racteristics but lose the neural crest markers Sox10 and p75 and produce This article has been accepted for MSC‐like cells. These findings show that embryonic NCSCs acquire a full CNS publication and undergone full identity in neurosphere culture. In contrast, MSC‐like cells are generated peer review but has not been from BMP NCSCs and pNCSCs, which reveals that postmigratory NCSCs are a through the copyediting, typeset‐ source for MSC‐like cells up to the adult stage. ting, pagination and proofreading STEM CELLS 2014; 00:000–000 process which may lead to differ‐ ences between this version and the Version of Record. Please cite this article doi: 10.1002/stem.1880 gratory neural crest are characterized by different com‐ INTRODUCTION positions of transcription factors that form a neural crest gene regulatory network [2, 3]. The neural crest gives rise to a broad array of deriva‐ The neural crest contains self‐renewing and multipotent tives including neurons and glia of the peripheral nerv‐ stem cell‐like cells that are present in migratory neural ous system, endocrine adrenal chromaffin cells, carti‐ crest and in postmigratory locations like gut and peri‐ lage and bone of the face, heart vasculature and mela‐ pheral ganglia [4]. Much of the current interest in neur‐ nocytes of the skin [1]. Neural crest cells are specified at al crest cells is due to their multipotent stem cell‐like the border between neural plate and non‐neural ecto‐ characteristics and their potential for regenerative med‐ derm, delaminate during neurulation from the dorsal icine [5]. We are referring to multipotent, self‐renewing neural tube to subsequently migrate to their various cells derived from embryonic, postnatal and adult PNS target sites. Neural plate border, premigratory and mi‐ a
STEM CELLS 2014;00:00‐00 www.StemCells.com
©AlphaMed Press 2014
2 tissues as neural crest‐derived stem cells (NCSCs). [6]although in the absence of melanocyte progeny these cells may also be considered as neural crest‐ derived progenitor cells [5]. NCSCs differ from neural stem cells in the central nervous system (CNS) by the expression of specific markers like Sox10, HNK1, p75, and by their potential to generate PNS progeny and mesenchymal derivatives. However, there is evidence that embryonic NCSCs can be reprogrammed to CNS neural stem cells in neurosphere cultures under the influence of culture conditions with heparin‐enhanced FGF‐signaling, developed for the maintenance of CNS stem cells [6‐8].. Reprogrammed NCSCs (rNCSCs) give rise to CNS progeny in vitro and in vivo, upon transplan‐ tation into embryonic and postnatal brain. The ability to produce myelinating oligodendrocytes that repair brain lesions in vivo is of considerable interest for regenera‐ tive medicine as selfrenewing rNCSCs can be expanded in culture and are not genetically modified [6, 7]. The aim of the present study was to characterize the generation of CNS neural stem cells from neural crest‐ derived stem cells in the PNS in more detail with re‐ spect to the underlying cellular and molecular mechan‐ isms. In addition, NCSCs from the adult PNS were ana‐ lyzed for their ability to acquire CNS properties to neu‐ rosphere cultures. Using global gene expression analysis we demonstrate that rNCSCs are indistinguishable from neural stem cells derived from embryonic spinal cord (SCSCs) and show no traces of their previous PNS identi‐ ty. Reprogramming proceeds rapidly by a simultaneous down‐ and upregulation of PNS and CNS markers, re‐ spectively, and leads to a homogenous population of Olig2+, Sox3+ and Lex+ CNS stem cells. Reprogramming does not involve an epigenetic resetting to a pluripotent state, but rather direct transdifferentiation from PNS to CNS stem cells. PNS stem cells can be prevented from acquiring a CNS identity by the application of BMP4 to the culture me‐ dium. The PNS identity of BMP NCSCs is shown by the expression of neural crest marker genes and their po‐ tential to generate characteristic PNS progeny, i.e. peri‐ pheral neurons and Schwann cells. In addition, neural crest derivatives with the characteristic expression pat‐ tern of pericyte progenitors and mesenchymal stem cells (MSCs) are generated in the presence of BMP4. As cultured MSCs and pericytes show a very similar pheno‐ type, ontogentic relationship, marker gene expression and differentiation potential we refer to these cells in the following as MSC‐like cells [9‐11]. Interestingly, NCSCs from adult mouse palate (pNCSCS) were also directed towards pericyte fates under conditions that allowed CNS reprogramming in embryonic NCSCs. These findings demonstrate that the potential of neural crest cells to generate MSC‐like cells [12, 13] is maintained in postmigratory embryonic and adult NCSCs. www.StemCells.com
Generation of CNS stem cells from PNS progenitors
MATERIALS AND METHODS Cell culture. Neurosphere cultures from embryonic DRG and spinal cord and from adult palate were generated with modifications of published procedures [6, 14], as detailed in the online Supporting information. Immunocytochemistry. Protocols of immunostainings for cell surface and intracellular antigens of short‐term cultures and differentiated neurosphere cultures are described in detail in Supplemental Materials. RT‐PCR. Total RNA was isolated from P3 rNCSCs, BMP NCSCs, SCSCs, pNCSCs and mouse ESCs (kindly provided by A. Smith, University of Cambridge, UK) using the RNeasy Kit. cDNA from RNA was synthesized using the M‐MLV Reverse Trancriptase Kit (Invitrogen). PCR pro‐ tocol and primer list (STable 1) are available in the on‐ line supporting information. Microarray analysis. Total RNA was isolated using TRI‐ zol reagent (Life Technologies), purified by RNeasy Mi‐ nElute kit (Quiagen) and analyzed for RNA integrity (2100 Bioanalyzer). RNA samples from three indepen‐ dent cell cultures of rNCSCs, BMP NCSCs, SCSCs and pNCSCs were labeled and hybridized to Affymetrix Mouse Genome 430 2.0 Arrays. Microarray hybridiza‐ tion was performed at the Expression Profiling Unit of the Medical University Innsbruck. Pre‐processing and differential expression analysis were performed in R (http://www.r‐project.org) using packages from the Bioconductor project [15]. Raw expression values were normalized and summarized using the GCRMA method [16]. The moderated t‐test [17] was applied to assess the significance of differential expression between the compared groups; the resulting p‐values were adjusted for multiple hypothesis testing following the method of Benjamini and Hochberg [18] for a strong control of false discovery rate. A gene was considered as differen‐ tially expressed when it was up‐ or down‐regulated at least 1.5‐fold and its adjusted p‐value was lower than 0.05. Gene ontology analysis was carried out to study the biological function of genes differentially expressed between rNCSCs, BMP NCSCs and pNCSCs using DAVID [19]. Principal component analysis (PCA) and unsuper‐ vised hierarchical clustering (HC) were performed on samples based on normalized expression of all genes. To correct for batch effects either ComBat package was used [20] or SCAN ‐ a single‐sample normalization ap‐ proach which allows concurrent use of independent gene expression datasets [21].. Profiling results have been deposited at the Gene Expression Omnibus (GEO) under accession GSE57003. The following publicly avail‐ able expression data downloaded from GEO were used for comparison with the microarray data generated in this study: GSE50824 (GSM1230380, GSM1230381, GSM1230382, GSM1230383); GSE8034 (GSM200043, GSM200044, GSM200045). ©AlphaMed Press 2014
3
RESULTS Olig2 is induced in DRG‐derived neurospheres under proliferation conditions (EGF/FGF) DRG‐derived NCSCs are reprogrammed in neurosphere cultures to rNCSCs that give rise under differentiation conditions to CNS progeny including Olig2+ cells with oligodendrocyte morphology and markers O4, CNPase and Sox10. [6] (SFig. 1A). To analyze the reprogramming rd efficiency we use short‐term cultures of 3 passage (P3) neurospheres in proliferation medium and characterize the properties of rNCSCs. Virtually all cells in P3 rNCSC neurospheres express the CNS‐specific gene Olig2 (99±0.4%; n=3; mean±sem; Fig. 1A; SFig. 1B), whereas the PNS marker p75 and the Schwann cell/oligodendroycte marker O4 are not de‐ tectable. Sox10+ cells, Tuj1+ neurons and GFAP+ glia are present in very low numbers (Fig. 1A, a‐f). Comparable results are obtained for P3 neurospheres derived from E12.5 mouse spinal cord (SCSCs) analyzed for compari‐ son (Fig. 1A, g‐l). These data indicate that reprogram‐ ming of embryonic DRG‐derived NCSCs occurs under conditions of neurosphere cell proliferation and results in a homogenous Olig2+ CNS stem cell population that does not maintain p75/Sox10 PNS markers. As these conclusions are based on the analysis of a small number of marker genes, genome‐wide expres‐ sion profiles of P3 rNCSCs and SCSCs were analyzed by Affymetrix microarrays. Notably, the gene expression profiles of SCSCs and rNCSCc are virtually identical (Fig. 1B). Among the 45.101 probe sets on the array, only one (Efcab7) was differentially expressed following the established criteria for differential expression (adjusted p‐value