Neuralized Mouse Embryonic Stem Cells Develop Neural Rosette-Like Structures in Response to Retinoic Acid and Produce Teratomas in the Brains of Syngeneic Mice

Cheryl L. Dunham; Mark D. Kirk

Abstract

Several induction protocols can direct differentiation of mouse embryonic stem cells （ESCs） to become neural cells. The B5 and B6 mouse ESC lines display different growth patterns in vitro， and when grown as adherent cultures， the B6 ESCs proliferated at a significantly lower rate than B5 ESCs. Remarkably， after a neural induction protocol that includes removal of LIF and addition of retinoic acid （RA）， mature B6 embryoid bodies （EBs） displayed a unique neural rosette-like morphology. On Day 8 of neural induction， B6 EBs revealed mature neuronal markers localized primarily to cells in the center of the EBs and glial markers expressed both in centrally and peripherally located cells. In contrast to B5 cells， when neuralized Day 8 B6 EB cells were dissociated and transplanted into the left striatum of syngeneic C57BL/6 mouse brains， teratomas formed. In addition， teratomas established from undifferentiated B6 cells grew more rapidly and achieved larger volumes when compared to those produced by Day 8， neuralized B6 EBs. The slow growth rate of B6 cells in vitro may have contributed to incomplete neuralization， formation of neural rosette-like structures， and a propensity to form teratomas.

Key words： C57BL/6 mouse； Embryoid bodies； Embryonic stem cells； Neural stem cells； Neural induction； Striatum； Teratoma

INTRODUCTION

Stem cells are classified in part by their ability to produce various types of specialized cells （Eisenberg & Eisenberg， 2003）. Pluripotent stem cells can differentiate into cells of all three primary embryonic germ layers， ectoderm， mesoderm， and endoderm and can contribute to germ line transmission. Embryonic stem cells （ESCs） are pluripotent and are found in the inner cell mass of the blastocyst（Eisenberg & Eisenberg， 2003； Mitalipov & Wolf， 2009； Solter， 2006； Alison & Islam， 2009）.

Due to their broad potential for differentiation， pluripotent stem cells are of particular importance in the fields of developmental biology and regenerative medicine. Before ESC lines were established， pluripotent cell lines were obtained from teratocarcinomas， tumors that can produce tissue types representative of all three primary embryonic germ layers （Solter， 2006； Li & Tanaka， 2011； Kleinsmith & Pierce， 1964）. In early studies， the 129SvJ mouse strain was found to have a 1% spontaneous rate of occurrence of testicular teratomas （Solter， 2006； Stevens & Little， 1954）. Cells from some of these teratomas could be extracted， cultured， and injected subcutaneously into syngeneic recipients to produce additional teratomas or injected intraperitoneally to form aggregates that were organized in a manner that resembled embryos， and were therefore called embryoid bodies （EBs）（Kleinsmith & Pierce， 1964）. Teratoma formation has become a standard by which to assess stem cell plasticity and is still used as a measure of pluripotency（Okita， Ichisaka， & Yamanaka， 2007）. Teratomas are frequently assessed via histological staining and by immunocytochemical detection of markers for tissues derived from the primary embryonic germ layers （Li & Tanaka， 2011； Shao et al.， 2007； Takahashi et al.， 2007； Takahashi & Yamanaka， 2006； Martin， 1981）.

An area of special interest for ESC-based cell therapies is treatment of neurodegenerative disorders. Diseases of the central nervous system （CNS）， such as amyotrophic lateral sclerosis and Parkinsons disease， are debilitating and currently have no cure； however， stem cell therapies could replace nerve cells lost due to the pathology and restore neurological function （Srivastava， Malhotra， Sharp， & Berggren， 2008）. Many induction protocols for ESCs have been developed that produce a variety of neural progeny including cholinergic， dopaminergic， serotonergic， GABAergic， and glutamatergic neurons， as well as glial cells， including oligodendrocytes and astrocytes （Cai & Grabel， 2007； Bibel et al.， 2004； Wang et al.， 2009； Bjorklund et al.， 2002； Lee， Lumelsky， Studer， Auerbach， & McKay， 2000； Meyer， Katz， Maruniak， & Kirk， 2004； Okabe， Forsberg-Nilsson， Spiro， Segal， & McKay， 1996； Bain， Kitchens， Yao， Huettner， & Gottlieb， 1995）.

Genetic， morphological， behavioral， and neuroanatomical variability exists between members of different mouse strains and even among substrains of specific murine lineages （Bryant et al.， 2008； Chen et al.， 2006； Crusio， Schwegler， & Van Abeelen， 1991； Auerbach et al.， 2000； Kawase et al.， 1994； Tabibnia， Cooke， & Breedlove， 1999； Wahlsten， Metten， & Crabbe， 2003； Ledermann， 2000； Cook， Bolivar， McFadyen， & Flaherty， 2002）， and based on these differences， one would predict differences in ESC lines derived from different mouse strains. Many transgenic mouse models are derived from the C57BL/6 mouse strain， and it would be advantageous to use C57BL/6-derived ESC cell lines， such as the B6 ESC cell line， for transplantation experiments （Kawase et al.， 1994； Ledermann & Burki， 1991； Gertsenstein et al.， 2010； Hughes et al.， 2007）. However， previous studies show that C57BL/6-derived ESCs can exhibit slower growth rates when compared to 129/SvJ-derived ESCs， thus leading to lower efficiency rates in establishing chimeras and knockout mice （Auerbach et al.， 2000； Kawase et al.， 1994； Ledermann & Burki， 1991； Gertsenstein et al.， 2010； Cheng， Dutra， Takesono， Garrett-Beal， & Schwartzberg， 2004； Brook & Gardner， 1997； Seong， Saunders， Stewart， & Burmeister， 2004； Collins， Rossant， & Wurst， 2007； Ware， Siverts， Nelson， Morton，& Ladiges， 2003； Keskintepe， Norris， Pacholczyk， Dederscheck， & Eroglu， 2007）.

To assess the potentials of different ESC lines for brain transplantation， in the current study we performed a comparative analysis of the ESC lines B5 （129/SvJderived） and B6 （C57BL/6-derived）. The current analysis included tests of their growth responses in culture， their in vitro response to the 4–/4+ retinoic acid neural induction protocol （Bain et al.， 1995）， and their ability to form teratomas in vivo after implantation into the brains of mature mice.

1. MATERIALS AND METHODS

1.1 Cell Culture

Mouse ESCs were cultured according to protocols described previously （Bain et al.， 1995； Meyer， Katz， Maruniak， & Kirk， 2006）. Briefly， enhanced green fluorescent protein （EGFP） expressing mouse B6 ESCs[C57BL/6J-Tg （ActinB：EGFP） OsbY01]， obtained from the Murine Embryonic Stem Cell Core （Washington University， St. Louis） and B5 mouse ESCs （S129） （Meyer et al.， 2004； Meyer et al.， 2006； Pierret et al.， 2007； Pierret et al.， 2010）， obtained from Dr. Andras Nagy， Samuel Lunenfeld Research Institute （Mt. Sinai Hospital， Toronto， Ontario， Canada） were cultured on feeder-free， gelatin-coated T25 flasks （Midwest Scientific， Catalog number： TP90025） containing 5 mL ESC growth medium（ESGM） and 1000 U/mL leukemia inhibitory factor （LIF； Millipore， Catalog number： ESG1106）. ESGM contained Dulbeccos modified Eagle medium （DMEM； Gibco， Catalog number： 11965-092）， 10% newborn calf serum， 10% fetal bovine serum， 5% nucleoside stock （consisting of 0.80 mg/mL adenosine， 0.85 mg/mL guanosine， 0.73 mg/mL cytidine， 0.73 mg/mL uridine， and 0.24 mg/mL thymidine）， and 1 mM β-mercaptoethanol （βME）. After growth to approximately 70% confluence， cells were exposed to 0.25% trypsin， mechanically dissociated， and passaged to new T25 flasks at a ratio of 1：2 or 1：3. The newly plated cells were checked using phase contrast microscopy to ensure that a majority existed as single cells in suspension.

The 4–/4+ protocol of Bain et al. （1995） was used to induce ESCs to a neural fate. To achieve this neural in duction， colonies were grown to 70% confluence， dissociated as described above， passed through a 40 microns cell strainer and transferred to uncoated petri dishes containing ESC induction media （ESIM） for 4 days. ESIM consisted of the same components as ESGM， but lacked LIF and βME. The cells established spherical aggregates （i.e.， EBs）. After four days， the EBs were then cultured for an additional four days in ESIM containing 500 nM all-trans retinoic acid （RA； Sigma-Aldrich， Catalog number： R2625）， a potent neuralizing agent（Cheng et al.， 2004； Faherty， Kane， & Quinlan， 2005； Yamanaka et al.， 2008）. Prior to transplantation on Day 8 of induction， the neuralized EBs were treated with 0.25% trypsin for 8 minutes and dissociated mechanically using repeated pipetting. Cell counts were performed using a hemocytometer， and trypan blue exclusion was used as a measure of cell viability.

1.2 Transplantations

Because both the B5 and B6 ESC lines were derived from male mouse strains， male mice were used for all transplant experiments to enhance immune compatibility. Mature C57BL/6 mice were anesthetized using mouse anesthetic cocktail [25 mg/mL Ketamine， 25 mg/mL LA Xylazine， and 0.5 mg/mL Acepromazine in 0.1M phosphate-buffered saline （PBS）； 0.1 cc per 25 gm body weight， given IM]. Anesthetized mice were placed in a stereotactic device， and an incision was made anteroposteriorly across the midline of the head. An electric drill and bit were used to make a burr hole in the cranium large enough in diameter to accommodate a 36 gauge needle while targeting the left striatum using the following coordinates： anteroposterior-0.72 mm from bregma； lateromedial -1.73 mm from midline； dorsoventral -2.00 mm from skull. A 10 μL Hamilton syringe with a 36 gauge needle was inserted 3 mm below the surface of the cerebrum and then retracted 1 mm. This method targeted the left striatum and created a space for the implanted cells during injection. Undifferentiated or induced ESCs were dissociated as described above and re-suspended at 500，000 cells per microliter in DMEM. Two microliters of cell suspension was injected for a total of 1 million implanted cells. Vetbond？ tissue adhesive was used to close the skin incision.

Mice were euthanized 3 or 6 weeks after transplantation， depending on the experimental design. In all cases， anesthetized mice were perfused transcardially with 0.1M PBS followed by 4% phosphate-buffered paraformaldehyde （PFA）. A volume of 30 mL of icecold PBS followed by 30 mL of ice-cold PFA were administered transcardially to each animal using a polystatic pump （Buchler Instruments） over a period of 12 minutes. Brains were removed and postfixed whole with 4% PFA for 36 hours and sucrose-protected by serial transfer into 10%， 20% and 30% sucrose in 0.17M sodium cacodylate buffer. Subsequently， brains were embedded in OCT and frozen at -20 °C. All animal care and treatment protocols followed MU Office of Animal Research guidelines and were approved by the MU Animal Care and Use Committee.

1.3 Immunohistochemistry

Embryoid bodies on Day 4 or Day 8 of neural induction（see above）， were pelleted gently， fixed overnight at 4 °C in 4% PFA in 0.1 M PBS， equilibrated in 20% sucrose in 0.17 M sodium cacodylate， embedded in OCT， and frozen at -20 °C. Serial frozen sections of EBs and of host brains were cut with a Leica Cryostat （Molecular Cytology Core， University of Missouri-Columbia） at a thickness of 10 μm， mounted on SuperFrost？ Plus slides（Fisher， Catalog number： 12-550-15）， immersed in 0.1 M PBS， and placed on a rocker at low speed for 15 minutes. Sections were then permeabilized and blocked using 0.3% Triton X-100 in 0.1 M PBS and 10% normal goat serum （NGS） for one hour at room temperature. Primary antibodies selective for pluripotent stem cells （mouse monoclonal for OCT3/4， also designated POU5F1， 1：250， Santa Cruz， Catalog number： sc-5279）， neural precursors [mouse monoclonal for nestin （NES）， 1：200， Millipore， Catalog number： MAB353]， immature neurons[mouse monoclonal for β-III Tubulin （TUBB3）， 1：100， Promega， Catalog number： G7121]， astrocytes [rabbit polyclonal for glial fibrillary acidic protein （GFAP）， 1：1000， Dako， Catalog number： z0334]， mature neurons[rabbit polyclonal for Neurofilament-M （NF-M）， 1：200， Millipore， Catalog number： AB1987]， mesoderm [rabbit polyclonal for Brachyury （BRY）， 1：100， Abcam， Catalog number： ab2068]， and endoderm [mouse monoclonal for α-fetoprotein （AFP）， 1：100， Cell Signaling， Catalog number： 3903S] were diluted in 0.1 M PBS containing 2% NGS and 0.5% Triton X-100 and applied to sections overnight at 4 °C. Samples were then washed 5 times with 0.1 M PBS at 5 minute intervals. Appropriate fluorescenttagged goat anti-mouse （Alexa Fluor？ 546， Molecular Probes， Catalog number： A11003） or goat anti-rabbit（Alexa Fluor？ 546， Molecular Probes， Catalog number： A11010） antibodies were diluted 1：200 in 0.1 M PBS containing 2% NGS， 0.5% Triton X-100， and the nuclear stain， DAPI （1：300， Molecular Probes， Catalog number： D3571） and applied for 1-4 hours at room temperature. Controls were performed in parallel with each treatment group and were processed using the above protocol， without the primary antibody. After application of the secondary antibody， samples were washed 8 times at 5 minute intervals with 0.1 M PBS. Coverslips were applied using a Mowiol solution [0.1 M TRIS at pH 8.5， 10% MOWIOL？ Reagent 4-88 （Calbiochem， Catalog number： 47904）， 25% glycerol]. Brain sections were imaged using a Leica M205 FA stereomicroscope with a Leica DFC 345 FX camera and processed using LAS AF Lite version 2.4.1， build 6384 and ImageJ version 1.43. Mounted sections of EBs were imaged using an Olympus IX70 inverted microscope equipped with a Hamamatsu ORCAAG deep-cooled CCD camera， and 8-bit TIF images were processed using MetaMorph version 6.3r6 and ImageJ version 1.43. Images of immunolabeled samples were captured and processed at the Molecular Cytology Core， University of Missouri-Columbia.