The cellular form of the prion protein guides the differentiation of human embryonic stem cells into neuron-, oligodendrocyte-, and astrocyte-committed lineages

Abstract

Prion protein, PrP^C, is a glycoprotein that is expressed on the cell surface beginning with the early stages of embryonic stem cell differentiation. Previously, we showed that ectopic expression of PrP^C in human embryonic stem cells (hESCs) triggered differentiation toward endodermal, mesodermal, and ectodermal lineages, whereas silencing of PrP^C suppressed differentiation toward ectodermal but not endodermal or mesodermal lineages. Considering that PrP^C might be involved in controlling the balance between cells of different lineages, the current study was designed to test whether PrP^C controls differentiation of hESCs into cells of neuron-, oligodendrocyte-, and astrocyte-committed lineages. PrP^C was silenced in hESCs cultured under three sets of conditions that were previously shown to induce hESCs differentiation into predominantly neuron-, oligodendrocyte-, and astrocyte-committed lineages. We found that silencing of PrP^C suppressed differentiation toward all three lineages. Similar results were observed in all three protocols, arguing that the effect of PrP^C was independent of differentiation conditions employed. Moreover, switching PrP^C expression during a differentiation time course revealed that silencing PrP^C expression during the very initial stage that corresponds to embryonic bodies has a more significant impact than silencing at later stages of differentiation. The current work illustrates that PrP^C controls differentiation of hESCs toward neuron-, oligodendrocyte-, and astrocyte-committed lineages and is likely involved at the stage of uncommitted neural progenitor cells rather than lineage-committed neural progenitors.

Keywords:

• prion protein
• human embryonic stem cells
• stem cell differentiation
• neuron-committed lineages
• neural progenitor cells

Abbreviations

PrPC	=	normal, cellular isoform of the prion protein
ESCs	=	embryonic stem cells
hESCs	=	human ESCs
Syn	=	synapsin I
TH	=	tyrosine hydroxylase
Olig1	=	a marker of oligodendrocyte-committed lineages
GFAP	=	glial fibrillary acidic protein
Tet	=	tetracycline
TetR	=	tetracycline repressor
Lenti-TetR	=	lentiviral vector expressing tetracycline repressor
Lenti-ShPrPC	=	lentiviral vector expressing short hairpin RNA against PrPC
Lenti-ShScram	=	lentiviral vector expressing scrambled shRNA
hES+TetR+ShPrPC	=	hESCs transfected with Lenti-TetR and Lenti-ShPrPC
hES+TetR+ShScram	=	hESCs transfected with Lenti-TetR and Lenti-ShScram
bFGF	=	basic fibroblast growth factor
MEFs	=	mouse embryonic fibroblasts
NIM	=	neural induction medium
NPM	=	neural proliferation medium
NDM	=	neuronal differentiation medium
MEF-CM	=	mouse embryonic feeder-conditioned medium
GRM	=	glial restrictive medium
EFG	=	epidermal growth factor
RA	=	retinoic acid
EBs	=	embryoid bodies
CNTF	=	ciliary neurotrophic factor

Introduction

Conversion of the cellular prion protein (PrP) into self-propagating, β-sheet rich, aggregated state, PrP^Sc, underlies several lethal transmissible neurodegenerative diseases in mammals, including Creutzfeldt-Jakob disease and bovine spongiform encephalopathy.¹ The normal, cellular isoform of the prion protein, PrP^C, is a sialoglycoprotein that is attached to the cell membrane via a C-terminal glycosylphosphatidyl-inositol anchor and has two N-linked glycans.^2,3 While PrP^C is expressed in various tissues and organs, the highest level of expression has been found in cells of the central nervous system and in spermatogenetic cells.^4-6

The spectrum of proposed biological functions of PrP^C has been expanding rapidly over the last decade. While several independent studies did not observed any deleterious effects in PrP^C knock out mice,^7,8 deficiency in one of the two genes encoding PrP^C in zebrafish led to an early developmental arrests.⁹ To date, the list of possible PrP^C functions includes those related to its Cu²⁺-binding properties such as Cu²⁺ sensing, Cu²⁺ transporting or acting as superoxide dismutase,^10–13 or activities not directly linked to Cu²⁺-binding such as signal transduction,¹⁴ neuroprotection^15–19 or regulation of the cell cycle.^20,21 Yet another set of PrP^C activities that has been under rapid expansion in recent years are likely related to PrP^C N-linked glycosylation and sialylation. They include neurotrophic activities,^19,22,23 cell adhesion,^9,23-25 and cell proliferation and differentiation.^19,21,26-31 The proportion of di- vs. mono-, and unglycosylated PrP^C glycoforms was found to increase in the course of neuronal differentiation, as well as upon an increase in the density of cells cultured in vitro.^32,33 The studies that used PrP^C knock out and overexpresser mice revealed that PrP^C expression in multipotent neural precursor cells correlates with differentiation of precursors into mature neurons during neural development and adult neurogenesis.²⁶ Cross-talk between PrP^C and the neural cell adhesion molecule NCAM was pointed out as critical for differentiation of neural stem cells into neurons.³⁴ In support of the hypothesis that PrP^C is involved in differentiation of neural precursor cells,²⁶ PrP^C was found on growing axons during development and along fiber bundles that contain elongating axons in the adult brain.^22,35 Moreover, PrP^C was also found to induce polarization, synapse development and neuritogenesis in embryonic neuron cultures.^18,36 Recently, ablation of PrP^C expression was found to cause chronic demyelinating polyneuropathy suggesting that PrP^C is required for myelin maintenance.³⁷

Several studies illustrated that the involvement of PrP^C in stem cell proliferation and differentiation is broader then just guiding differentiation of neural stem cells. PrP^C was found on the surface of hematopoietic stem cells and proposed to be important for their self-renewal.²⁷ Moreover, studies conducted using PrP^C knock out mice revealed that PrP^C is important for regeneration of adult muscle tissue and guides the proliferation and differentiation of myogenic precursor cells.³⁸ Extremely high levels of Prnp (prion protein gene) mRNA were found in Sertoli cells that support development and high proliferating activity of spermatogonial stem cells.⁶

In part due to limitations in usage of human embryonic stem cells, the vast majority of previous work on elucidating the role of PrP^C in stem cell differentiation was limited to animal models or animal-derived cell lines. To test whether PrP^C plays a role in early human embryogenesis, our previous studies examined the effects of recombinant PrP or the impact of PrP^C suppression or overexpression on human embryonic stem cells (hESCs) during spontaneous differentiation.^21,28 hESCs are pluripotent cells with high self-renewal and proliferation activities that can be differentiated into any cell type of the three germ layers and subsequently any tissue.³⁹ The developmental sequence of hESCs embryonic bodies (EB) resembles the process of human embryogenesis.⁴⁰ Using a panel of lentiviral vectors we previously found that PrP^C is involved in key cellular activities that determine the status of hESCs: (1) it regulates the cell cycle, (2) it controls the cellular switch between self-renewal and differentiation, and (3) contributes to regulating a differentiation balance between the lineages of the three germ layers, endoderm, mesoderm and ectoderm.²¹ In particular, we found that silencing of PrP^C under spontaneous differentiation conditions selectively suppressed differentiation toward ectodermal progenitors without affecting differentiation toward endodermal and mesodermal progenitors.

The current study asks the question whether PrP^C selectively controls differentiation of hESCs into cells of three neural lineages: neuronal cells, astrocytes and oligodendrocytes. Using three protocols for differentiating hESCs into predominantly neuronal, oligodendrocyte and astrocyte lineages, we found that suppression of PrP^C expression delays differentiation of all three lineages. Switching PrP^C expression on and off during early differentiation revealed that PrP^C expression during the very early stages that correspond to embryonic bodies has a more significant impact than silencing PrP^C at the later stages. The extent of differentiation delay was found to be roughly proportional to the total length during which the expression of PrP^C was suppressed. Together with previous studies, this work illustrates that PrP^C is involved in controlling a balance between differentiation toward three germ layers, but not within neural lineages. The current work also suggests that PrP^C is involved at the early stages of neural differentiation, perhaps, at the stage of uncommitted neural progenitor cells rather than lineage-committed progenitors.

Results

Silencing of PrP^C delays differentiation of neuronal, oligodendrocytic, and astrocytic lineages

To examine the effect of PrP^C on differentiation of hESCs toward three neural lineages, we employed an hESC line expressing inducible short hairpin RNA that suppresses PrP^C expression under control of a tetracycline repressor (designated as Sup). As controls, an hESC line that expresses only the tetracycline repressor (designated as C1) or an hESC line that expresses inducible scrambled short hairpin RNA under control of the tetracycline repressor (designated as C2) were used. Three experimental protocols that were previously shown to induce differentiation of hESCs into predominantly neuronal cells, oligodendrocytes and astrocytes were employed.^41–43 In all three protocols, PrP^C expression was suppressed starting from day zero and the suppression was continued until the experimental end-point, day 40.

As judged from cell shape and growing patterns, PrP^C silencing delayed differentiation of hESCs conducted according to all three protocols. Notable differences in cell morphology between control lines and hESCs with suppressed PrP^C expression were observed by the day 30 regardless of the differentiation protocol employed (Figs. 1A, 2A, and 3A). By the day 40, most of hESCs with silenced PrP^C remained aggregated and lacked well-defined cell shapes, whereas cells in control groups acquired well-defined shape, a sign of differentiation (Figs. 1A and B, 2A and B, and 3A and B). By the day 40 control lines subjected to neuro-, oligodendrocyte- or astrocyte-specific differentiation protocols expressed tyrosine hydroxylase (TH, a marker of neuronal cells), Olig1, (a marker of oligodendrocyte-committed lineages) or glial fibrillary acidic protein (GFAP, a marker of astrocyte-commited lineages) across cell populations, respectively (Figs. 1B and C, 2B and C, and 3B and C). hESC lines with silenced PrP^C expressed substantially lower amounts of protein markers specific for neuronal, oligodendrocytic or astrocytic lineages than in corresponding control cell groups (Figs. 1B and C, 2B and C, and 3B and C).

Figure 1. Suppression of PrP^C delays differentiation into neuronal cells. (A) Phase-contrast images of hESCs with suppressed PrP^C-expression (Sup) and in two control lines (C1 and C2) taken at 10th, 20th, 30th, and 40th day of differentiation. (B) Immunostaining for PrP^C (red) and TH (green) on 40th day of differentiation. Hoechst 33342 was used for staining of nuclei (blue). Scale bar = 50 μm. (C) Quantification and statistical analyses of PrP^C- and TH-positive cells on 40th day of differentiation. The data represent a mean ± SD from three independent experiments for each cell line. Statistical significance was determined by Student's t test: *, P < 0.05.

Figure 2. Suppression of PrP^C delays differentiation into oligodendrocytes. (A) Phase-contrast images of hESCs with suppressed PrP^C-expression (Sup) and in two control lines (C1 and C2) taken at 10th, 20th, 30th, and 40th day of differentiation. (B) Immunostaining for PrP^C (red) and Olig1 (green) on 40th day of differentiation. Hoechst 33342 was used for staining of nuclei (blue). Scale bar = 50 μm. (C) Quantification and statistical analyses of PrP^C- and Olig 1-positive cells on 40th day of differentiation. The data represent a mean ± SD from three independent experiments for each cell line. Statistical significance was determined by Student's t test: *, P < 0.05.

Figure 3. Suppression of PrP^C delays differentiation into astrocytes. (A) Phase-contrast images of hESCs with suppressed PrP^C-expression (Sup) and in two control lines (C1 and C2) taken at 10th, 20th, 30th, and 40th day of differentiation. (B) Immunostaining for PrP^C (red) and GFAP (green) on 40th day of differentiation. Hoechst 33342 was used for staining of nuclei (blue). Scale bar = 50 μm. (C) Quantification and statistical analyses of PrP^C- and GFAP-positive cells on 40th day of differentiation. The data represent a mean ± SD from three independent experiments for each cell line. Statistical significance was determined by Student's t test: *, P < 0.05.

Western blotting for synapsin I (Syn, a neuronal marker), Olig1 or GFAP revealed that neuron-, oligodendrocyte- or astrocyte-specific differentiation protocols had a noticeable impact on the outcomes of differentiation, although the resulting cell populations were found to be heterogeneous in cultures produced according to three protocols (Fig. 4A). For instance, cell differentiated according to the neuronal protocol expressed Syn, but also oligodendrocyte- and barely detectible astrocyte-specific markers (Fig. 4A). Cells differentiated according to the oligodendrocytic protocol expressed Olig1 and to a lesser extent Syn and GFAP. Cell treated according to astrocyte-specific protocols expressed Olig1 and GFAP, but barely detectible amounts of Syn (Fig. 4A). Nevertheless, regardless of the differentiation protocol, hESC lines with silenced PrP^C displayed substantially lower levels of neuron-, oligodendrocyte- or astrocyte-specific markers in comparison to the corresponding control lines (Fig. 4A and B). hESC with silenced PrP^C as expected also showed the lowest level of PrP^C expression relative to the control cell lines (Fig. 4A and B). In summary, the current experiments revealed that suppression of PrP^C expression blocked or substantially delayed differentiation of hESCc cells into three neuronal lineages (neuronal cells, oligodendrocytes and astrocytes); the effect of PrP^C was observed regardless of the differentiation protocol employed.

Figure 4. western blotting analysis of expression of PrP^C and three marker proteins. Western blotting (A) and quantification with statistical analyses (B) of the expression level of PrP^C, Syn, Olig 1, and GFAP in hESCs with suppressed PrP^C expression (Sup) and in two control lines (C1 and C2) cultured for 40 d under lineage-preferred differentiation conditions in the presence of tetracycline. β-actin was used as a loading control in (A). In (B) the expression level of each protein was normalized relative to that of β-actin in each differentiation protocol. The data represent a mean ± SD from three independent experiments. Statistical significance was determined by Student's t test: *, P < 0.05.

The effect of PrP^C silencing at different time points on differentiation of hESCs

To test whether it is a specific time point of turning PrP^C expression on and off or the total duration of PrP^C silencing that has the most profound impact on delaying hESC differentiation, PrP^C was silenced at different time points during spontaneous differentiation as shown in Figure 5A. The differentiation status was determined at the day 15 by analyzing the expression of protein markers Syn, Olig1 and GFAP, specific for neuronal, oligodendrocytic and astrocytic lineages, respectively (Fig. 5B and C). Previously we showed that under spontaneous differentiation conditions hESCs cells gave rise to neuronal cells, oligodendrocytes and astrocytes.²¹

Figure 5. The effect of PrP^C silencing at different time points on differentiation of hESCs into neuron-, oligodendrocyte-, and astrocyte-specific lineages. (A) Schematic representation of eight experimental formats, where PrP^C was temporarily silenced for 5, 10, or 15 d. hES+TetR+ShPrP^C (Sup) cells were cultured under spontaneous differentiating conditions either in the absence (gray arrows, PrP^C expression is on) or presence of 1 μg/ml tetracycline (white arrows, PrP^C expression is silenced). Western blotting (B) and normalized expression level (C) of PrP^C, Syn (synapsin I), Olig 1 and GFAP in hESCs after 15 d of culturing according to eight experimental formats described in panel A. β-actin was used as a loading control. In (C), the expression level of each protein was normalized relative to that of β-actin in each differentiation format. The data represent a mean ± SD from three independent experiments.

In the absence of PrP^C silencing, all three markers and PrP^C were detected by day 15, confirming on-going differentiation (format 6, Fig. 5). Expression of Syn was relatively weak, as its expression typically increases at the later stages of neuronal differentiation. Comparison of the eight formats revealed that regardless of the specific time point at which PrP^C was turned on or off, the extent to which differentiation was suppressed was roughly proportional to the length of time during which PrP^C was silenced. Indeed, PrP^C silencing for 5 d (formats 1, 5 and 7) showed relatively minor effects on expression of Olig1 and GFAP regardless of whether PrP^C was silenced at the beginning, in the middle or at the end of the 15-d experiment (Fig. 5B, C). PrP^C silencing for 10 d (formats 2, 4, and 8) showed more substantial effects than silencing for 5 d. Notably, in the experiment where PrP^C was continuously silenced during the last 10 d of experiment (format 2), the negative effect on differentiation was minimal in comparison to the formats 4 and 8 (Fig. 5B, C). In fact, the negative effect of PrP^C silencing for 10 d in format 2 was similar to those observed in formats # 1, 5 and 7, where PrP^C was silenced for 5 d. This result illustrates that PrP^C expression during the very initial stages might be more important than at the later stages. None of the three markers could be detected in the format where PrP^C was silenced for the whole duration of the experiment (format 3, Fig. 5B,C), confirming again that PrP^C is important for differentiation of cells in all three lineages.

Careful comparison within each group where PrP^C was silencing for 5 or 10 d revealed relatively minor variations between the ratios of neuron-, oligodendrocyte- or astrocyte-specific markers (Fig. 5B,C). This suggests that PrP^C expression might have a somewhat limited impact on tuning the ratio between neuronal cells, oligodendrocytes and astrocytes at different time-points of spontaneous differentiation. However, temporal silencing of PrP^C during different stages of differentiation was not able to alter drastically the cell fate of hESCs.

Discussion

In previous studies, we showed that ectopic expression of PrP^C in hESCs under self-renewal conditions triggered differentiation toward lineages of three germ layers.²¹ Silencing of PrP^C in hESCs cultured under spontaneous differentiation conditions suppressed differentiation of cells toward ectodermal lineages, but did not affect differentiation of endodermal and mesodermal lineages, thereby changing the balance between the lineages of three germ layers.²¹ Because all three protein markers used to examine ectodermal differentiation in previous studies were neuron-specific (growth-associated protein 43, synaptophysin and TH), it remained unclear whether PrP^C was involved only in differentiation of neuronal cells or also in cells of other ectodermal lineages, such as oligodendrocyte- or astrocyte-committed lineages. The question of whether PrP^C is expressed in oligodendrocytes, astrocytes or their immediate precursors remains controversial. The studies that examined different mouse brains areas including neurogenetic regions did not find PrP^C expression in astrocytes or oligodendrocytes,^26,44 whereas several other studies that employed different animal models demonstrated expression of Prnp mRNA or PrP^C in astrocytes or oligodendrocytes.^45-48

In the current study, we showed that silencing of PrP^C in hESCs suppressed differentiation toward neuron-, oligodendrocyte-, and astrocyte-committed lineages (Fig. 6). Similar results were observed in three different experimental protocols, arguing that the effect of PrP^C was independent of differentiation conditions employed. Notably, while each protocol was developed for differentiating hESCs into three different cell types, the resulting cell populations were found to be heterogeneous in each condition and consisted of oligodendrocytes, astrocytes and neuronal cells. Notably, PrP^C silencing suppressed differentiation toward all three lineages under each differentiation condition.

Figure 6. A schematic model illustrating PrP^C effects on self-renewal and differentiation of hESCs into neural lineages. Expression of PrP^C in pluripotent hESCs triggers their differentiation into neuron-, oligodendrocyte- and astrocyte-committed lineages. PrP^C guides differentiation at the very early stage, presumably the stage of uncommitted progenitor, via a yet unknown signaling pathway or through inhibiting cell-renewal.

While the current work suggests that PrP^C deficiency might lead to significant delay or even arrested of neural development in human, it is currently difficult to project these results on development of human embryo in part because such conclusions are not well supported by previous studies that employed animal models. Previous studies using PrP^C knock out mice revealed only subtle phenotypes such as minor neurophysiological or behavioral abnormalities,^49-51 but no obvious problems in development.^7,8 Several studies pointed out that PrP^C might be essential only under circumstances when animals are subjected to stress conditions, such as regeneration of adult muscle tissues after damage³⁸ or self-renewal of hematopoetic cells after irradiation.²⁷ It is difficult to comprehend the extent to which PrP^C is essential for neurodevelopment in part because of some differences in biochemistry and physiology between humans and other mammals. A number of functional activities of PrP^C and in particular those involved in neurotrophic activities,^19,22,23 cell adhesion,^9,23–25 cell proliferation and differentiation^19,21,26-31 are likely to depend on sialylation of PrP^C N-linked glycans. Due to an irreversible mutation in the gene encoding human N-acetylneuraminic acid hydroxylase, different kinds of sialyc acid residues are synthesized in humans compared with the rest of mammalian species: N-acetyl neuraminic acid (Neu5Ac) is produced in human, while N-glycolylneuraminic acid (Neu5Gc) is in all other mammals.⁵² Among other factors, the difference in structures of sialic acid residues might account for the differences in the spectrum of PrP^C biological activities in humans and other mammals. Notably, early developmental arrests were described in PrP knock out models of zebrafish⁹ supporting the role of PrP^C in early development in at least one taxonomic class.

The recent article by Miranda et al. provides comprehensive review of the studies on the role of PrP^C in stem cell pluripotency, self-renewal and differentiation,⁵³ therefore we will only discuss the results that are the most relevant to the current finding. During early embryogenesis, PrP^C expression was found to be largely limited to cells that undergo neural differentiation.^54,55 The fact that PrP^C expression accompanies neuronal differentiation during embryo development was supported by work on in vitro models that employed neural precursors and ESCs.^26,28,33 Treatment of primary neuronal culture or embryonic neurons with recombinant PrP in α-helical conformation helped neuronal survival and stimulated neurite outgrowth.^36,56 The current finding that differentiation toward all three neural lineages was delayed, suggests that PrP^C is involved at the early stages of differentiation, perhaps, at the stage of uncommitted neural progenitor cells rather than lineage-committed neural progenitors (Fig. 6). In support of this hypothesis, the experiment on silencing PrP^C expression at different stages of early differentiation revealed that PrP^C expression during the very early stage of embryonic bodies had a more significant impact than silencing at the later stages (Fig. 5). These results expands the previous view and suggests that not only does PrP^C expression accompany cell fate restriction of multipotent stem cells toward neural lineages,³¹ but is actively involved at the early stages of differentiation toward neural-committed lineages. While no deleterious effects were reported in PrP^C knock out mice, a double knock out of PrP^C and its paralog Shadoo was found to have lethal effect on mouse development at the E10.5 stage.⁵⁷ The current work is consistent with the results of the recent studies that illustrate a delay in differentiation toward oligodendrocytes in embryonic primary cell cultures isolated at P0-P2 stage from Prnp^0/0 mouse pups relative to cultures from Prnp^+/+ mouse pups.⁵⁸ Other studies reported faster astrocyte maturation in cultured embryonic astrocytes from PrP^C -overexpressing mice compared with wild-type mice,⁵⁹ and a slower rate of astrocyte development in cultures from PrP^C-null animals then wild-type mice.⁶⁰ These results are in accord with the current work that PrP^C silencing delays differentiation of the astrocyte-committed lineage. We did not observe significant effect of PrP^C silencing in changing the balance between neuron-, oligodendrocyte- or astrocyte-specific differentiation pathways. This is consistent with the idea that PrP^C is involved in differentiation of cell of all three neural lineages at very early stages (Fig. 6).

How does PrP^C regulate ESCs self-renewal and differentiation? Several plausible PrP^C interacting partners and pathways have recently been described. Knocking out the Prnp gene or blocking PrP^C with PrP^C-specific antibodies was shown to downregulate the mRNA expression level of Nanog, a transcription factor which is involved in ESCs self-renewal and controlling pluripotency.⁶¹ Knocking down PrP^C expression by siRNA in mouse ESCs had a negative effect on transcription of the ectodermal marker Nestin.⁶² This result is consistent with the current study that PrP^C is involved at the stage of stem cell differentiation into neural progenitor cells. Other studies reported that the positive effect of PrP^C on neuronal differentiation and neurite outgrowth could be mediated via interaction of PrP^C with the stress-inducible protein STI1,^18,30 laminin,⁶³ or the neural cell adhesion molecule NCAM.³⁴ In addition, PrP^C could be involved in changing the balance between self-renewal and differentiation via interaction with cdk2, which is known to participate in the cell cycle G1 to S phase transition.⁶⁴ Our previous studies showed that at the early stages of hESCs differentiation, PrP^C silencing changed the structure of cell cycle by inhibiting the cell cycle G1 to S phase transition²¹ (Fig. 6).

Importance of PrP^C for early neural differentiation might also play a critical role in prion pathogenesis. As supported by recent study, PrP^C depletion due to PrP^Sc spread and replication in CNS is likely to have a negative impact on the ability of CNS to replace lost neurons via stimulating differentiation of adult stem cells during prion disease.⁶⁵ Consistent with this hypothesis, PrP^C expression level in the mouse subventricular zone was found to correlate with differentiation of multipotent neural precursors during adult neurogenesis.²⁶

Materials and Methods

hESCs culturing

The protocol for using hESCs (H9, National Stem Cell Bank, Madison, WI, USA) was reviewed and approved by the University of Maryland, Baltimore Embryonic Stem Cell Research Oversight Committee and Institutional Review. The following H9 hESCs were employed in this study: i) hES+TetR is a hESC line transfected with a lentiviral vector expressing tetracycline repressor (referred to as C1); ii) hES+TetR+ShPrP^C is a hESC line transfected with lentiviral vectors expressing tetracycline repressor and inducible short hairpin RNA for suppression of PrP^C expression (referred to as Sup); iii) hES+TetR+ShScram is a hESC line transfected with lentiviral vectors expressing tetracycline repressor and inducible scrambled short hairpin RNA as a control for non-specific effects (referred to as C2). These hESCs lines were described in detail in a previous study.²¹ H9 hESCs were maintained on mitomycin C (Sigma, St. Louis, MO, USA) -treated mouse embryonic fibroblasts (MEFs) (American Type Culture Collection, Manassas, VA, USA) feeder layers in DMEM/F12 (Invitrogen, Carlsbad, CA, USA) supplemented with 20% knockout serum replacement (Invitrogen), 0.1 mM β-mercaptoethanol (Sigma), 0.1 mM non-essential amino acids (Invitrogen), 50 U/ml penicillin G (Invitrogen), 50 μg/ml streptomycin (Invitrogen) and 4 ng/ml human recombinant basic fibroblast growth factor (bFGF, Invitrogen) at 37°C in 5% CO₂. hESC colonies were subcultured on new feeder cells every 5–7 d.

Induced differentiation of hESCs

For differentiation into neuronal cells, the protocol described by Nat and coauthors was used as previously described.⁴³ Briefly, after culturing in feeder free conditions, hESC colonies were incubated in neural induction medium (NIM, DMEM:Neurobasal (1:1) supplemented with 1x N2 and 1x B27 supplements and 2mM Glutamax) for one day. Then, hESCs were cultured on non-adherent culture dishes in NIM until they form uniformed aggregates for 2–4 d. Floating aggregates were transferred to neural proliferation medium (NPM, DMEM:Neurobasal (1:1) supplemented with 0.5x N2 and 0.5x B27 supplements, 2mM Glutamax and 20 ng/ml FGF2) on day 5. At day 7, aggregates were plated on laminin-coated culture dishes in NPM and cultured for an additional 10 d. The medium was changed with neural differentiation medium (NDM, Neurobasal supplemented 1x B27 supplement, 2mM Glutamax, 10 ng/ml brain derived neurotrophic factor, BDNF) for additional 20 d.

For differentiation into oligodendrocytes, the protocol described by Sharp and coauthors was used as previously described.⁴² Briefly, mouse embryonic feeder-conditioned medium (MEF-CM), glial restrictive medium (GRM, DMEM with 1x B27 supplement, 1x Insulin/Transferrin/Selenium Solution (Gibco) and Transition medium (MEF-CM:GRM = 1:1) were used. To induce differentiation into oligodendrocyte progenitor cells, hESCs were treated with collagenase IV (Sigma), and cell aggregates were distributed to low-attachment cell culture plates. After cultivation for two days with transition medium, cell aggregates were cultured with GRM+20 ng/ml epidermal growth factor (EFG, Sigma) supplemented with 10 μg/ml retinoic acid (RA, Sigma) for 10 d, then cultured with GRM supplemented with 20 ng/ml EGF medium for additional 15 d. Cell aggregates were transferred to Matrigel (BD Biosciences)-coated plates and cultured with GRM medium supplemented with 20 ng/ml EGF for additional 20 d while changing medium every other day.

For differentiation into astrocytes, the protocol described by Emdad and coauthors was used.⁴¹ Briefly, cells were mechanically detached from culture plates in undifferentiated condition, and cell aggregates were cultured in low-attachment plates to generate EBs. At day six, EBs were cultured in Matrigel-coated plates using neurobasal medium with 2x N2 supplement, 10 ng/ml FGF2, 20 ng/ml EGF, 500 ng/ml Noggin for 3 d, then cultured in neurobasal medium with 2x N2 supplement, 10 ng/ml FGF2, 20 ng/ml EGF, 500 ng/ml Noggin, 20 ng/ml ciliary neurotrophic factor (CNTF, R&D Systems) for additional 3 d. Cell aggregates were then transferred and cultured with neurobasal medium supplemented with 2x N2, 10 ng/ml FGF2, 500 ng/ml Noggin, 20 ng/ml CNTF for 3 d, then in neurobasal medium supplement with 2x N2, 20 ng/ml CNTF for additional 3 d. Neural tube-like rosettes were detached mechanically at the day 15 of differentiation and transferred to Matrigel-coated plates in the same medium.

For spontaneous differentiation, mechanically dissociated and harvested hESCs cells were grown in suspension culture without MEFs using the same medium as described in hESCs culturing protocol, but without bFGF for 5 d in absence or presence of 1 μg/ml tetracycline. At the day 5 of differentiation, formed EBs were attached onto 0.1% gelatin-coated culture plates, then cultured for additional 5 d in the absence or presence of 1 μg/ml tetracycline, as indicated. At the day 10 of differentiation, the culture medium was changed, and cells were cultured in the absence or presence of 1 μg/ml tetracycline for additional 5 d as indicated. At the day 15 of differentiation, cells were harvested and cell lysates were used for western blotting analysis.

Immunostaining

Cells were washed with phosphate buffered saline (PBS), fixed with 4% paraformaldehyde (Sigma) for 15 min at room temperature (RT), washed again three times with PBS, then incubated with PBS-T (PBS + 0.25% triton X-100) for 10 min at RT followed by incubation with 1% bovine serum albumin (BSA) in PBS-T for 30 min at RT. After blocking with BSA, cells were stained in 1% BSA in PBS-T for one hour at RT with the following antibodies: mouse anti-TH (Millipore), rabbit anti-Olig1 (Sigma), rabbit anti-GFAP (Novus), mouse anti-PrP 3F4 (Signet), or humanized anti-PrP P (Signet). Double immunostaining was visualized using combination of goat anti-mouse secondary antibodies conjugated with Alexa 488 or Alexa 546 and goat anti-rabbit secondary antibodies conjugated with Alexa 488 or Alexa 546 (Invitrogen). For antibody P, donkey anti-human secondary antibody conjugated with Alexa 546 (Invitrogen) was used. Hoechst 33342 was used for staining the nucleus. After staining, cells were mounted with antifade mounting fluorescence medium (Dako). Microscopy was performed on an inverted microscope (Nikon Eclipse TE2000-U) with an illumination system X-cite 120 (EXFO Photonics Solutions). Digital images were acquired using a cooled 12-bit CoolSnap HQ CCD camera (Photometrics). Images were processed with WCIF ImageJ software (National Institute of Health, Bethesda, MD). The experiments were performed in triplicate; at least three fields of view were analyzed for each repeat for statistical analysis.

Western blots and data analysis

Cells were lysed in ProteoJET Mammalian Cell Lysis Reagent (Fermantas) with a protease inhibitor cocktail (Roche), and insoluble materials were precipitated by centrifugation at 16,000 X g for 15 min at 4°C. The supernatant was transferred to a new tube and protein concentrations were determined by absorbance spectroscopy. Twelve micrograms of each lysate was resolved by 12% sodium dodecyl sulfate-PAGE (SDS-PAGE), transferred to a polyvinylidene fluoride (PVDF) membrane (Millipore), and then immunoblotted with rabbit anti-Syn (Cell Signaling), rabbit anti-Olig1 (Sigma), rabbit anti-GFAP (Novus) or mouse anti-β-actin (Sigma) antibody. Binding of primary antibody was detected by incubating blots with horseradish-peroxidase-conjugated goat anti-mouse or anti-rabbit antibody, and blots were developed using chemiluminescence (Thermo Scientific).

To analyze expression level of specific proteins, densitometry analysis for each protein band and corresponding background was performed using WCIF ImageJ software, then background was subtracted. Then, the resulting net value representing the expression level of an individual protein was normalized to net value representing the expression level of a loading control (β-actin) in a corresponding sample. Three independent experiments were performed and the data represent a mean ± SD from three experiments. Statistical significance was determined by Student's t test: *, P < 0.05.

Acknowledgment

We thank Pamela Wright for editing the manuscript.

Funding

This work was supported by a Postdoctoral Fellowship Grant of Maryland Stem Cell Commission to YJL and NIH grant NS045585 to IVB.