Highly efficient differentiation and enrichment of spinal motor neurons derived from human and monkey embryonic stem cells

Abstract

Background: There are no cures or efficacious treatments for severe motor neuron diseases. It is extremely difficult to obtain naïve spinal motor neurons (sMNs) from human tissues for research due to both technical and ethical reasons. Human embryonic stem cells (hESCs) are alternative sources. Several methods for MN differentiation have been reported. However, efficient production of naïve sMNs and culture cost were not taken into consideration in most of the methods.

Methods/principal findings: We aimed to establish protocols for efficient production and enrichment of sMNs derived from pluripotent stem cells. Nestin+ neural stem cell (NSC) clusters were induced by Noggin or a small molecule inhibitor of BMP signaling. After dissociation of NSC clusters, neurospheres were formed in a floating culture containing FGF2. The number of NSCs in neurospheres could be expanded more than 30-fold via several passages. More than 33% of HB9+ sMN progenitor cells were observed after differentiation of dissociated neurospheres by all-trans retinoic acid (ATRA) and a Shh agonist for another week on monolayer culture. HB9+ sMN progenitor cells were enriched by gradient centrifugation up to 80% purity. These HB9+ cells differentiated into electrophysiologically functional cells and formed synapses with myotubes during a few weeks after ATRA/SAG treatment.

Conclusions and significance: The series of procedures we established here, namely neural induction, NSC expansion, sMN differentiation and sMN purification, can provide large quantities of naïve sMNs derived from human and monkey pluripotent stem cells. Using small molecule reagents, reduction of culture cost could be achieved.

Figure 1. Noggin Treatment Induces Neural Differentiation of Primate Embryonic Stem Cells.

(A) Schematic procedures of a highly efficient motor neuron differentiation system from primate embryonic stem cells are shown. Red text shows medium components and white text in blue boxes shows culture days. (B–G) KhES-1 human or CMK6 cynomolgus monkey ESCs were cultured in the presence of recombinant mouse Noggin for 10 days (Passage 1; P1, B and F) and 7 more days (Passage 2; P2, C and G). Cells were fixed with 4% PFA and stained with antibodies against specific markers: Nestin (NES: neuroectoderm), Desmin (DES: mesoderm), α-fetaprotein (AFP: endoderm), and Oct4 (OCT: undifferentiated ESC) at P1 and P2. Black bar in BF image and white bar in NES image indicate 25 µm (B). Dorsomorphin-induced neural differentiation is shown in D and E.

Figure 2. ATRA/Shh Promotes Differentiation to Spinal Cord Motor Neurons.

Dissociated NRs, derived from hESCs or monkey ESCs, were cultured in the presence of ATRA and Shh for 7 days. When 1 µM ATRA and 500 ng/ml Shh were supplemented in the differentiation culture (B, D, F and H), larger numbers of βIII-tubulin+/HB9+ (B and F) and βIII-tubulin+/Isl1+ cells (D and H) were observed. Insets in A–H show high magnification. (I and J) The ratio of HB9- or Isl1-positive cells was calculated from the hESC differentiation culture (I) and monkey ESC differentiation culture (J). Scale bar in A indicates 100 µm. Inset scale bar indicates 10 µm. *p<0.05 (n = 5).

Figure 3. ATRA/Shh Strongly Changes Characteristics Rostrally to Caudally.

(A–D) Spinal motor neuron marker HB9 and rostral neural marker BF1 in the control (A and C) or 1 µM ATRA and 500 ng/ml Shh culture conditions (B and D). (E) HB9 expression was strongly induced by ATRA/Shh treatment while BF1 expression was strongly repressed. *p<0.05 (n = 3).

Figure 4. ATRA/Shh Induces the Expression of Spinal Cord-Specific and Motor Neuron-Specific Genes.

(A–C) The expression of sMN-specific genes such as HB9, Isl1, and Olig2 was upregulated by the treatment of 1 µM ATRA and 500 ng/ml Shh. (D) A spinal cord-specific gene, HoxB4, was also upregulated in ATRA/SAG-treated cells compared to the control without ATRA/Shh. In contrast, the rostral brain marker BF1 expression was downregulated in ATRA/Shh-treated culture compared to the control culture (E). (F) There was no statistically significant difference in the expression levels of the neuronal marker βIII-tubulin between ATRA/Shh culture and the control. The gene expression levels in the control culture were defined as 1.0.*p<0.05, **p<0.005 (n = 3).

Figure 5. Shh Agonist Acts as a Motor Neuronal Differentiation Factor Equivalent to Shh Protein.

(A) ATRA (1 µM) and SAG (10 to 1,000 nM) treatments as well as 1 µM ATRA and 500 ng/µl Shh treatment greatly increased both HB9+ and Is11+ cells. (B) Isl1/βIII-tubulin double-positive cells were observed in ATRA/SAG culture. White bar indicates 20 µm. (C–G) Quantitation of the gene expression levels in ATRA/SAG-treated culture. Two sMN-specific markers, HB9 and Isl1, were upregulated by ATRA/SAG treatment (C, D). A spinal cord marker, HoxB4, was also upregulated (E), while a forebrain marker, BF1, was downregulated by addition of ATRA/SAG (10 and 100 nM) (F). The change in expression of the neural cell marker βIII-tubulin was not statistically significant (G). *p<0.05, **p<0.005 (n = 4).

Figure 6. Mature Motor Neurons are Electrophysiologically Functional and Form Synapses with Myotubes.

(A) A mature sMN marker, choline acetyl transferase (ChAT), was detected in hESC derived-neurons in the long-term culture. Mouse E14.5 spinal cord primary culture (mouse SC) was used as a positive control. White bar indicates 20 µm. (B) A current clamp recording showed repetitive firing of action potentials in response to the injection of a 1-second current pulse (50 pA). (C) A voltage clamp recording showed an inward current in response to the addition of 100 µM glutamate (V_holding = −60 mV). Red line indicates the glutamate application (D and E). A voltage clamp recording (D) and current density-voltage relationship (E) show inward currents in response to depolarizing voltage steps from −60 mV to 40 mV. (F and F') Terminals of βIII-tubulin+ neurons (green) were co-localized with the signals of an acetylcholine receptor marker α-bungarotoxin (red) on C2C12-derived myotubes (arrowheads). High magnification picture (F') indicates the area of open white square in F. (G and G') α-bungarotoxin signals (red, arrowheads) was detected nearby neural synaptic marker, Synapsin (green). Inset in G shows high magnification picture. Bars in F' and G' indicate 20 µm.

Figure 7. FGF2-treated Neurospheres Produce More sMNs Than Do EGF/FGF2-treated Neurospheres.

(A) Both FGF2- and EGF/FGF2-treated neurospheres passaged though sphere numbers were gradually decreased depending on the number of passages. (B) Neurospheres in two different culture conditions showed no differences in morphology. Black bar indicates 100 µm. (C) The HB9+ sMN differentiation ratio in EGF/FGF2-treated neurospheres decreased more rapidly than in FGF2-treated neurospheres. (D) Immunocytochemistry of differentiated cells derived from EGF/FGF2-treated neurospheres. White bar indicates 20 µm. *p<0.05, **p<0.005 (n = 4).