Engineering of midbrain organoids containing long-lived dopaminergic neurons.

TLDR: It was found that 3-week-old neurospheres were optimal to generate 3D tissue containing DA neurons with typical A9 morphology, and should also become a useful biomaterial for studies on Parkinson's disease.

Abstract: The possibility to generate dopaminergic (DA) neurons from pluripotent stem cells represents an unlimited source of material for tissue engineering and cell therapy for neurodegenerative disease. We set up a protocol based on the generation of size-calibrated neurospheres for a rapid production (3 weeks) of a high amount of DA neurons (>60%) oriented toward a midbrain-like phenotype, characterized by the expression of FOXA2, LMX1A, tyrosine hydroxylase (TH), NURR1, and EN1. By using γ-secretase inhibitors and varying culture time of neurospheres, we controlled maturation and cellular composition of a three-dimensional (3D) engineered nervous tissue (ENT). ENT contained neurons and glial cells expressing various markers of maturity, such as synaptophysin, neuronal nuclei-specific protein (NeuN), and glial fibrillary acidic protein (GFAP), and were electrophysiologically active. We found that 3-week-old neurospheres were optimal to generate 3D tissue containing DA neurons with typical A9 morphology. ENT generated from 4-week-old neurospheres launched glial cell type since astrocytes and myelin could be detected massively at the expense of TH-immunoreactive neurons. All γ-secretase inhibitors were not equivalent; compound E was more efficient than DAPT in generating DA neurons. This DA tissue provides a tool for drug screening, and toxicology. It should also become a useful biomaterial for studies on Parkinson's disease.

Figures

FIG. 2. ENT generation and electrophysiological and morphological analysis. (A) Schematic representation of air–liquid interface culture principle. Three- or 4-week-old neurospheres (21 or 28 days) were plated onto PTFE semipermeable membrane. (B) Neurospheres fused rapidly and formed a compact cell mass observed in 24 h that gave rise to a 3D engineered neural tissue (ENT). (C) Macroscopic (inset) and microscopic views of ENT; arrow indicates neurite outgrowth in the periphery. (D) Electrophysiological analysis of 4-week-old ENT. A paired-pulse stimulation of the ENT with 50-ms interval induced EFP (1st upper diagram). Stimulation artifacts are indicated by arrows. When interval between the two stimuli was set to 15 ms (2nd upper diagram), we observed a net diminution of the second EFP indicating a paired-pulse inhibition. Spontaneous activity from four electrodes (D, lower diagram). An enlargement of one spike is displayed in the inset. Dual-smad inhibition with small molecules (LDN/SB) produced a homogenous neural population in 3D culture. ENTs derived from differentiation of hPSCs were fixed in paraformaldehyde after 1 month in culture and sectioned for analysis. (E) Cresyl-violet-stained transversal section displayed the homogenous aspect of the 3D culture. Immunohistochemistry showed nestin-immunoreactive cells present in the whole ENT (F), confirming the neural homogeneity of the 3D structure. b3-Tubulin immunoreactivity was distributed heterogeneously; it was absent from rosette-like structures that were nestin immunoreactive. The asterisk (*) indicates the air–liquid culture interface and (m) indicates the PTFE membrane. Scale bar: (F) 50 mm, (B, C) 1 mm. 3D, three-dimensional; EFP, evoked field potential; ENT, engineered nervous tissue; PTFE, polytetrafluoroethylene. Color images available online at www.liebertpub.com/scd

FIG. 5. (Continued).

FIG. 5. (Continued).

FIG. 3. Patterning toward midbrain development. (A) Expression of floor plate markers (FOXA2 and LMX1A) and midbrain markers (NURR1 and EN1) was detected in 3-week-old neurospheres by qualitative PCR. GAPDH and CYCLOPHILIN served as housekeeping genes. (B) TH expression in DA ENTs issued from 2-, 3-, and 4-week-old neurospheres was evaluated by quantitative real-time PCR. After 1 week of culture, optimal expression of TH was detected in ENTs issued from 3-week-old neurospheres (data represent mean–SEM, n= 3, significance level *P< 0.05, one-way ANOVA). (C) Comparison of the dopaminergic potential of two ways to produce NPCs for ENT production. NPCs from neurospheres (HS415 3D) generated more TH expression in ENTs than NPCs from HS415 2D culture (data represent mean–SEM, n= 3, significance level P< 0.001, Student’s t-test). (D) Complete dissociation of 3-week-old neurospheres and plating onto 2D polyornithin/laminin surface gave rise to high amount of cells expressing immunoreactivity for floor plate markers FoxA2 and Lmx1a (> 80%) and a high amount of TH-immunoreactive cells coexpressing FoxA2 and Nurr1 (> 70%). (E) When 3- week-old neurospheres are used to form ENTs, the cells show a decrease of expression for all these markers with a change in localization of Nurr1, becoming cytoplasmic, and FoxA2 and TH less colocalized (for 2D plating and ENT, immunofluorescence and cell counting, mean–SEM, n= 3). PCR, polymerase chain reaction; TH, tyrosine hydroxylase. Color images available online at www.liebertpub.com/scd

FIG. 4. Characterization of TH cells present in 2D and ENT cultures. (A) TH-immunoreactive neuron morphology inside an ENT derived from neurosphere culture analyzed by confocal microscopy; the arrow shows typical varicosities of DA neurons (left-hand-side picture). TH-immunoreactive neuron morphology on 2D culture surface coated with laminin after dissociation of neurospheres (right-hand-side picture). (B) Top view of total ENT immunostained for b3-tubulin or TH and generated from hES cells (upper row) or from iPSCs (lower row). In the case of hES cells large peripheral bundles of neurites were found outgrowing from ENT (delimited by dotted outline). These neurite extensions were immunoreactive for b3-tubulin and TH. In the case of iPSC line no neurite extensions were observed although b3-tubulin and TH were detected inside the ENT. (C) Analysis of dopamine present in 2D cultures and ENT: representative high-pressure liquid chromatography chromatogram from ENT lysate for adrenaline and dopamine analysis (arrow, left side of the figure) and quantification of dopamine present in ENT (n = 5) and 2D (n= 3) culture lysates (right-hand-side picture). Each ENT was formed from 106 cells (DA quantification was reported for 0.5 · 106 cells compare with 2D culture containing 0.5 · 106 cells/well). The cell nuclei are stained in blue by DAPI. Scale bar: (A) 50mm, (B) 1 mm. DA, dopaminergic; hES, human embryonic stem. Color images available online at www.liebertpub.com/scd

FIG. 1. Principle of the generation of neurospheres. (A) Standardized protocol for generation of calibrated neurospheres derived from hPSCs. Diagram of culture conditions (A). hPSCs (HS415 cell line) cultured on human foreskin feeder cells (B) were shifted onto human extracellular matrix for expansion (C). Neural induction started at day 0 with cells cultured on human matrix. At day 1, cells detached from matrix as single cells were seeded in microwell plates (Aggrewell ) (D). After 24 h of culture, spherical aggregates, visible within each microwell (E), were harvested for cell culture in suspension in a six-well plate in rotation (F). Pluripotent stem cell markers detected by flow cytometry (G, upper row and histograms) were expressed in the majority of cells seeded on human matrix (> 90% of cells were Oct, Nanog, and Sox2 positive). After 1 week of culture in suspension (G, lower row and histograms), sphere analysis showed a rapid loss of pluripotency markers ( < 1% cells express Oct 3⁄4 or Nanog). A complete conversion of PSCs into NPCs was observed ( > 98% of cells within neurospheres express Sox2). Flow cytometry data represent mean – SEM, n= 5. LDN: LDN193189, SB: SB431542. Maturation compounds consist of glial-cell-derived neutrophic factor, brain-derived neutrophic factor, transforming growth factor type 3, fibroblast growth factor 20, dibutyryl cAMP, ascorbic acid, trichostatin A, compound E, or DAPT. hPSCs, human pluripotent stem cells; NPC, neural progenitor cell. Color images available online at www.liebertpub.com/scd

Frequently asked questions

Q1. What contributions have the authors mentioned in the paper "Engineering of midbrain organoids containing long-lived dopaminergic neurons" ?

In this paper, a 3-week-old neurosphere was used to generate a 3D brain organoid from human pluripotent stem cells ( HPSC ). 

Q2. What is the reason for the higher number of TH cells in 2D culture?

One explanation concerning the higher number of TH cells in 2D culture after neurosphere dissociation in comparison to ENT culture could be due to laminin-coated plates in 2D culture; the matrix could participate in increasing the survival of TH cells. 

Q3. What is the effect of compound E on the development of mature neurons?

Immunoreactivities for the vesicular glutamate transporter VGLUT1 for glutamatergic neurons were strongly enhanced under compound E, suggesting the development of specific subtypes of mature neurons. 

Q4. What are the main advantages of using hPSCs in tissue engineering?

In addition to translational applications, hPSC-based technologies are important for basic research, providing unique insides into mechanisms of human brain development. 

Q5. How was the separation of the analytes achieved?

Separation of the analytes was achieved on a reversed-phase column, Symmetry C-18, 5mm (4.6 · 150 mm2) (Waters Corporation), in isocratic mode at a flow rate of 1 mL/min. 

Q6. Under what conditions did GFAP affect the differentiation of ENTs?

it is worth noting that under DMSO conditions only few areas were GFAP positive and hence many PCNA-positive cells were GFAP negative, possibly corresponding to nestin-immunoreactive neural progenitors. 

Q7. What are the drawbacks of using two-dimensional cell culture models?

Classical two-dimensional (2D) cell culture models are valuable tools for studying cell behavior under stringently controlled conditions. 

Q8. What is the role of g-secretase inhibitor in ENT?

The use of a g-secretase inhibitor (compound E) seemed to be another key element to accelerate ENT maturation and DA neuron production by acting likely through inhibition of the Notch pathway [21–23]. 

Q9. What is the way to generate neurospheres?

generation of neurospheres consists of unwieldy methods with manual scraping of adherent ES cells and/or mechanically cut of rosette structures containing NPCs, in order to release clumps of cell aggregates in suspension culture [29,30]. 

Q10. What was used to maintain pluripotency features of ES cells?

Media used to maintain pluripotency features of ES cells were removed and changed to a xeno-free media containing a serum replacement (X-vivo 10; Lonza). 

Q11. What was the reaction time for the paired pulses?

Paired-pulse evoked field potentials(EFPs) were recorded in response to stimulation of typically 2–4,000 mV every 5 s, with paired-pulse intervals of 50 or 15 ms. 

Q12. How many microwells were used to produce ES cells?

Human ES cells were then detached from matrix and aggregated inside aggrewell plates containing up to 4,700 microwells per well according the model used (Stemcell Technologies). 

Q13. What is the effect of compound E on the development of TH-immunore?

Another striking effect of compound E was the enhancement of TH- immunoreactive cells (Fig. 5A), However, other g-secretase inhibitors, such as DAPT, were not as efficient in generating TH-immunoreactive neurons. 

Q14. What is the role of gamma secretase in neurogenesis?

Among gamma-secretase substrates Notch1 alone is sufficient to block neurogenesis but does not confer selfrenewal properties to neural stem cells. 

Q15. What was the reason for the inhibition of the second pulse?

This type of pairing of two pulses in rapid succession led to an inhibition known as1538 TIENG ET AL.1539paired-pulse inhibition and was generally due to the recurrent inhibition of the subsequent response initiated in the neural network by the second stimulus delivered shortly (15 ms) after the first one. 

Q16. How many TH cells were detected after 1 week of culture?

Surprisingly the authors observed a clear reduction of cells expressing the previous markers (FoxA2, Lmx1a, TH, and Nurr1) and the number of TH cells fell down from > 60% to *30% after 1 week of culture, an observation confirmed over this period and up to 1-month culture (data not shown) (Fig. 3E). 

Q17. How can the authors generate a 3D ENT from hPSCs?

The authors have designed a protocol to generate 3D DAengineered human nervous tissue from hPSC-derived NPCs capable of long-term survival. 

Q18. How did the ENTs appear after 1 month in culture?

After 1 month in culture, macroscopically the ENTs appeared as a homogenous tissue with neurite outgrowth at the periphery (Fig. 2C, arrow).