establishment of an alginate based 3d culture system for

75
Michael von Troyer, 01216463 Establishment of an alginate based 3D culture system for the generation of dopaminergic neurons MASTER THESIS Submitted at the Faculty of Biology Master's Program Molecular Cell and Developmental Biology LEOPOLD-FRANZENS-UNIVERSITÄT INNSBRUCK and performed at the Institute for Biomedicine – Neuromedicine group Eurac Research, Bolzano Supervisor: Univ-Prof. Dr. Frank Edenhofer Institute of Molecular Biology, Genomics, Stem Cell Biology and Regenerative Medicine Faculty of Biology External Supervisor Dr. Alessandra Zanon Institute for Biomedicine at Eurac Research Innsbruck, 2020

Upload: others

Post on 30-Nov-2021

3 views

Category:

Documents


0 download

TRANSCRIPT

Michael von Troyer, 01216463

Establishment of an alginate based 3D culture system for

the generation of dopaminergic neurons

MASTER THESIS

Submitted at the Faculty of Biology

Master's Program Molecular Cell and Developmental Biology

LEOPOLD-FRANZENS-UNIVERSITÄT INNSBRUCK

and

performed at the

Institute for Biomedicine – Neuromedicine group

Eurac Research, Bolzano

Supervisor:

Univ-Prof. Dr. Frank Edenhofer

Institute of Molecular Biology, Genomics, Stem Cell Biology and Regenerative Medicine

Faculty of Biology

External Supervisor

Dr. Alessandra Zanon

Institute for Biomedicine at Eurac Research

Innsbruck, 2020

I

TABLE OF CONTENTS

TABLE OF CONTENTS ................................................................................................................................ I

1. ACKNOWLEDGMENTS ..................................................................................................................... 1

2. ABSTRACT ........................................................................................................................................ 2

3. INTRODUCTION ............................................................................................................................... 4

3.1. Parkinson’s disease (PD) .......................................................................................................... 4

3.1.1. Clinical and pathophysiological features of PD ............................................................... 4

3.1.2. Genetic causes of PD ....................................................................................................... 6

3.2. Modeling human neurological disease by using induced pluripotent stem cells (iPSCs) ....... 9

3.2.1. iPSCs and disease modeling............................................................................................. 9

3.2.2. Midbrain dopaminergic neuronal differentiation ......................................................... 10

3.3. 3D cell culture models and techniques ................................................................................. 12

3.3.1. Extracellular Matrices (ECM) and Hydrogels ................................................................. 14

3.3.2. Alginate .......................................................................................................................... 15

3.3.3. Applications of 3D cultures in vitro ............................................................................... 17

4. AIM OF THE STUDY ........................................................................................................................ 19

5. PATIENTS, MATERIALS AND METHODS ......................................................................................... 20

5.1. Patients .................................................................................................................................. 20

5.2. Materials ................................................................................................................................ 20

5.2.1. Chemicals ....................................................................................................................... 20

5.2.2. Western blot solutions .................................................................................................. 21

5.2.3. Cell culture reagents, differentiation factors, and coating materials ........................... 22

5.2.4. Cell culture media .......................................................................................................... 23

5.2.5. Kits ................................................................................................................................. 23

5.3. Methods ................................................................................................................................ 24

5.3.1. hiPSC culture and passaging .......................................................................................... 24

5.3.2. Harvesting and counting cells for neuronal in vitro differentiation .............................. 24

5.3.3. Monolayer midbrain dopaminergic differentiation of iPSCs ......................................... 24

5.3.4. Alginate bead midbrain dopaminergic differentiation of iPSCs .................................... 25

5.3.5. Preparation of alginate solution .................................................................................... 25

5.3.6. Alginate bead formation and culture ............................................................................ 26

5.3.7. Decapsulation and replating of differentiated cells from alginate beads ..................... 26

5.3.8. Cell viability assay .......................................................................................................... 26

5.3.9. Real-time quantitative PCR (RT-qPCR) .......................................................................... 27

II

5.3.10. Gel analysis of RT-qPCR products .................................................................................. 28

5.3.11. Western blot analysis .................................................................................................... 28

5.3.12. Mitochondrial superoxide detection ............................................................................. 30

5.3.13. Immunofluorescence staining ....................................................................................... 30

5.3.14. Vibratome sectioning of alginate beads/cell aggregates .............................................. 33

5.3.15. Electrophysiological characterization ............................................................................ 33

6. RESULTS ......................................................................................................................................... 34

6.1. Test of different matrix compositions ................................................................................... 34

6.1.1. Validation of the beneficial effect of Rho-associated kinase inhibitor treatment on

alginate encapsulated iPSCs .......................................................................................................... 34

6.1.2. Number of cell aggregates within alginate of different compositions .......................... 36

6.1.3. Size of cell aggregates within alginate of different compositions ................................. 36

6.2. Fibronectin supports cell viability during differentiation ...................................................... 38

6.3. Comparative RT-qPCR suggests enhanced mDA differentiation ........................................... 40

6.4. Molecular analysis of early mDA neural differentiation ....................................................... 42

6.4.1. Immunofluorescence staining ....................................................................................... 42

6.4.2. Western blot analysis .................................................................................................... 47

6.5. Electrophysiological analysis of the neuronal cultures ......................................................... 49

6.6. Mitochondrial reactive oxygen species (mROS) .................................................................... 51

6.7. Molecular analysis of long-term mDA neural differentiation on 3D platform ...................... 52

7. DISCUSSION ................................................................................................................................... 57

7.1. Patient-specific neuronal cell models for PD research ......................................................... 57

7.2. Setup of 3D system by encapsulation of hiPSCs with alginate hydrogels ............................. 57

7.3. mDA neuronal differentiation of hiPSCs in 3D alginate culture system ................................ 60

8. SUMMARY ..................................................................................................................................... 63

9. PERSPECTIVES ................................................................................................................................ 64

10. REFERENCES .................................................................................................................................. 65

1

1. ACKNOWLEDGMENTS

At this point, I would like to thank all the people that supported me during the pursuit of the present

thesis. First, I would like to thank my supervisor at the Institute for Biomedicine at EURAC Research,

Dr. Alessandra Zanon, for her advice, constant motivation and great patience.

Special thanks go to Diana Riekschnitz, who was always ready to kindly provide me technical support

regarding the daily laboratory routine, Dr. Alexandros Lavdas for his dedication to the analysis of the

large number of images I produced, Dr. Luisa Foco for providing her expertise regarding the analysis

of RT-qPCR data, as well as Dr. Irene Pichler and Dr. Daniel Wojciechowski for their suggestions

regarding parts of the written thesis. I am also grateful to the Group Leader of the Neuromedicine

group, Dr. Andrew A. Hicks, for making it possible for me to perform this master thesis at the

Institute for Biomedicine at EURAC Research. I want to emphasize my gratitude to all the other

members of the Institute for Biomedicine that I had the pleasure of getting to know and thank them

for their professional support, interesting discussions and the enjoyable conversations during lunch

break which often brightened up my day. Furthermore, I want to thank my internal supervisor at the

Leopold Franzens University of Innsbruck, Univ-Prof. Dr. Frank Edenhofer for opening up the

possibility to perform this external thesis at the Institute for Biomedicine at EURAC Research.

Finally, I would like to give my special thanks to my girlfriend, family and friends for their love and

support.

2

2. ABSTRACT

During the last century, the prevalence of chronic age-related neurodegenerative diseases increased

due to prolonged ageing. Parkinson’s disease (PD) is the second most common neurodegenerative

disorder, affecting 2-3% of the population over age 65 (Poewe et al. 2017). It is characterized by the

deterioration of melanin-containing midbrain dopaminergic neurons in the substantia nigra, which

leads to a set of motor dysfunctions accompanied by non-motor symptoms, both strongly affecting

the patient’s quality of life.

The use of induced pluripotent stem cells (iPSCs) enables the generation of patient-specific neuronal

subpopulations, posing a possibility to overcome the restricted accessibility to disease-affected tissue

for mechanistic studies on PD. However, standardized protocols are needed allowing highly efficient,

cell type specific differentiation. In fact, existing culture systems for neuronal differentiation show a

variable reproducibility and yield neurons with a relatively low degree of functional maturation.

Thus, in this work we attempted to evaluate a new in vitro cell culture model for the neuronal

differentiation of iPSCs within a three dimensional (3D) alginate hydrogel matrix in order to obtain a

cell population more enriched with terminally differentiated neurons relevant for the cellular

modeling of PD (midbrain dopaminergic neurons). We studied the suitability of alginate as 3D

biomaterial scaffold to support mDA differentiation of iPSCs in vitro. Therefore, we evaluated 1% and

2% alginate solutions alone and in combination with the extracellular matrix component fibronectin

in terms of their effect on cell viability and differentiation.

Our results show that iPSCs embedded in alginate proliferate and form cellular aggregates. We found

that the treatment with Rho associated kinase inhibitor Y-27632, previously reported to be essential

for embryonic stem cell survival after cell encapsulation in crosslinked alginate, is crucial also for iPSC

survival after encapsulation. Furthermore, we demonstrate that the addition of fibronectin

significantly increases the number and size of aggregates forming in the alginate scaffold.

By using RT-qPCR at day 10 and 20 of differentiation, we have found that overall neuronal as well as

mature dopaminergic neuron marker gene expression was significantly increased in 1% and 2%

alginate with fibronectin with respect to two-dimensional cultures. These findings are further

supported by immunofluorescence staining generated for a subset of neuronal markers.

To assess functional maturation, we evaluated the electrophysiological functionality of our

differentiated cells at day 30 via whole-cell patch clamp analysis. Active as well as passive

electrophysiological parameters support the increased functionality of neurons grown in alginate

scaffolds with respect to the 2D differentiated cells.

3

Furthermore, we have shown that differentiated dopaminergic neurons can be cultured in our

alginate based 3D culture system for more than 200 days and have confirmed the presence of

synaptic connections via immunofluorescence staining for pre- and post-synaptic markers at this

stage.

In summary, we demonstrate that the described alginate based 3D cell culture system is a viable

alternative to conventional adherent culture methods.

4

3. INTRODUCTION

3.1. Parkinson’s disease (PD)

3.1.1. Clinical and pathophysiological features of PD

Parkinson’s disease (PD) is known as the second most frequent neurodegenerative disorder affecting

2-3% of the population older than 65 years of age (Poewe et al. 2017). It is characterized clinically by

the cardinal motor symptoms resting tremor, bradykinesia (slow movements), rigidity (stiffness) and

postural instability (Lang and Lozano 1998) and pathologically by the selective degeneration of the

dopaminergic (DA) neurons residing in the substantia nigra pars compacta (SNpc) of the ventral

midbrain. Another PD hallmark is the occurrence of intracellular protein aggregates called Lewy

bodies (Wakabayashi et al. 2013; Damier et al. 1999). Even though the clinical picture of PD was

already known in the ancient indian medical system Ayurveda (from 4500-1000 B.C), it was first

described in modern times by James Parkinson, a british doctor and surgeon in 1817 (Manyam 1990;

Lang and Lozano 1998). Since then advances in treatment allow for the effective alleviation of

symptoms and management of the disease but none of the developed treatment methods is curative

(Poewe et al. 2017). In addition to the cardinal motor symptoms, PD patients may show secondary

motor symptoms such as gait disturbance, micrographia, precision grip impairment, and speech

problems. (Moustafa et al. 2016).

Besides these frequently cited motor symptoms, a variety of widespread non-motor symptoms

(NMSs) associated with PD have been described. An Italian study found that the most frequently

reported NMSs are of psychiatric nature. Less frequently, sleep symptoms, gastrointestinal

symptoms and pain are reported (Barone et al. 2009). NMSs are generally already present at disease

onset and some might be observable in the prodromal phase of the disease (Trinh and Farrer 2013).

Hyposmia, constipation, and sleep disorders might even be present up to 20 years before

manifestation of the characteristic motor symptoms (Ascherio and Schwarzschild 2016; Poewe 2017).

PD is assumed to be a multisystemic disorder caused by progressive death of selected but

heterogeneous populations of neurons. Among these neuronal populations, the dopaminergic, and

neuromelanin-containing neurons residing in the SNpc are thought to be highly susceptible.

Estimations indicate that 60-70% of such neurons have already been lost in the ventrolateral layer of

the SNpc at disease onset (Fig. 1). This results in dopamine depletion of the striatum (caudate

nucleus and putamen) as these neurons are forming the nigrostriatal pathway (Lang and Lozano

1998).

5

At the cellular level, an important hallmark of PD is the presence of compact neuronal cytoplasmic

inclusions named Lewy bodies (LBs). The protein α-synuclein is the main component of LBs. α-

synuclein-oligomers and protofibrils are known to be cytotoxic. However, it is unclear whether the

formation of LBs causes cell death or, at least initially, represents a protective mechanism (Beyer et

al. 2009; Braak and Del Tredici 2017). More than 70 other proteins from a variety of classes were

described to contribute to these inclusions (Beyer et al. 2009; Deng et al. 2018). Notably, also other

gene products associated with familial PD, discussed in the following section, were found among the

components forming LBs (Beyer et al. 2009). Neither depletion of DA neurons in the SNpc nor the

presence of LBs alone are of diagnostic value for PD as the latter are found also in the context of

other neurodegenerative disorders, such as Alzheimer’s disease, in the absence of nigral

degeneration. However, their combination is specific for a definitive diagnosis of idiopathic PD

(Poewe et al. 2017; Corti et al. 2011).

FIGURE 1. Neuropathological hallmarks of PD. (a) Macroscopical (inset) and transverse sections of

the midbrain upon immunohistochemical staining for tyrosine hydroxylase (TH), the rate limiting

enzyme for the synthesis of dopamine, are shown. Depigmentation of the SNpc and selective loss of

the ventrolateral parts of the SNpc (right panel) are visible compared to control (left panel). (b-d)

Haematoxylin and eosin staining of the ventrolateral region of the SN showing a normal distribution

of pigmented neurons in a healthy control (part b) and diagnostically significant moderate (part c) or

severe (part d) pigmented cell loss in PD . (e-g) Immunohistochemical staining of α-synuclein shows

round, intracytoplasmic LBs (arrow in part e), more diffuse, granular deposits of α-synuclein (part e

and part f), deposits in neuronal cell processes (part f), extracellular dot-like α-synuclein structures

(part f) and α-synuclein spheroids in axons (part g ) (Poewe et al. 2017).

6

3.1.2. Genetic causes of PD

Prior to 1997, epidemiological findings and twin studies led to the believe that the occurrence of PD

was merely determined by environmental factors (Farrer 2006). This assumption was supported by

the fact that a large number of parkinsonism cases were discovered to be linked with exposition to

pathogens or chemical substances. The term parkinsonism refers to a clinical syndrome found in PD

but also many other conditions. It is characterized by bradykinesia and rigidity and/or rest tremor

(Poewe et al. 2017).

However, beginning with the discovery of SNCA (also PARK1/4) mutations linked to PD in 1997 and

the demonstration that the corresponding presynaptic protein α-synuclein is a major component of

filamentous LBs, the link between genetic mutations and PD has become increasingly clear (Corti et

al. 2011).

TABLE 1. Gene loci and disease-causing genes in PD (Deng et al. 2018).

LOCUS LOCATION GENE SYMBOL

FULL GENE NAME INHERITANCE DISEASE ONSET

LEWY BODIES

PARK1/4 4q22.1 SNCA Alpha synuclein AD Early onset, late-onset*

C

PARK2 6q26 PRKN Parkin RBR E3 ubiquitin protein ligase AR Early-onset NC PARK3 2p13 Park3 Parkinson disease 3 AD Late-onset NC PARK5 4p13 UCHL1 Ubiquitin C-terminal hydrolase L1 AD Early-onset, late-

onset* NC

PARK6 1p36 PINK1 PTEN induced putative kinase 1 AR Early-onset NC PARK7 1p36.23 PARK7 (DJ-1) Parkinsonism associated deglycase AR Early-onset NC PARK8 12q12 LRRK2 Leucine rich repeat kinase 2 AD Late-onset C PARK9 1p36.13 ATP13A2 ATPase 13A2 AR Early-onset NC PARK10 1p32 PARK10 Parkinson disease 10 Unclear Late-onset NC PARK11 1q37.1 GIGYF2 GRB10 interacting GYF protein 2 AD Late-onset NC PARK12 Xq21-q25 PARK12 Parkinson disease 12 X-linked

inheritance Late-onset NC

PARK13 2p13.1 HTRA2 HtrA serine peptidase 2 AD Late-onset, early-onset*

NC

PARK14 22q13.1 PLA2G6 Phospholipase A2 group VI AR Early-onset NC PARK15 22q12.3 FBXO7 F-box protein 7 AR Early-onset NC PARK16 1q32 PARK16 Parkinson disease 16 Unclear Late-onset NC PARK17 16q11.2 VPS35 VPS35, retromer complex component AD Late-onset NC PARK18 3q27.1 EIF4G1 Eukaryotic translation initiation

factor 4 gamma 1 AD Late-onset NC

PARK19 1p31.3 DNAJC6 DnaJ heat shock protein family (Hsp40) member C6

AR Early-onset NC

PARK20 21q22.1 SYNJ1 Synaptojanin 1 AR Early-onset NC PARK21 20p13 TMEM230 Transmembrane protein family 230 AD Late-onset, early-

onset* C

PARK22 7p11.2 CHCHD2 Coiled-coil-helix-coiled-coil-helix domain containing 2

AD Late-onset, early-onset*

NC

PARK23 15q22.2 11p15.4

VPS13C RIC13

Vacuolar protein sorting 13 homolog C RIC3 acetylcholine receptor chaperone

AR AD

Early-onset Late-onset, early onset*

NC NC

AD: autosomal dominant. AR: autosomal recessive. *: few cases. C: confirmed. NC: not confirmed.

To date, 23 loci and 19 autosomal disease-causing genes linked to parkinsonism, including 10

dominant and 9 recessive genes were identified (Tab. 1). About 15% of PD patients have family

7

history and 5-10% suffer from a monogenic form of the disease with Mendelian inheritance (Deng et

al. 2018). These findings led to the distinction between familial PD, which is thought to be caused by

single mutations in a variety of genes inherited in a dominant or recessive Mendelian manner, and

idiopathic (sporadic) PD, a multifactorial disease modulated by both environmental factors and

genetic susceptibility. Notably, the genetic aspect of idiopathic PD was highlighted by genome-wide

association studies, which detected a number of respective genetic risk loci and variants. Overall,

these molecular findings point to a series of pathological events involving deficits in synaptic

exocytosis and endocytosis, endosomal trafficking, lysosome‑mediated autophagy, mitochondrial

maintenance and kinase-signaling pathways (see Fig. 2; Trinh and Farrer 2013; Deng et al. 2018).

FIGURE 2. Molecular mechanisms involved in Parkinson disease. An overview of the major

molecular pathways implicated in the pathogenesis of Parkinson disease and their interactions

(Poewe et al. 2017).

Even though the familial cases of PD account only for a small part of the overall disease burden, they

deserve much attention. Their exploration may greatly enhance the understanding of the more

common idiopathic condition by revealing underlying molecular disease mechanisms (Trinh and

Farrer 2013).

SNCA was the first gene found to be associated with PD and encodes the presynaptic neuronal

protein α-synuclein. It contains 6 exons, spans 117 kb and is translated to a small, approximately 15

8

kDa protein comprised of 140 amino acids. The PD-linked mutations in SNCA include dominantly

inherited amino acid substitutions and genomic multiplications (Trinh and Farrer 2013). Abnormal

aggregation of α-synuclein, a consequence of its misfolding, leads to the formation of toxic fibrils,

which are thought to affect cellular homeostasis and to cause neuronal death. Interestingly, an

overexpression of wildtype α-synuclein suffices to cause dose dependent disease phenotypes in

human carriers of SNCA multiplications (Fuchs et al. 2007; Corti et al. 2011). Several post-

translational modifications as well as impaired degradation by the ubiquitin-proteasome system

(UPS) and lysosomal pathways are influencing its propensity to aggregate (Fig. 2). Moreover, a prion-

like propagation has been proposed as mechanism of α-synuclein aggregation (Poewe et al. 2017).

This hypothesis suggests that aggregated α-synuclein can spread intra-axonally and seed aggregation

of endogenous α-synuclein upon endocytosis in new host cells.

The large 51-exon gene coding for the multi-domain protein leucine-rich repeat kinase 2 (LRRK2; also

PARK8) represents another important gene linked to autosomal dominantly inherited PD. The LRRK2

protein has well-defined GTPase and kinase functions and multiple biological roles in striatal

neurotransmission, neuronal arborization, endocytosis, autophagy and immunity. The clinical

phenotype of patients with mutations in this gene consists, in most cases, in a late-disease onset and

closely resembles idiopathic PD (Corti et al. 2011; Trinh and Farrer 2013). The pathology of LRRK2

related PD is heterogenous and characterized by incomplete penetrance. LRRK2 mutations usually

result in gain-of-function, with increased GTPase and kinase activity (Deng et al. 2018).

Less than 4% of PD cases in the community are classified as early-onset (≤45 years of age at

diagnosis) or juvenile parkinsonism (≤20 years of age at diagnosis). The majority of early-onset

autosomal recessive PD cases (about 15%) can be explained by parkin loss-of-function mutations

(Trinh and Farrer 2013). The PRKN gene (also PARK2) encodes for parkin, a 465 amino acid protein

which functions as an E3 ubiquitin ligase and is associated to mitochondrial quality control and

turnover. Parkin is primarily cytosolic but is recruited selectively and rapidly to the outer

mitochondrial membrane of depolarized mitochondria upon mitochondrial membrane dissipation to

act as a protective and anti-apoptotic factor by promoting mitophagy (Narendra et al. 2008). Parkin

activity is regulated by phosphorylation through the mitochondrial serine/threonine kinase PTEN-

induced putative kinase 1 (PINK1). In vitro findings demonstrated that PINK1-parkin signaling plays an

important role in mitochondrial quality control by regulating stimulus-induced mitochondrial

clearance (McWilliams and Muqit 2017).

Autosomal recessively inherited loss-of-function mutations in the PINK1 (also PARK6) gene represent

the second most frequent cause of autosomal recessively inherited early-onset PD. In normal

mitochondria, PINK1 transfers to the inner mitochondrial membrane were it is cleaved. In

9

dysfunctional mitochondria, it accumulates in the outer mitochondrial membrane and, in

collaboration with parkin, facilitates the removal of damaged mitochondria. PINK1-KO rats and

PINK1-mutant Drosophila both show DA neuronal loss and motor dysfunctions while DA

neurodegeneration has been identified in homozygous PINK1-KO mice only upon induction of α-

synuclein or MPTP-toxicity (Deng et al. 2018).

The least common form of early-onset autosomal recessively inherited parkinsonism is caused by

mutations in the DJ-1 gene (also PARK7). DJ-1 codes for a 189 amino acid protein, which is a member

of the ThiJ/Pfp1 family of molecular chaperons. Like parkin, DJ-1 is predominantly cytosolic but it

translocates from the cytoplasm to the outer mitochondrial membrane in the presence of oxidative

stress. Thus, it is thought to be involved in neuroprotection. Interestingly, overproduction of DJ-1

partially rescues the phenotype caused by loss of PINK1 but not loss of parkin function in Drosophila.

This points to DJ-1 as acting in a parkin-independent pathway downstream of PINK1. There is no

possibility to distinguish between the clinical phenotypes of PRKN-, PINK1-, and DJ-1- linked PD (Corti

et al. 2011).

3.2. Modeling human neurological disease by using induced pluripotent stem

cells (iPSCs)

3.2.1. iPSCs and disease modeling

In 2006, Takahashi and Yamanaka were able to show the feasibility of cellular de-differentiation from

mouse embryonic fibroblasts (MEF) to iPSCs. They narrowed the selection of genes necessary for the

reprogramming process down to the four so-called Yamanaka factors Klf4, Oct4, Sox2 and c-Myc. The

resulting iPSCs were able to differentiate into derivatives of all three germ layers (ectoderm,

mesoderm, and endoderm) and hold the potential of unlimited capacity for self-renewal (Takahashi

and Yamanaka 2006).

Several approaches have been established for the delivery of the reprogramming factors including

integrating and non-integrating viral vectors, BAC transposons, episomal vectors, proteins and RNA

delivery (Hu and Li 2016). Additionally, during the last decade, substantial efforts have been made to

replace the protein reprogramming factors by small molecules. These are thought to be safer, more

effective and easier to apply (Ma et al. 2017).

The advent of iPSC technology in combination with subsequent differentiation into various cell-types,

such as neurons, cardiomyocytes, myocytes or hepatocytes, has been a breakthrough for the

biomedical field. It enabled patient specific cellular modeling of a large number of diseases and high

throughput drug validation in the context of personalized precision medicine. One of the greatest

10

advantages of this approach is the generation of nearly unlimited numbers of tissue-specific cells

from autologous patient material, which allows for examination of disease specific cellular

phenotypes on the level of individual patients. This is especially important when the accessibility of

affected tissue is not given or linked to a high risk of collateral damage as in the case of neurological

diseases (Playne and Connor 2017).

Importantly, disease related cellular phenotypes could be observed for inherited and also idiopathic

forms of PD using cellular reprogramming technology in several studies (Jungverdorben et al. 2017;

Playne and Connor 2017). Byers et al. showed an increased expression of genes involved in oxidative

stress, protein aggregation and cell death in patient iPSC-derived DA neurons with an SNCA

triplication compared to controls (Byers et al. 2011).

Although the field of PD modeling has advanced with the help of cell-reprogramming, there remain a

number of challenges. iPSC lines can vary significantly regarding their disease-specific phenotype,

even if they are derived from the same individual (Jungverdorben et al. 2017). Even within a cell line,

cell-to-cell variability of the re-differentiated cells is high. Thus improving and choosing the

appropriate differentiation protocols will be essential for the validity of future studies.

3.2.2. Midbrain dopaminergic neuronal differentiation

The in vivo specification and differentiation of midbrain dopaminergic (mDA) neurons has been an

intense area of investigation during the last decades. This research has enabled the in vitro

generation of neurons as a potential cell source for cell replacement therapies, disease modeling,

drug discovery and other applications including the study of DA neuron physiology (Arenas et al.

2015).

DA neurons are characterized by the release of the catecholaminergic neurotransmitter dopamine

and the expression of tyrosine hydroxylase (TH), the rate limiting enzyme in the synthesis of

catecholamines. These neurons are found all over the mammalian central nervous system (CNS),

including the ventral midbrain, and can be divided into three distinct nuclei: the SNpc (A9 group), the

ventral tegmental area (A10 group) and the retrorubral field (A8 group) (Dahlstroem and Fuxe 1964;

Björklund and Hökfelt 1983). The A10 and A8 group neurons co-determine emotional behavior,

natural motivation, reward and cognitive function. In contrast, the A9 group neurons, residing in the

SNpc of the ventral midbrain, are intimately connected to the dorsolateral striatum forming the

nigrostriatal pathway, which predominantly influences motor function and is disrupted in PD.

11

Protocols to induce the differentiation of DA neurons starting from iPSCs can rely on different

principles. One strategy is the blockage of activin/TGFb and BMP signaling pathways by dual SMAD

inhibition. These and other morphogenic signaling pathways are active in embryonic regions with a

mesodermal or endodermal fate and are inhibited by various inhibitors in regions destined to

become neuroectoderm (Alberts 2015; Sanes et al. 2012). However, endogenous inhibitors can be

replaced by small molecules such as LDN193189 for in vitro neural induction as shown in a number of

studies (Boergermann et al. 2010).

Once neural induction is initialized, cell fate has to be further specified in order to obtain the desired

cell-type corresponding to the DA neurons degenerating in the ventral midbrain of PD patients. This

can be achieved by exposition to the morphogen sonic hedgehog (SHH). SHH is expressed in the

notochord during embryogenesis which induces SHH expression in the most ventral region of the

neural tube, the floorplate (FP), which serves as a signaling center controlling ventral identities

(Placzek and Briscoe 2005; Sanes et al. 2012). Smoothened agonists like Purmorphamine and SAG can

be used to substitute or support recombinant human SHH for in vitro differentiation (Sinha and Chen

2006).

In addition to the exposition to SHH as a ventralizing factor, there has to take place specification

towards midbrain specific lineages. This can be attained by the exposition to recombinant Wnt1 and

FGF8, two signaling molecules that together play an important role in vivo in the formation of the

isthmic organizer (IsO), a signaling center defining the midbrain-hindbrain boundary. For in vitro

differentiation the small molecule GSK3-inhibitor CHIR99021 can be used to stimulate the canonical

Wnt signaling pathway and substitute recombinant Wnt1 (Reinhardt et al. 2013).

Interestingly, at a certain point SHH has to be downregulated in order to permit mDA neurogenesis.

The main factor repressing SHH seems to be Wnt1 (Joksimovic et al. 2009). However there is

evidence that BMP5 and BMP7 signaling, mediated by SMAD1 and SMAD5, may also be involved in

this process and promote mDA neurogenesis in vivo as well as in human iPSCs and neural stem cells

(Jovanovic et al. 2018).

The differentiation of mDA neurons after neurogenesis is regulated by early factors, which in turn

promote the activity of late transcription factors which regulate the production of appropriate

neurotrophic factors and neurotransmitters. A great collection of neurotrophic factors enhancing

mDA neuron survival in vitro has been identified including the transforming growth factor (TGFβ2/3),

brain-derived neurotrophic factor (BDNF) and members of the glial cell-line derived neurotrophic

factor (GDNF) family (Hegarty et al. 2013, 2014; Arenas et al. 2015). These neurotrophic factors

together with molecules shown to increase the yield of differentiation, such as ascorbic acid and

12

cAMP, are added to the DA progenitors to induce the final differentiation (Volpicelli et al. 2004;

Michel and Agid 1996). The duration of this final step varies between protocols and significantly

influences the neuronal maturity (Hartfield et al. 2014; Sanes et al. 2012).

3.3. 3D cell culture models and techniques

The functionality and health of organs and tissues relies on the functionality of the cells, their

integrity within the tissue and the extracellular microenvironment. Techniques for ex-vivo culturing

single cells and tissues have been developed largely during the last century. The culture of cancerous

cell lines has been established to overcome limited cell division capacities of primary animal cells.

These achievements greatly expanded the knowledge about cellular and molecular biology and

enabled in vitro drug discovery and diagnostics (Dutta and Dutta 2018).

To date, adherent cells are mostly grown on rigid plastic surfaces of culture vessels were they form a

monolayer. Non-adherent cells are cultured suspended in growth medium. Although these methods

provided many insights and enabled the development of vaccines and production of recombinant

proteins, in the long run, these cells do not fully replicate the functionality of their in vivo

counterparts. The rigid plastic or glass surfaces and growth media supplemented with fetal serum

used for 2D cell culture create a non-natural environment (Dutta and Dutta 2018). Moreover,

mammalian cells are structurally supported by the extracellular matrix (ECM), a hydrated

composition of fibrous proteins and glycosaminoglycans. The local composition and properties of the

ECM were revealed to play important roles during embryonic development because they affect cell

growth, differentiation (Bissell and Barcellos-Hoff 1987; Frantz et al. 2010) and cell migration (Perris

and Perissinotto 2000). Consistently, it has been demonstrated that cells or tissues cultured on 2D

substrates do not recapitulate in vivo cell growth and expression of tissue-specific genes (Geckil et al.

2010).

Three dimensional (3D) cell culture methods were already established in the middle of the 20th

century (Dutta and Dutta 2018). Later it was shown in collagen gel cultures that tissue architecture

represents an important factor affecting growth and drug sensitivity (Miller et al. 1985). This

knowledge is well established by now, and modern 3D culture techniques in combination with

genome editing methods are important approaches enabling disease modeling, tissue engineering,

drug screening and ultimately cell replacement therapies (Fig. 3).

13

FIGURE 3. Different approaches for deriving human 3D cultures. Human pluripotent stem cells

(hPSCs) derived from a blastocyst or by reprogramming of somatic cells, adult stem cells or cancer

cells derived from primary tissue can be used to derive microfluidics-based organs-on-a-chip (top),

undirected organoids (middle), and region-specific brain organoids or organ spheroids (bottom).

These 3D cultures can be manipulated with CRISPR–Cas9 and other genome-editing technologies,

transplanted into animals or used for drug screening (Pașca 2018).

The behavior of stem cells is regulated by the mechanical properties of their microenvironment

independent of biochemical cues (Saha et al. 2008). More specifically, the early neurogenic

differentiation of human pluripotent stem cells (hPSCs) is supported by soft substrates with

stiffnesses comparable to neural tissue. When the optimal substrate modulus for neurogenic

differentiation is exceeded this support vanishes, while hPSC colony spread area increases (Keung et

al. 2012). Even cell geometrical cues seem to play an important role as they are thought to alter

various cellular characteristics and thereby cell fate (Bao et al. 2017).

In this context, it is worth noticing that neurons in 3D cultures, in contrast to those grown in two

dimensions, embed themselves in a secreted ECM network (Wang et al. 2005; Grayson et al. 2004;

Bonaventure et al. 1994). Besides cell-ECM contacts, also cell-cell interaction and communication are

thought to be important for in vitro cultured cells to acquire tissue like functionality (Pașca 2018;

Wang et al. 2009). This is highlighted by the study of modular brain spheroids used to model inter-

regional communication. While the migration of GABAergic interneurons towards glutamatergic

14

cerebral cortex neurons could be observed in 3D cultures, it was almost absent in 2D systems (Birey

et al. 2017).

Spheroids are generally defined as cellular aggregates not adhering to a culture substrate. Their

generation is achieved by non-adherent cell culturing methods and does not rely on a provided

scaffold or foreign materials. Typically, spheroids can be formed by dissociation and reaggregation

via culture in low cell adhesion plates, in hanging drops (Fennema et al. 2013) or by exposing the cells

to shear forces in spinner flasks as described by Sutherland et al., who employed the method for the

modeling of carcinoma tissue (Sutherland et al. 1971). However, more recent techniques enable the

increasingly reproducible fabrication of large numbers of complex organotypical spheroids (see Fig.

3; Fennema et al. 2013).

The growth and differentiation pattern of different progenitor cells was found to be greatly

enhanced by the formation of cellular aggregates. This is reflected in the initial steps of many cellular

differentiation protocols (Dutta and Dutta 2018).

3.3.1. Extracellular Matrices (ECM) and Hydrogels

The ECM functions as a network that physically connects cells and governs the functional mechanical

properties of many tissues such as bone, cartilage and skin. The immediate microenvironment of

cells is an important mediator of signals and stimuli, and the mechanical ECM properties alone can

support or inhibit the differentiation of certain cell lineages (Saha et al. 2008; Keung et al. 2012).

ECM components can also bind signaling molecules such as growth factors, cytokines or hormones

and alter their presentation via conformational changes. The characteristics of the ECM can in turn

be controlled by the residing cells. It has become apparent that cellular processes like migration,

adhesion, survival, proliferation and differentiation are altered through ECM modulation (Frantz et

al. 2010; Josephine C. Adams and Fiona M. Watt 1993).

The specific constituents of the ECM include a network of insoluble and soluble protein fibrils

surrounded by and embedded in a hydrated gel made of polysaccharides, glycoproteins,

proteoglycans, and other soluble factors. Collagen, elastin, laminin and fibronectin are some of the

major ECM proteins found in almost all vital organs.

Collagens are the most abundant fibrillary proteins constituting up to 30% of the protein mass of a

multicellular animal. Their fibrous nature is derived from a unique triple helix arrangement (Frantz et

al. 2010; Josephine C. Adams and Fiona M. Watt 1993). Elastin is a structural protein and a major

constituent of elastic fibers that allow tissues to endure repeated stretching and deformation.

Laminin and fibronectin are both glycoproteins, which interconnect ECM components with each

15

other or with cells. Fibronectin is a V-shaped protein consisting of two nearly identical subunits

joined by a pair of disulfide bonds. It has binding sites for both collagen and glycosaminoglycans

(GAGs), which enable it to crosslink these ECM components and facilitates the linking of ECM

components to cells. The proteoglycans are hetero-polymers composed of a protein core associated

to GAG side chains. GAGs are polymers of repeating disaccharide units in the form of a bottle brush

(Dutta and Dutta 2018).

Mechanical, structural and compositional cues of the ECM can drastically alter cell function (Caliari

and Burdick 2016). Cells lacking this extracellular microenvironment tend to display aberrant

behaviors like flattened shape, abnormal polarization, altered drug response and loss of

differentiated phenotype (Engler et al. 2006; Rowlands et al. 2008).

These problems motivated the development of well-defined biomaterials, which are able to partially

mimic the cues provided by the actual ECM. Particularly hydrogels, water swollen networks of

polymers, have proven to be useful for this purpose since they have mechanical properties similar to

many soft tissues and can support cell adhesion and protein sequestration (Geckil et al. 2010). The

advantageous properties of natural polypeptides such as collagen, fibrinogen and fibronectin include

the presence of cellular binding domains allowing for cell adhesion. Furthermore, natural

polysaccharide polymers like alginate, hyaluronic acid and chitosan, structurally resemble GAGs and

provide hydrogels with controlled permeability. However, their mechanical properties are generally

restricted to a certain range, they are often degraded enzymatically or hydrolytically in an

uncontrolled manner and their derivation from natural sources negatively influences the

reproducibility.

Synthetic polymers offer more systematically controllable properties. A prominent bio-compatible

synthetic polymer is poly(ethylene glycol) (PEG). PEG-based polymers are frequently used for 3D cell

culture studies and can be easily modified to hold various functional groups, bioactive epitopes and

biodegradable units. However, synthetic polymers are generally biologically inert, not biodegradable

and often rely on harsh gelation techniques (DeVolder and Kong 2012).

Of course, the possibilities to chemically or physically combine different natural or synthetic

polymers in order to tackle their individual disadvantages are nearly endless (Gribova et al. 2011).

3.3.2. Alginate

Alginate is a natural polymer widely used as biomaterial for 3D cell culture and tissue engineering.

Commercial alginates are extracted from marine brown algae (Pawar and Edgar 2012). Due to its high

water retention, cation binding, biocompatibility and non-immunogenicity, this FDA-approved

16

biomaterial has found numerous applications in biomedical science and engineering including food

industry. It is a linear polysaccharide with blocks of (1-4)-linked β-d-mannuronic acid (M) and α-l-

guluronic acid (G) monomers. The relative amount of each monomer can vary among different

sources. Furthermore, the monomers can be linked randomly or grouped in homo-polymeric blocks.

In order to obtain a hydrogel, the individual linear alginate polymer chains have to be crosslinked.

The most common crosslinking procedure is ionic crosslinking (Lee and Mooney 2012), which consists

in the exposure of the unmodified alginate to divalent cations. Typically, Ca2+ is the first choice, but

Sr2+ or Ba2+ may be used as well. During this process, the cations are bridging G residues on different

chains, which results in the formation of an insoluble alginate mesh.

The mechanical properties of alginate hydrogels such as viscosity, stiffness and degradability depend

on chemical characteristics and can be varied by altering composition, concentration, molecular

weight or gelation procedure. The ion type chosen for ionic crosslinking governs the strength and

uniformity of produced hydrogels. Two main gelation methods can be distinguished: for diffusion

gelation, ions are allowed to diffuse from a large outer reservoir (usually a CaCl2 solution) into an

alginate solution, which causes almost instantaneous gelation; internal gelation is based on the

controlled release of gelling ions from an inert ion source that is dissolved or suspended within the

alginate solution. This procedure usually results in a slower and more uniform gelation. As ionic

crosslinking depends on the presence of G blocks, a decreasing M/G-ratio results in stiffer hydrogels

with smaller pore size (Huang et al. 2012; Andersen et al. 2015; Kuo and Ma 2001). The gelation

procedure described in these publications makes scaffold fabrication simple and non-toxic to cells

(Rowley et al. 1999).

Alginates can be considered as inert, as they are not specifically recognized by cell receptors. Thus,

the behavior of cells is mainly influenced by mechanical properties of alginates and can be concerted

via addition of adhesive ligands, fibrous proteins or specific carbohydrates (Huang et al. 2012;

Andersen et al. 2015; Bozza et al. 2014). The tailoring of alginate derivative properties is possible in

principle but can be very challenging in practice due to the key alginic acid properties including

solubility, pH sensitivity, and complexity (Pawar and Edgar 2012).

Alginate hydrogels can be dissolved and cells can be recovered by several agents including chelating

agents such as EDTA, phosphate, citrate or non-gelling agents such as sodium or magnesium ions

(Huang et al. 2012). This also implies that initial mechanical properties will, in the long run, depend

on the environment surrounding the hydrogel (Sakai et al. 2004).

17

3.3.3. Applications of 3D cultures in vitro

One of the fields that adopted 3D cell culture methods is cancer research. Miller et al. compared the

growth properties and sensitivity to chemotherapeutic drugs of mouse mammary tumor

subpopulation cell lines when grown in a collagen matrix or as monolayer culture (Miller et al. 1985).

The growth of the 3D cell aggregates in the collagen matrix was reduced non-exponentially by

chemotherapeutic drugs. This resembled the in vivo behavior of tumor cells more closely than the

exponential growth reduction observed for monolayer cultures (Godugu et al. 2013).

3D cell culture is also routinely used for the support of stem cell growth and differentiation into

several lineages including chondrogenic (Olderøy et al. 2014), hepatic (Baharvand et al. 2006),

pancreatic (Wang et al. 2009) and neuronal lineages (Kim et al. 2013; Li et al. 2011).

Interestingly, 3D cell culture helped to investigate the role of tissue level elasticity as an important

contributor to the specification towards specific lineages in cell cultures. It was shown that

mesenchymal stem cells grown on polyacrylamide gels coated with collagen I commit to the lineage

specified by matrix elasticity. Matrix elasticities corresponding to brain, muscle and bone where

shown to be neurogenic, myogenic and osteogenic, respectively. Soluble induction factors were

found to be less selective than matrix stiffness in governing specification (Engler et al. 2006).

Barharvand et al. reported that the differentiation of human embryonic stem cells (hESCs) into

hepatocyte-like cells is accelerated when they are cultured in 3D collagen scaffolds compared to

collagen-coated dishes (Baharvand et al. 2006).

The study of the central nervous system (CNS) is particularly challenging due to its complexity and

poor accessibility. 2D cell cultures are still the method of choice for many applications such as large-

scale drug screening and imaging. However, cell morphology, direct cell-cell interactions as well as

the cross-talk between specific cell types, seem to be more informative and reminiscent of in vivo

neural tissue in 3D cultures (Pașca 2018; Centeno et al. 2018). Kim et al. proposed that a 3D

environment based on microencapsulation, meaning the embedding of cells in small alginate beads

facilitated the differentiation of human embryonic stem cells to DA neurons (Kim et al. 2013).

The adoption of 3D cell culture platforms suitable in terms of consistency, scale and cost could

greatly increase the relevance of results obtained from high throughput drug screenings. Particularly,

the integration of microfluidics technologies and microfabrication techniques, borrowed from the

microchip industry, holds great promise regarding disease modeling, drug development and

evaluation (Huh et al. 2011; Jackson and Lu 2016; McKee and Chaudhry 2017).

Stem cells, progenitor cells, and lineage-committed cells are considered to be possible drug depots

upon implantation in patients. They could offer sustained release of therapeutic biomolecules.

18

Hydrogels are often used to separate the implanted cells from the host in order to protect them from

the immune system (Schmidt et al. 2008; Wilson and Chaikof 2008). This principle has been applied

by Lim and Sun to correct the diabetic state of rats by the implantation of microencapsulated islets

(Lim and Sun 1980). An even longer lasting effect was obtained with a similar approach, using a more

stable alginate high in guluronic acid, by Soon-Shiong et al. in a spontaneous diabetic dog model

(Soon-Shiong et al. 1993).

Innumerable experimental manipulations are thinkable with 3D organ type cultures, as they could be

used to elucidate known or novel molecular signaling pathways. This enterprise relies naturally on 3D

cell culture and the modeling of an appropriate ECM environment. For instance, the formation of

contractile muscle tissue by neonatal rat cardiac cells was shown to be supported by the integration

of two adhesion peptides into an unmodified alginate scaffold (Sapir et al. 2011).

Engineered tissue is created by harvesting cells, expanding them in culture and subsequently on

scaffolds which could ultimately be used for implantation. Interestingly, alternative technologies

were investigated in which cells could be mixed with a crosslinkable biomaterial and minimal

invasively injected into specific tissues of model organisms. The gel crosslinking would then take

place in situ. This requires less cells, time and reagents and possibly even reduces the chance of

failure due to contaminations. The subcutaneous injection of murine fibroblasts along with a

combination of crosslinked hyaluronan and gelatin or chondroitin sulfate and gelatin into nude mice

for example resulted in viable and uniform soft tissue (Shu et al. 2006).

19

4. AIM OF THE STUDY

The aim of the present thesis was the evaluation of a new in vitro cell culture model system for the

mDA neuronal differentiation of iPSCs within a 3D alginate hydrogel matrix.

For this purpose, the following specific aims were addressed:

1. Testing of different alginate based scaffold biomaterial compositions for the differentiation

of iPSCs to mDA neurons in 3D culture.

2. Comprehensive characterization of the differentiated cells in terms of neuronal and mDA

marker gene expression as well as electrophysiological functionality.

3. Demonstration of long-term maintenance of differentiated mDA neurons with the presented

alginate based 3D culture system.

20

5. PATIENTS, MATERIALS AND METHODS

5.1. Patients

Two hiPSC lines from two control individuals (iPS-802 and iPS-SFC084-03-02), which do not carry any

mutations in known PD genes, were used. iPSC-802 was generated in the laboratory of the Institute

for Biomedicine by electroporation of the parental fibroblast line with non-integrating episomal

plasmids carrying OCT3/4, SOX2, KLF4, and L-MYC (Meraviglia, Zanon et al., 2015). iPSC-SFC084-03-02

was provided by Prof. Philip Seibler at the University of Lübeck. An iPSC line of a PD patient carrying a

heterozygous triplication of the SNCA gene (iPS-ND34391) was provided by the Coriell Institute

biobank. Phenotypic and genotypic data are listed in Table 2. The study was approved by the Ethics

Committee of the Azienda Sanitaria dell’Alto Adige (Italy). For all the experiments, either one or the

other control line was used.

TABLE 2. Genotypic and phenotypic characterization of PD patient and controls

ID Sex Age at biopsy (yr)

Mutation Zygosity Clinical status

Control 802 F 60 - - unaffected

Control SFC084-03-02 F 63 - - unaffected

Mutant ND34391 F 55 SNCA triplication heterozygous affected

5.2. Materials

5.2.1. Chemicals

1kb Plus DNA Ladder Invitrogen

All-In-One qPCR mix Genecopoeia

All-In-One qPCR mix Genecopoeia

Antioxidant Life Technologies

Blotting-Grade Blocker Bio-Rad

Bovine serum albumin (BSA) Sigma-Aldrich

cOmplete™ Protease Inhibitor Cocktail Roche

Ethanol (C2H5OH) J.T.Baker

Ethanol (C2H5OH) J.T.Baker

GelStar™ Nucleic Acid Gel Stain 10,000X Lonza

IGEPAL CA-630 (NP-40) ((C2H4O)nC14H22O) Sigma-Aldrich

21

Isopropanol Sigma-Aldrich

Isopropanol Sigma-Aldrich

Loading Buffer 5X Thermo Fisher Scientific

Methanol (CH3OH) Sigma-Aldrich

Methanol (CH3OH) Sigma-Aldrich

NuPage MES running buffer 20X Life Technologies

NuPage sample buffer 4X Life Technologies

NuPage transfer buffer 20X Life Technologies

Paraformaldehyde (HO(CH2O)nH) Sigma-Aldrich

Paraformaldehyde (HO(CH2O)nH) Sigma-Aldrich

PhosSTOP EASYpack Roche

Precision Plus Protein Western Blotting Standards Bio-Rad

Reducing agent 10X (DTT 500mM) Life Technologies

Sodium azide (N3Na) Sigma-Aldrich

Sodium azide (N3Na) Sigma-Aldrich

Sodium chloride (NaCl) Sigma-Aldrich

Sodium chloride (NaCl) Sigma-Aldrich

Sodium dodecyl suphate (SDS) Fluka

Triton X-100 Sigma-Aldrich

Triton X-100 Sigma-Aldrich

Trizima Base (NH2C(CH2OH)3) Sigma

TRIzol Reagent Thermo Fisher Scientific

TRIzol Reagent Thermo Fisher Scientific

Tween-20 Millipore

5.2.2. Western blot solutions

RIPA buffer 25 mM Tris-HCl pH 7.6, 150 mM NaCl, 1% IGEPAL, 1% sodium

deoxycholate, 0.1% SDS in ddH2O

22

Running buffer 50 NuPAGE MES Running buffer 20X (Thermo Fisher Scientific), 950

mL ddH2O

Transfer buffer 50 mL NuPAGE MES Transfer Buffer 20X (Thermo Fisher Scientific),

100 mL MeOH, 850 mL ddH2O

Ponceau Red 0.1% Ponceau S, 5% acetic acid in ddH2O

TBS 10X 200 mM Tris, 9% NaCl, pH 7.5 in ddH2O

TBS-T 1X 100 mL TBS 10X, 0.1% Tween20, ddH2O to 1 liter

Blocking buffer 5% Blotting-Grade Blocker in TBS-T 1X

5.2.3. Cell culture reagents, differentiation factors, and coating materials

Agarose Bio-Rad

Ascorbic acid Sigma-Aldrich

B-27 Plus Supplement 50X Thermo Fisher Scientific

CHIR99021 Stemgent

Collagenase type IV Thermo Fisher Scientific

DAPT Tocris Bioscience

Dibutyryl-cyclic-AMP EnzoLifescience

DMEM/F12 + GlutaMAX Thermo Fisher Scientific

Fibronectin Corning

KnockOut Serum Replacement Thermo Fisher Scientific

Laminin (LA) Sigma-Aldrich

LDN-193189 Stemgent

Matrigel Basement Membrane Matrix BD Biosciensces

MitoSox Life Technologies

Neurobasal Medium Life Technologies

Penicillin/Streptomycin (P/S) Life Technologies

Phosphate buffered saline (PBS) Life Technologies

23

Poly-D-lysine (PDL) Thermo Fisher Scientific

Purmorphamine Stemgent

Recombinant Human BDNF Peprotech

Recombinant Human FGF-8a R&D System

Recombinant Human GDNF Peprotech

Recombinant Human Sonic Hedgehog R&D System

Recombinant Human TGF-beta 3 Peprotech

SB-431542 Tocris Bioscience

StemMACS iPS-Brew XF Miltenyi

StemMACS Y27632 (RI) Myltenyi Biotech

TrypleExpress Life Technologies

5.2.4. Cell culture media

iPSC medium StemMACS iPS-Brew XF (Myltenyi) (includes iPS-Brew XF 50X

supplement), 1% penicillin/streptomycin (Life Technologies)

Neural differentiation medium Knockout DMEM (Life Technologies), 15% knockout serum (Life

Technologies), 1% L-glutamine (Life Technologies), 1% nonessential

amino acids (Life Technologies), 0.2% -mercaptoethanol (Life

Technologies).

Neural medium Neurobasal medium (Life Technologies), B27 Supplement 1X, 1% L-

glutamine (Life Technologies).

5.2.5. Kits

BCA Protein Assay Kit Pierce

Countess Cell Counting Chamber Slides Thermo Fisher Scientific

Fluo-4 Calcium Imaging Kit Thermo Fisher Scientific

Human Neural Stem Cell Immunocytochemistry Kit Thermo Fisher Scientific

LIVE/DEAD Viability/Cytotoxicity Kit, for mammalian cells Thermo Fisher Scientific

24

Pluripotent Stem Cell 4-Marker Immunocytochemistry Kit Thermo Fisher Scientific

5.3. Methods

5.3.1. hiPSC culture and passaging

hiPSCs were cultured under feeder-free conditions on Matrigel matrix (Corning) in StemMACS™ iPS-

Brew XF (Miltenyi Biotech) with 1% penicillin-streptomycin (Thermo Fisher Scientific). Cells were

maintained in a saturated humidified atmosphere at 37°C and 5% CO2, and medium was changed

every day. iPSC colonies were passaged enzymatically using 1 mg/ml Collagenase type IV (Thermo

Fisher Scientific) in DMEM/F12 (Thermo Fisher Scientific). Briefly, differentiated areas of the iPSC

colonies were removed by slow-vacuum aspiration under a stereomicroscope, and cells were

incubated with collagenase for 10 min at 37°C (approximately until the edges of the colonies are

visibly curling). Next, the enzymatic solution was replaced by DMEM/F12, and the colonies were

detached with a cell scraper and centrifuged at 200 xg for 5 min. The cell pellet was resuspended in

StemMACS iPS-Brew XF medium, and colony fragments were triturated in smaller pieces ( 5̴0-200

cells per fragment). The colony suspension was seeded at a ratio of 1:3-1:5 on Matrigel coated cell

culture plates.

5.3.2. Harvesting and counting cells for neuronal in vitro differentiation

When seeded for neuronal differentiation, cells were passaged at a maximum of 90% confluency

using TrypLE Express (Thermo Fisher Scientific) for 3-5 min at 37°C. Enzymatic detachment was

stopped by the addition of StemMACS iPS-Brew XF. Cells were suspended via pipetting, and the cell

suspension was centrifuged at 200 xg for 3 min. The cell pellet was resuspended in medium

supplemented with 10 µM ROCK inhibitor (RI). For counting, 10 µL of cell suspension were mixed

with 10 µL of Trypan Blue Stain (0.4%) (Thermo Fisher Scientific). 10 µL of this mixture were

transferred into the chamber ports of a Countess ™ Cell Counting Chamber Slide (Thermo Fisher

Scientific), and the cell number was determined using Countess® Automated Cell Counter (Thermo

Fisher Scientific).

5.3.3. Monolayer midbrain dopaminergic differentiation of iPSCs

Prior to 2D monolayer differentiation cells were harvested and counted as described in section 5.3.2.

105 cells/12 well were seeded on cell culture plates coated with 350µg/ml Matrigel (Corning) in

25

StemMACS iPS-Brew XF medium supplemented with 10 μM RI. Daily medium changes were

conducted in the following days with fresh medium without RI.

Once the cells reached 80-90% confluency, the differentiation of iPSCs into midbrain DA neurons was

initiated following the floor-plate-based neuron induction protocol established by Kriks et al., with

minor modifications (Kriks et al. 2011; Zanon et al. 2017). The timed exposure to differentiation

factors, meaning signaling proteins and small molecules, was started by supplying the cells with

knockout serum replacement (KSR) medium supplemented with SMAD pathway inhibitors SB431542

(SB, 10 µM, Tocris Bioscience) and LDN-193189 (LDN, 100nM, Stemgent). LDN was added on days 1-

10 while the supplementation with SB was stopped upon day 6. On days 2-7 Recombinant Human

Sonic Hedgehog (SHH, 100 ng/ml, R&D System), Recombinant Human FGF-8a (FGF8, 200 ng/ml, R&D

System) and Purmorphamine (Pu, 2 µM, Stemgent) were added to the medium. Furthermore, on

days 4-12 the Wnt pathway activator molecule CHIR99021 (CH, 3 µM, Stemgent) was added to the

differentiation medium.

During days 7–10 of differentiation, increasing amounts of Neurobasal medium (Life Technologies)

plus B27 supplement (Life Technologies) (referred to as NB-B27) were added to the KSR medium

(33%, 50%, and 66%). Thereafter it was completely shifted to NB-B27. From day 8, the medium was

changed every other day.

On day 12, maturation of DA neurons was initiated by adding Recombinant Human BDNF (20 ng/ml,

Peprotech), Recombinant Human GDNF (20 ng/ml, Peprotech) ascorbic acid (AA, 0.2 mM, Sigma),

Recombinant Human TGF-beta 3 (ß3, 1 ng/ml, Peprotech), dibutyryl-cyclic-AMP (dbcAMP, 0.5 mM,

EnzoLifescience) and DAPT (10 mM, Tocris) (together referred to as BAGTCD). On day 20 of

differentiation, cells were passaged as single cells and replated at 1.5x105 cells/cm2 on 12 well plates

with coverslips coated with PDL/LA (100 µg/ml; 10 µg/ml). The medium was supplemented with

BAGTCD maturation factors until day 35.

5.3.4. Alginate bead midbrain dopaminergic differentiation of iPSCs

The same differentiation conditions as described above in section 5.3.3 were applied for the 3D

cellular model system.

5.3.5. Preparation of alginate solution

Alginate solutions (1% and 2% w/v) were prepared freshly the day before cell encapsulation by

mixing alginic acid sodium salt (Sigma Aldrich) in a 0.025 M HEPES and 0.15 M NaCl buffer (pH 7.4).

26

The alginate solution was heated to dissolve alginate and subsequently autoclaved and sterile

filtered.

5.3.6. Alginate bead formation and culture

At the day of cell encapsulation, iPSCs were harvested and counted as described in section 5.3.2.

1.0x106 cells were gently mixed with 1 ml alginate solution (1% or 2% w/v) with or without

fibronectin (Fn) to a final concentration of 100 µM. Alginate beads were formed by slowly dripping

the prepared alginate-cell suspension in 0.1 M CaCl2 with a 19 gauge needle. The beads were

incubated for 20 min in the CaCl2 solution and then washed 3 times with 1x PBS pH 7.4 (Thermo

Fisher Scientific). Subsequently they were transferred to cell culture plates (15-20 beads/6 well) with

a cell scraper and supplied with StemMACS iPS-Brew XF medium supplemented with 10 μM RI.

Alginate beads were cultured with StemMACS iPS-Brew XF medium supplemented with 10 μM RI for

three days prior to differentiation if not otherwise stated.

5.3.7. Decapsulation and replating of differentiated cells from alginate beads

Prior to decapsulation alginate beads were washed inside the well with PBS without Calcium and

Magnesium. Subsequently, the beads were distributed to 1.5 ml Eppendorf tubes (3-4 per tube)

which were filled up with PBS without Calcium and Magnesium and shaken on a thermoblock at 37°C

and 700 rpm for 20 min. Then they were centrifuged at 300 xg for 5 min, the supernatant was

discarded, and they were seeded at an appropriate density on cell culture plates with round coverlips

coated with 1 mg/ml Matrigel.

5.3.8. Cell viability assay

The Live/Dead cell viability assay (Live/Dead cell viability assay kit, Life Technologies) was performed

at day 5, 10, and 20 of differentiation. For the assay, beads were collected and washed briefly twice

with PBS. Each bead was then placed in a live imaging chamber and incubated for 20 min with a 2 µM

Ethidium Homodimer-1 (EH-1) and 8 µM Calcein (AM) solution in PBS in dark conditions. Images were

acquired using a Leica SP8-X confocal microscope (Leica) and analyzed using Imaris software.

27

5.3.9. Real-time quantitative PCR (RT-qPCR)

Cells were directly lysed in 1 ml of TRIzol (Invitrogen) reagent at day 10, 20 and 35 of differentiation.

Total RNA isolation was conducted following manufacturer's instructions.

1-2 µg RNA were reverse-transcribed using the SuperScript® VILO™ cDNA Synthesis Kit (Invitrogen) in

a 20 µl reaction. The cDNA synthesis was carried out in a thermocycler following the manufacturers’

temperature protocol (10 min at 25°C, 60 min at 42°C and 5 min at 85°C). The cDNA was then diluted

to 10 ng/µl with RNAse free water. RT-qPCR was performed in triplicate on the 96CFX Manager (Bio-

Rad) in 20 µl, using 25 ng of reverse transcribed product, the All In One qPCR mix (Genecopoeia) and

0.2 µM of primers (Genecopoeia; Table 3).

The PCR program used was the following:

Number of Cycles Step Temperature Time

1 Pre-denature 95°C 10 min

40

Denature 95°C 10 sec

Annealing 60°C 20 sec

Extension 72°C 15 sec

Melting curve analysis was performed immediately after the amplification reactions.

Temperature range Rate of temperature change Duration

72°C ~ 95 °C 0.5°C/ step 6 sec/ step

30°C

30 sec

Relative gene expression was calculated using the ΔΔCt method and the Ct value of the target gene

was normalized to the Ct value of two reference genes (β-actin and RPL13) using the following

equation:

Fold change ≈ 2−ΔΔC𝑡 = 2[ΔC𝑡sample A (target−reference)− ΔC𝑡sample B (target−reference)]

Efficiencies and correlation coefficients (r2) were calculated for all primers.

28

TABLE 3. Overview of primers used for RT-qPCR

Gene Manufacturer Catalog Number

PAX6 Genecopoeia HQP012219

FOXA2 Genecopoeia HQP008906

TUJ1 Genecopoeia HQP057484

TH Genecopoeia HQP018064

DAT Genecopoeia HQP053961

GIRK2 Genecopoeia HQP010010

5.3.10. Gel analysis of RT-qPCR products

In order to check whether the set of primers used for the RT-qPCR experiments gave the correct and

expected amplicon, 5 µl of the PCR products were loaded on 1.5% agarose gel prepared with

Certified™ Molecular Biology Agarose (Bio-Rad) and 0.5X TBE buffer by heating it in a microwave.

GelStar™ Nucleic Acid Gel Stain 10,000X (Lonza) was added to the liquid agarose. Then, the solution

was poured into a casting tray for solidification at room temperature. A molecular weight standard

(Low Molecular Weight DNA Ladder, New England Biolabs) was loaded on the first lane for size

estimation. Electrophoresis was performed at 100 V for 20 min in 0.5 X TBE buffer. The gel was

analyzed using the ChemiDoc™ Touch Imaging System (Bio-Rad).

5.3.11. Western blot analysis

5.3.11.1. Sample preparation

For western blot analysis, whole cell homogenates were used. Cell pellets were resuspended in the

appropriate volume of cold RIPA buffer supplemented with protease inhibitors (cOmplete™ Protease

Inhibitor Cocktail, Roche) and phosphatase inhibitors (PhosSTOP EASYpack, Roche) and then

incubated 30 min on ice. After incubation, the cell lysate was centrifuged for 20 min at 14,200 x g at

4°C. The supernatant, containing the cell proteins, was transferred into a new tube. Protein

concentration was determined using BCA Protein Assay Kit (Thermo Fisher Scientific) according to the

manufacturer’s protocol. Absorbance was measured using VICTOR™ X3 Multilabel Plate Reader

(PerkinElmer) at 562 nm, and sample protein concentration was calculated using a standard

calibration curve (0-2.5 µg/µL). 10 µg of proteins were loaded per well on the SDS-PAGE gel. Protein

lysate was mixed with NuPAGE LDS Sample Buffer (ratio 1:4) and NuPAGE Sample Reducing Agent

29

(ratio 1:10) (Thermo Fisher Scientific). RIPA-Buffer was added in the amount needed to obtain an

equal volume for all samples. Next, the samples were denaturated by incubation for 5 min at 95°C.

5.3.11.2. SDS-PAGE and protein transfer

After denaturation, sample mixtures were loaded on a NuPAGE 4-12% Bis-Tris Gel (Thermo Fisher

Scientific). Molecular weight standard Precision Plus Protein WesternC Blotting Standards (Bio-Rad)

was run on the same gel. The electrophoresis was performed in Running buffer for 5 min at 100 V

and for 90 min at 150 V. After electrophoresis, proteins were transferred at 32 V for 70 min onto a

nitrocellulose membrane (Biorad) in a CellX Mini Blotting Module (Thermo Fisher Scientific) containing

Transfer buffer.

5.3.11.3. Antibody staining for western blot

The blots were blocked in Blocking buffer (see section 5.2.2). Primary antibodies were appropriately

diluted as reported in Table 4, and the blots were subsequently incubated overnight at 4°C on a

radial rotor. The next day, the membrane was washed three times with TBS-T for 5 min and then

incubated with a horseradish-peroxidase (HRP)-conjugated secondary antibody (Millipore) at a

dilution of 1: 10,000 for 1 h at room temperature. This was followed by other three washing steps in

TBS-T. Target proteins were detected by enhanced chemiluminescence using Clarity™ ECL Western

Kit (Bio-Rad), according to the manufacturers’ protocol. The chemiluminescence signal was detected

using the ChemiDoc™ Touch Imaging System (Bio-Rad). Equal loading was assessed by use of an

antibody against a housekeeping protein. The signal of the target protein was quantified by Image

Lab 5.2.1 analyzer software (BioRad) by normalizing the expression of the target protein in relation to

the housekeeping proteins.

30

TABLE 4. Overview of antibodies used for western blot analysis and their respective dilution factor

Primary antibodies Species Manufacturer Dilution factor

FOXA2 (#sc374376) Mouse SantaCruz 1:1000

LMX1A (#ab10533) Rabbit Millipore 1:1000

TUJ1 (#MMS-435P) Mouse Biolegend 1:1000

TH (#657012) Rabbit Callbiochem 1:1000

GAPDH (#MAB 374) Mouse SantaCruz 1:500

Secondary antibodies

Horseradish-peroxidase (HRP)-

conjugated antibody (#AP308P)

Mouse Millipore 1:10,000

Horseradish-peroxidase (HRP)-

conjugated antibody (#12-348)

Rabbit Millipore 1:10,000

5.3.12. Mitochondrial superoxide detection

At days 10, 20 and 35 one well of each differentiated cell line, replated as indicated in section 5.3.7,

was used for mitochondrial superoxide detection. Medium was removed by washing three times with

PBS and replaced by HBSS containing 5 µM MitoSox Red mitochondrial superoxide indicator (Thermo

Fisher Scientific). After an incubation of 15 min at 37°C and 5% CO2, the cells were washed for three

times with HBSS. Subsequently, the cells were fixed with 4% paraformaldehyde at room temperature

and the coverslips were mounted on an object slide as described for immunofluorescence staining.

Images were acquired using a Leica SP8-X confocal microscope (Leica) and analyzed using Imaris

software.

5.3.13. Immunofluorescence staining

Cells differentiated in alginate beads were replated at days 10, 20 and 35 of differentiation on 12 well

plates with Matrigel coated coverslips. Cells differentiated as monolayer were either replated at day

10 of differentiation on 12 well plates with 250 µg/ml Matrigel coated coverslips or at day 20 of

differentiation at 1.5x105 cells/cm2 on 12 well plates coated with PDL/LA (100 µg/ml; 10 µg/ml). Cells

were fixed in a 4% paraformaldehyde solution in PBS for 15 min at room temperature. Afterwards,

cells were washed three times with PBS.

31

All steps of the immunofluorescence stainings were performed on a shaker. If not stated otherwise,

cells were permeabilized for 5 min with 0.5% TritonX100 in PBS, then blocked for 1 h with 3% BSA in

PBS at room temperature and incubated overnight with primary antibodies in 3% BSA at 4°C. For all

stainings, cells were washed three times for 5 min with PBS the next day, incubated with secondary

antibodies for 1h at RT, washed three times for 5 min with PBS again and then mounted on an

object-slide using ProLong Diamond Antifade Mountant with DAPI (Thermo fisher scientific).

For FOXA2 and LMX1A co-stainings, blocking and primary antibody incubation was conducted in PBS

with 3% BSA and 0.05% TritonX100.

For TH and DAT and TH and GIRK2 co-stainings, cells were directly blocked in 10% FBS in PBS without

any permeabilization and primary antibody incubation was performed in 5% FBS in PBS. Images were

acquired using a Leica SP8-X confocal microscope (Leica) and analyzed using Imaris software. The

antibodies used are listed in Table 5.

32

TABLE 5. Overview of antibodies used for immunofluorescence stainings and

respective dilution factors

Primary antibodies Species Manufaturer Dilution factor

FOXA2(HNF-3ß(H-4)) (sc374376) Mouse Santa Cruz 1:50

LMX1A (ab31006) Rabbit Abcam 1:200

MAP2 (ab5392) Chicken Abcam 1:1000

TH (MAB318) Mouse Millipore 1:500

GIRK2 (ab65096) Goat Abcam 1:100

DAT (sc-32259) Rat Santa Cruz 1:50

LMX1A (AB10533) Rabbit Millipore 1:1000

TH (sc-25269) Mouse Santa Cruz 1:50

TUJ1 (MMS-435P) Mouse Covance 1:1000

TH (657012) Rabbit Millipore 1:200

PSD95(7E3) (36233S) Mouse Cell Signaling 1:150

Synapsin 1(D12G5) (52975) Rabbit Cell Signaling 1:100

Secondary antibodies

Alexa Fluor 488 Anti-Mouse IgG Goat Invitrogen 1:1000

Alexa Fluor 555 Anti-Rabbit IgG Goat Invitrogen 1:1000

Alexa Fluor 647 Anti-Mouse IgG Goat Invitrogen 1:1000

Alexa Fluor 488 Anti-Chicken IgG Goat Invitrogen 1:1000

Alexa Fluor 488 Anti-Rat IgG Goat Invitrogen 1:1000

Alexa Fluor 555 Anti-Mouse IgG Donkey Invitrogen 1:1000

Alexa Fluor 647 Anti-Goat IgG Donkey Invitrogen 1:1000

Alexa Fluor 555 Anti-Goat IgG Donkey Invitrogen 1:1000

Alexa Fluor 488 Anti-Mouse IgG Donkey Invitrogen 1:1000

33

5.3.14. Vibratome sectioning of alginate beads/cell aggregates

Cell aggregates of mDA-neurons differentiated and cultured in 1% alginate for more than 200 days

were fixed with 4% paraformaldehyde overnight at 4°C, washed three times with 1x PBS and

embedded in 3-4% agarose. The solid agarose block was sectioned into 80 µm slices using a Leica

VT1000 S vibratome. The sections were stored in 1X PBS at 4°C until used for experimental analysis.

Immunofluorescence stainings on these sections were carried out as indicated in section 5.3.13.

5.3.15. Electrophysiological characterization

Whole-cell recordings in current-clamp mode have been performed in a temperature-controlled

recording chamber (36-37°C) mounted on an inverted Eclipse-Ti microscope (Nikon, Tokyo, Japan)

and using a MultiClamp 700B amplifier (Molecular devices, LLC). Current-command protocols and

data acquisition have been performed using pClamp 10.0 software and the Digidata 1550 interface

(Molecular Devices, LLC). Data were lowpass-filtered at 3 kHz and sampled at 10 kHz. Patch

electrodes, fabricated from thick borosilicate glass capillaries, have been made using a Sutter P-1000

puller (Sutter Instruments) to a final resistance of 4-6 MΩ when filled with the intracellular solution

containing (in mM): 120 KGluconate, 25 KCl, 10 EGTA, 10 HEPES, 1 CaCl2, 4 Mg-ATP, 2 Na-GTP and 4

Na2-Phosphocreatine (pH 7.4, adjusted with KOH). Cells have been bath-perfused with a Krebs

solution composed by (in mM): 129 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 30 D-glucose, 25 Hepes; pH 7.3 with

NaOH. Spontaneous action potentials have been recorded in a gap-free mode while the presence of

the Ih current was evaluated by applying long hyperpolarizing steps at different current intensities (-

30 pA increments). Series resistance value has been monitored along the experiment and recordings

with changes over 20% of its starting value have been discarded.

34

6. RESULTS

The present thesis describes the establishment and validation of a new in vitro cell culture model for

the neuronal differentiation of iPSCs within a 3D alginate hydrogel matrix to mimic the elastic

environment encountered in the developing brain. In the first part of the experimental work, it was

attempted to identify a hydrogel composition and initial culturing conditions, suitable to sustain cell-

viability and propagation. In the second part, iPSC control lines were differentiated into mDA neurons

and the expression of differentiation markers was examined via RT-qPCR and immunofluorescence

staining at different stages of the process. Additionally, the functionality of the differentiated cells

was assessed by electrophysiological characterization. At last the ability of the cells to recapitulate a

known mitochondrial dysfunctional phenotype was examined and they were kept in culture for more

than 200 days. .

6.1. Test of different matrix compositions

6.1.1. Validation of the beneficial effect of Rho-associated kinase inhibitor

treatment on alginate encapsulated iPSCs

Alginates are widely used hydrogels because of their easy gelation at physiological conditions, lower

batch-to-batch variations than provided for animal materials, the possibility of gentle dissolution,

transparency allowing microscopic observations and the formation of a pore network allowing

nutrient, synthesized product and waste diffusion. In addition, they can be considered to be inert as

they are not recognized by mammalian cell surface receptors (Andersen et al. 2015).

It has been reported previously that the treatment with Rho-associated kinase inhibitor (RI) prior to

differentiation is essential to sustain the cell viability of hESCs encapsulated in alginate microcapsules

(Chayosumrit et al. 2010; Kim et al. 2013). For this reason, we tested its beneficial effect also for

encapsulated human iPSCs.

IPSCs were encapsulated in 1% or 2% alginate with or without fibronectin. At the day of

encapsulation, iPSCs were treated with RI. The next day, cells were divided into two groups, one of

which was cultured in StemMACS medium with RI for three additional days, while the other group

was cultured without RI. Subsequently, mDA neuronal differentiation was induced in both conditions

(Fig. 4A). The proliferating cells formed aggregates within the alginate beads, which were counted

manually (Fig. 4B). As whole beads did not fit into the field of view, the counts refer to cell

aggregates per field of view (Fig. 4C). The number of aggregates was low when iPSCs were not

treated with RI for three days prior to differentiation and no significant difference in number of

35

aggregates between the different matrix compositions could be identified in this condition (Fig. 4B,

C). On the other hand, it was apparent that the number of aggregates in the RI pre-treated alginate

beads was significantly higher than in the non-pre-treated beads, with the only exception of 2%

alginate. In addition, there were significant differences between the numbers of aggregates per field

arising from cells encapsulated in different hydrogel compositions within this group (Fig. 4B, C).

FIGURE 4. RI is beneficial for iPSCs encapsulated in alginate beads. (A) Differentiation protocol

adapted from Kriks et al. 2011. SM, StemMACS medium; RI Rho associated kinase inhibitor; BAGTCD,

BDNF, L-Ascorbic Acid, GDNF, TGFß3, dbcAMP, DAPT. For details see methods section. Quantification

of cell aggregates of a control iPSC line encapsulated in alginate beads of different compositions (Alg,

alginate; Fn, fibronectin) at day 5 of neuronal differentiation, with and without pre-treatment with RI

in StemMACS medium (3 days RI versus no RI) immediately after encapsulation; at encapsulation, RI

was added in both conditions.

36

(B) Control iPSCs at day 5 of neuronal differentiation in alginate beads of different compositions

corresponding to the quantification in C. (C) Proliferating cells form aggregates inside the 3D

matrices, which vary in abundance and shape. The number of aggregates is quantified as cell

aggregates per field of view. Data from one differentiation experiment. Statistical differences were

calculated by one-way ANOVA followed by Tukey's post hoc test to correct for multiple comparisons

(n=3 to 11 fields of view per condition). Data is plotted as ± SEM. **** p ≤ 0.0001.

6.1.2. Number of cell aggregates within alginate of different compositions

To examine which alginate concentration would best support the survival and propagation of the

hiPSCs, we tested alginate concentrations of 1% or 2% w/v. Both concentrations have been indicated

as suitable for neuronal differentiation in previous publications (Li et al. 2011; Wang et al. 2009;

Bozza et al. 2014) as their elastic modulus has been found to be in the range resembling the stiffness

of neural tissue, which seems to have a strong impact on the lineage differentiation of pluripotent

stem cells (Keung et al. 2012; Saha et al. 2008; Engler et al. 2006).

Moreover, as alginate alone is thought not to provide adequate cell adhesion (Rowley et al. 1999), it

was tested whether cell survival and proliferation was enhanced by the addition of the glycoprotein

fibronectin, a prominent ECM component, which is known to be involved in neural development,

pathfinding and regeneration (Perris and Perissinotto 2000; Luckenbill-Edds 1997). This protein is

characterized by the presence of specific domains, which facilitate the interaction with other matrix

molecules as well as with cell surface receptors, most notably integrins (Colognato and Yurchenco

2000; Schwarzbauer and DeSimone 2011).

Because of the low number of cell aggregates in the beads not treated with RI, only the group, which

received the three day pre-treatment with RI was analyzed further. The cells were differentiated for

more than 20 days and images were taken at days 5, 10 and 20 (Fig. 5A). Even though the cell

aggregate number tended to be higher in 1% alginate, we could not observe a significant difference

at any time point in number of aggregates per field between 1% and 2% alginate in this experiment

(Fig. 5B). However, a significantly higher number of aggregates in both 1% and 2% alginate matrices

containing fibronectin was observed compared to alginate alone (Fig. 5A, B). The highest number of

aggregates was obtained for iPSCs encapsulated in 1% alginate with fibronectin (Alg+Fn) at day five

of differentiation (n=35 on average).

6.1.3. Size of cell aggregates within alginate of different compositions

To further investigate whether there was a difference in terms of growth of cell aggregates between

the various alginate compositions, the size of the aggregates was measured. The length of aggregates

37

was used as a proxy to measure their size as their shape was mostly non-spherical. At day 5, no

significant difference was detected between the different compositions. At day 10, the average size

of aggregates was higher in 2% Alg+Fn than in 1% Alg+Fn. This difference became highly significant at

day 20. Additionally, at this time point the average size of aggregates was significantly higher in 2%

Alg+Fn than in 2% alginate alone. Similarly, it was higher in 1% Alg+Fn than in 1% alginate alone (Fig.

5C). The largest aggregates were observed in the condition of 2% Alg+Fn (Fig. 5A, C).

B C

38

FIGURE 5. The number and size of aggregates increase by the addition of fibronectin. (A) Control

iPSCs differentiated into mDA-neurons in alginate beads of different compositions (Alg, alginate; Fn,

fibronectin) following the protocol outlined in Figure 4A. Proliferating cells form aggregates inside

the 3D matrices, which vary in abundance and size. D5, 10 and 20 refer to days after initiation of

differentiation (after 3 day treatment with Rho associated kinase inhibitor (RI) in StemMACS

medium). (B) Quantification of cell aggregates from differentiations shown in A. The number of

aggregates is given in cell aggregates per field of view. (C) Size of cell aggregates quantified in B.

Length of aggregates has been used for non-spherical cell aggregates as indicator of size. Data from

one (2% Alg, 2% Alg+Fn) or two differentiation experiments. Statistical differences were calculated by

one-way ANOVA followed by Tukey's post hoc test to correct for multiple comparisons (B: n=3 to 11

fields of view per condition; C: n=4-128 cell aggregates per condition). Data is plotted as ± SEM. * p ≤

0.05, **** p≤ 0.0001.

6.2. Fibronectin supports cell viability during differentiation

In order to examine the viability of cells within different matrix compositions, we performed a

Live/Dead two-color assay on the intact beads. For this purpose, we used a commercially available Kit

for mammalian cells (Thermo Fisher). The essential components of the Kit are Calcein-AM, marking

living cells with green fluorescence, and the Ethidium Homodimer-1 (EH-1) which shows red

fluorescence when it reaches the nuclei of dead cells passing through disrupted membranes. Cellular

viability was calculated as the green/red ratio. In most cases, the alginate beads remained roughly

spherically shaped and the first cellular aggregates were evaluated after 5 days of differentiation (Fig.

6A).

From day 0 to day 20 of differentiation the number of viable cells in 1% and 2% Alg+Fn was higher

compared to the other conditions in alginate alone. In addition, at day 20, almost all cells

encapsulated in 1% alginate alone resulted dead, and the cell viability within 2% alginate without

fibronectin was clearly decreased compared to the conditions with fibronectin (Fig. 6B).

Furthermore, we investigated the size of the aggregates formed by viable cells. The volume of

aggregates could be determined by 3D reconstruction (Fig. 6A). In line with the results obtained from

bright field microscopy (Fig. 5C), we found that the cells in 2% Alg+Fn formed the biggest aggregates

followed by the ones in 1% Alg+Fn. In both conditions, the aggregate size increased steadily over

time. In contrast, the cell aggregates in 1% and 2% alginate alone did not increase in size during the

examined differentiation period. These data support the crucial role of fibronectin in supporting cell

viability during the differentiation process.

39

B

C

40

FIGURE 6. Viability of iPSCs encapsulated and differentiated in alginate of different compositions.

The LIVE/DEAD assay kit for mammalian cells (Thermo Fisher) was used for the distinction of live and

dead cells. (A) Representative 3D reconstructions of cell aggregates stained with Calcein-AM (live)

and Ethidium Homodimer-1 (dead). Scalebar represents 100 µm. (B) Cell viability over time given as a

percentage of green/red ratio. (C) Average size of the observed cell aggregates given in number of

voxels.

6.3. Comparative RT-qPCR suggests enhanced mDA differentiation

The data collected so far highlighted the key role of fibronectin in supporting the differentiation

performance in terms of number, size and viability. Thus, we decided to further investigate the mDA

neuronal marker gene expression of the neurons cultured in alginate (1% and 2%) with fibronectin

using quantitative real-time PCR (RT-qPCR) and to compare it with the gene expression of neurons

differentiated in parallel using our conventional 2D protocol.

A number of well-defined marker genes were investigated at day 10 and 20 of differentiation. In

Figure 7 the normalized gene expression patterns of the 3D neuronal cultures (1% Alg+Fn and 2%

Alg+Fn) are compared to the gene expression observed in the 2D culture (set to 1).

At day 10, the early neuronal commitment marker Pax6 was expressed up to 10 times more in the 3D

conditions (Fig. 7A). This indicates a stronger neuronal induction under 3D conditions. Our analysis

further showed a stronger expression of FOXA2 at day 10 under 3D compared to 2D conditions (Fig.

7A), which points toward a floorplate derived midbrain identity (Kriks et al. 2011).

The class III member of the beta tubulin family TUJ1 is involved in axon guidance and frequently used

as a neuronal marker (Tischfield et al. 2010). We could detect an increased expression level of TUJ1

for both day 10 and 20 of differentiation under 3D conditions when compared to the 2D culture (Fig.

7A and B).

Interestingly, the expression of tyrosine hydroxylase (TH), the rate limiting enzyme in dopamine

production, was always higher under 3D conditions when compared to the 2D culture. Furthermore,

a noteworthy difference between the alginate compositions was observed for TH expression, which

was stronger for cells encapsulated in 1% Alg+Fn at both time points (Fig. 7A and B). The expression

of the dopamine transporter (DAT), investigated at day 20 of differentiation, was significantly higher

for both 1% and 2% Alg+Fn compared to the neurons generated using the 2D protocol.

41

A

B

FIGURE 7. Comparative gene expression analysis at day 10 (A) and 20 (B) of differentiation

between mDA neurons generated in alginate of different compositions (Alg, alginate; Fn,

fibronectin) and 2D culture (2D). Data are presented as the mean ± standard deviation from

triplicates normalized to the gene expression in 2D culture. Statistical differences were calculated by

two-way ANOVA followed by Tukey post hoc test to correct for multiple comparisons. * p ≤ 0.05, **p

≤ 0.01, **** p≤ 0.0001

Day 20

Day 10

42

In order to check whether the set of primers used for the RT-qPCR experiments gave the correct and

expected amplicon, we loaded the qPCR products on an agarose gel. The sizes of the PCR products

matched the expected amplicon lengths (Fig. 8).

FIGURE 8. Gel electrophoresis of qPCR products amplified in the RT-qPCR experiments.

6.4. Molecular analysis of early mDA neural differentiation

The alginate compositions (1% and 2%) with fibronectin performed consistently better than alginate

alone regarding cell aggregates number and size and cell viability (see Fig. 5 and 6). Furthermore,

gene expression analysis of cells encapsulated in 1% and 2% Alg+Fn showed that these two matrix

compositions resulted in comparable gene expression patterns, except for a higher TH expression in

1% Alg+Fn (Fig. 7). Additionally, the handling of 1% alginate was easier during cell encapsulation

compared to the 2% alginate. For these reasons, we chose to rely on 1% Alg+Fn as biomaterial

composition for the following experiments.

6.4.1. Immunofluorescence staining

To further demonstrate the ability of our matrix to sustain neural differentiation, we decided to

analyze the gene expression of early neuronal and mDA neuron marker genes via

immunofluorescence staining at day 10 and day 20 of differentiation. In order to quantify the signals,

we either counted positively stained nuclei (counts/nuclei) or measured the volume of positive

43

stained cells for each gene as well as the overlapping co-stained regions (voxels/nuclei) and

normalized them to the number of nuclei.

At day 10 of differentiation, no statistically significant difference in the expression of the neuronal

progenitor markers Pax6 and Nestin could be observed between 2D and 3D conditions (Fig. 9A and

B).

FIGURE 9. Immunofluorescence stainings for Pax6 and Nestin. (A) Representative

immunofluorescence images showing the neuronal progenitor markers Pax6 and Nestin at day 10 in

2D and 3D (1% Alg+Fn) cultures. (B) Quantification of Pax6 and Nestin immunofluorescence stainings

generated in 2D and 3D cultures.

The combined expression of the transcription factors FOXA2 and LMX1A is essential for formation of

a ventral midbrain identity (Arenas et al. 2015) and the in vitro generation of a floorplate derived

A

B

44

midbrain cell lineage (Kriks et al. 2011). Therefore, we quantified immunofluorescence stainings of

FOXA2 and LMX1A for neurons generated in 2D as well as 3D cultures at day 10 of differentiation.

Although the difference was not significant, we observed that the percentages of LMX1A+ and

FOXA2+ cells were increased in 1% Alg+Fn compared to cells generated using the 2D protocol.

Additionally, the signal from double positive LMX1A+/FOXA2+ neuronal cells was significantly

increased in our 3D platform compared to the conventional 2D model (Fig. 10A and B).

FIGURE 10. Immunofluorescence stainings for FOXA2 and LMX1A. (A) Representative

immunofluorescence images showing the expression of the transcription factors LMX1A and FOXA2

at day 10 in 2D and 3D (1% Alg+Fn) cultures. (B) Quantification of LMXA1 and FOXA2

immunofluorescence stainings/co-stainings in 2D and 3D cultures. Statistical differences were

calculated by one-way ANOVA followed by Tukey’s post hoc test to correct for multiple comparison*

p ≤ 0.05

A

B

45

To further characterize the extent of differentiation, we analyzed the co-staining of the DA neuron

marker TH together with the pan-neuronal marker TUJ1 and the mDA neuronal maturation marker

GIRK2 at day 20 of differentiation.

We were able to show that the proportion of both TH+ and TUJ1+ cells was significantly higher under

3D condition when compared with the 2D protocol. Furthermore, a more than three times higher

fraction of TH+/TUJ1+ double positive cells was observed under 3D conditions (Fig. 11A and B).

FIGURE 11. Immunofluorescence stainings for TH and TUJ+. (A) Representative immunofluorescence

images showing the co-staining of TH with the neuronal marker TUJ1 at day 20 of differentiation in

2D and 3D (1% Alg+Fn) cultures. (B) Quantification of TH and TUJ1 immunofluorescence stainings/co-

stainings in 2D and 3D cultures. Statistical differences were calculated by one-way ANOVA followed

by Tukey’s post hoc test to correct for multiple comparison * p ≤ 0.05, ***p ≤ 0.001

A

B

46

Next, we determined TH in co-staining with GIRK2. We were able to clearly identify GIRK2+ neurons

even though they were much less numerous than TH+ neurons (Fig. 12A). Also for this condition, we

were able to show that the proportion of both TH+ and GIRK2+ cells was significantly higher in the 3D

model system compared to the 2D system. Additionally, the fraction of TH+/GIRK2+ was again higher

under 3D conditions, although not significantly so (Fig. 12B).

FIGURE 12. Immunofluorescence stainings for TH and GIRK2. (A) Representative

immunofluorescence images showing the co-staining of TH with mDA neuronal maturation marker

GIRK2 at day 20 of differentiation in 2D and 3D (1% Alg+Fn) cultures. (B) Quantification of TH and

GIRK2 immunofluorescence stainings/co-stainings in 2D and 3D cultures. Statistical differences were

calculated by one-way ANOVA followed by Tukey’s post hoc test to correct for multiple comparison *

p ≤ 0.05

A

B

47

6.4.2. Western blot analysis

Western blot analysis was an additional method we used to further characterize the early mDA

neuronal differentiation of neurons in terms of marker gene expression. We found that the

expression levels of the transcription factors FOXA2 and LMX1A were markedly higher in the 3D

model system, compared to 2D differentiated cells at day 20 of differentiation.

In line with our immunofluorescence and RT-qPCR data, we observed a stronger upregulation in the

expression of the neuronal marker TUJ1 and the DA neuron marker TH in the neurons obtained using

the 3D platform at day 20 (Fig. 13).

FIGURE 13. Western blot analysis of whole cell lysates of control iPSC derived mDA neurons at day

20 of differentiation, cultured in 2D and 3D. Relative density values were normalized to the loading

control GAPDH.

48

When we analyzed the same markers at day 35 of differentiation, we could still observe an

upregulation in the transcription factors FOXA2 and LMX1A in the 3D culture system, but not as

strong as at day 20 of differentiation. The same was true also for the other two markers, TUJ1 and

TH. It is also important to note the high variability in the expression of key neuronal markers

observed in the 2D model system, which on the contrary was not present for the neurons generated

in the alginate microcapsules (Fig. 14).

FIGURE 14. Western blot analysis of whole cell lysates of control iPSC derived mDA neurons at day

35 of differentiation, cultured in 2D and 3D. Relative density values were normalized to the loading

control GAPDH.

49

6.5. Electrophysiological analysis of the neuronal cultures

Endogenous DA neurons in the human midbrain display spontaneous rhythmic firing of action

potentials from 2 to 10 Hz (Grace and Bunney 1984; Ramayya et al. 2014). The Ih current is an inward

depolarizing current mediated by the hyperpolarization-activated cyclic nucleotide-gated (HCN)

cation channels. It plays an essential role in regulating neuronal properties, synaptic integration and

plasticity, and synchronous activity in the brain. It is a typical characteristic of mature mesencephalic

DA neurons (Chu and Zhen 2010). Moreover, during neuronal maturation, the expression of Nav

channels increases and the cell becomes capable of generating APs spontaneously or following

depolarization.

In order to functionally assess the mDA neuronal maturation of our 2D and 3D cultures, we

performed electrophysiological recordings at day 30 of differentiation. Individual neurons were

recorded in whole-cell configuration in order to evaluate the presence of spontaneous firing activity

(APs) and the hyperpolarization-activated current (Ih).

At day 30 of differentiation, we observed that 35% of the recorded cells in the 3D neuronal culture

were able to generate spontaneous as well as evoked APs. The average spontaneous firing rate of

our 3D cultured neuronal cells was 2.7 ± 0.97 Hz (mean ± SEM; range 0.43-6.82 Hz, N=6; Tab. 6). This

frequency is in line with the values reported in relevant publications (Grace and Bunney 1984; Chu

and Zhen 2010; Ramayya et al. 2014; Hartfield et al. 2014). On the contrary, we could not record any

APs for the cells generated using our conventional 2D protocol at this stage of the differentiation

(Fig. 15A).

Moreover, 41% of the patched neurons, generated by the 3D system, displayed the voltage-sag

response upon hyperpolarizing current injections, which is a distinctive feature of DA neurons,

indicating the presence of the HCN-mediated currents in these cells (Fig. 15A). Again, none of the 2D

recorded neurons at day 30 was characterized by the presence of HCN-mediated currents.

Additionally, we examined the passive electrophysiological properties of our differentiated mDA

neurons. These properties are inherent to the cell membranes and could modulate neuronal

integration and firing patterns. We have found that the resting membrane potential (mV) was

significantly more negative in the cells of the 3D neuronal culture. Furthermore, membrane

capacitance (pF) was significantly higher in the same cells compared with cells cultured in 2D. We

could not observe a significant difference for the input resistance (MΩ).

50

FIGURE 15. Electrophysiological analyses of 2D and 3D cultures. (A) Representative traces from

whole cell patch clamp recordings of cells in 2D or 3D cultures at day 30. APs, Spontaneous action

potentials were observed only in cells cultured in 3D; Ih, the presence of the HCN-mediated current

was observed upon application of hyperpolarizing current steps at different intensities with -30 pA

increments. No spontaneous or evoked APs and no Ih could be recorded for the 14 recorded cells

cultured in 2D. In contrast, 35% of the 17 examined cells cultured in 3D showed spontaneous as well

as evoked APs, while the presence of Ih was recorded in 41% of these neurons. (B) Quantification of

passive electrophysiological properties of cells in 2D or 3D cultures at day 30. Statistical differences

were calculated by Mann-Whitney test * p ≤ 0.05

The passive and the active electrical properties, exhibited by the cells generated in the 2D and 3D

cultures, are summarized in Table 6.

TABLE 6. Summary of passive and active electrophysiological properties of the recorded neuronal cells generated in the 2D and 3D system.

Number of cells Vm (mV) Rm (M) Cm (pF) Spontaneous APs / Ih Evoked AP Na Currents K Currents

2D 14 -33 ± (-2) 848 ± 127 13 ± 1.4 0% / 0% 0/14 (0%) 86% 100%

3D 17 -41 ± (-2) 705 ± 78 19 ± 1.9 35% / 41% 6/17 (35%) 100% 100%

Values represent mean ± SEM, otherwise the percentage of cells is given. Vm, resting membrane potential; Rm, input resistance; Cm,

membrane capacitance; APs, spontaneous action potentials; Ih, Hyperpolarization-activated mixed cation current.

A

B

51

Overall, these data indicate that the DA neurons generated using the 3D model system reached

earlier a relatively advanced functional level of in vitro maturity when compared to neurons

generated using our conventional 2D protocol. Moreover, 3D neurons showed the two main

electrophysiological characteristics of mesencephalic DA neurons: the pacemaking firing capacity (i.e.

presence of spontaneous APs) and the functional expression of the Ih current.

6.6. Mitochondrial reactive oxygen species (mROS)

Oxidative stress, caused by disequilibrium between generation and detoxification of reactive oxygen

species (ROS), plays an important role in the cellular-damage and -death associated with

neurodegenerative diseases. The generation of ROS correlates with mitochondrial dysfunction, which

is a widely accepted pathogenic mechanism implicated in PD (Lin and Beal 2006; Dias et al. 2013).

Furthermore, increased oxidative stress is a well-known feature of iPSC-derived neurons with

mutations in PD related genes (Imaizumi et al. 2012; LaMarca et al. 2018).

We compared the mDA neurons differentiated with our 3D platform to those differentiated in 2D in

terms of mitochondrial superoxide (mROS) generation, as this phenotype could be the basis of future

experiments. For this purpose, we used the mitochondria targeted, superoxide sensitive, fluorogenic

dye MitoSox Red to detect mROS levels at day 10, 20 and 35 in both our healthy control line and an

iPSC-line carrying an α-synuclein (SNCA) triplication mutation.

Under both 2D and 3D conditions, we consistently observed higher mROS levels, indicated by

MitoSox fluorescence mean intensity, in the α-synuclein triplication mutant compared to the control

cell line. The difference was statistically significant for both 2D and 3D conditions at day 20 and day

35 of differentiation. The aberrant phenotype was observed already at day 10 of differentiation for

both conditions but at this time point it was significant only for the neurons generated using the 3D

alginate platform (Fig. 16).

These findings indicate that our 3D model system can recapitulate one important mitochondrial

phenotype observed in PD patients’-derived neurons and that this phenotype might be detectable

earlier during neuronal differentiation.

52

FIGURE 16. Mitochondrial superoxide levels (mROS) quantified with the fluorogenic dye MitoSox

Red. (A) Representative fluorescence images at different time points. (B) Quantification of MitoSox

intensity. The mean MitoSox intensity ± SEM is given as a measure of mROS levels. mDA neurons

differentiated from a control and an iPSC line carrying an α-synuclein triplication mutation (3x SNCA)

are compared at day 10, 20 and 35. Data are derived from two independent differentiation

experiments. Statistical differences were calculated by one-way ANOVA followed by Tukey's post hoc

test to correct for multiple comparisons * ≤ 0.05, ** p ≤ 0.01, **** p≤ 0.0001

6.7. Molecular analysis of long-term mDA neural differentiation on 3D

platform

PD is thought to be the consequence of damage accumulating progressively in the course of years,

possibly decades, in specific populations of post mitotic neurons. Thus, the finding that ageing is the

most important risk factor for PD is easy to explain. On the other hand, iPSCs represent rejuvenated

cells that lost the hallmarks imposed by ageing. This is bedeviling the development of a successful

iPSC-based cell culture system for the modeling of disease. Several approaches have been developed

Day 10 Day 20 Day 35 A

B

53

in order to deal with this complication. One of them is the long-term culture of differentiated mDA

neurons in order to recapitulate the ageing process in vitro. This is a challenging undertaking

considering that cells, especially neurons, cultured for long periods tend to lose viability.

Therefore, it was important to show that our 3D culture method is suitable also for long-term

maturation of iPSC-derived mDA neurons. For this reason, we kept mDA neurons differentiated both

in 2D or encapsulated in 1% Alg+Fn in culture for more than 200 days. At the end of this period, we

analyzed the expression of mature mDA neuronal and synaptic marker proteins in both groups via

RT-qPCR and immunofluorescence staining performed on cell aggregate sections.

The gene expression of the mature mDA marker proteins DAT, TH and GIRK2, as well as the pan-

neuronal marker TUJ1 was determined via RT-qPCR. We could observe significantly higher expression

levels of TH, GIRK2, and DAT in the neurons generated in the 3D model system in comparison to our

conventional 2D protocol (Fig. 17).

FIGURE 17. Comparative gene expression analysis via RT-qPCR between mDA neurons differentiated

in 1% Alg+Fn (3D) and 2D culture for more than 200 days. Data are presented as the mean ± standard

deviation from technical triplicates normalized to the gene expression in 2D culture. Statistical

differences were calculated by two-way ANOVA followed by Bonferroni post hoc test to correct for

multiple comparison * p ≤ 0.05, **p ≤ 0.01

In Figure 18A, we show that both mDA marker proteins DAT and TH are expressed in cells

encapsulated in alginate. The stained areas are partially overlapping indicating the co-expression of

both proteins. The most intense staining for both proteins was observed on the margins of the

54

sections. This could be due to a higher mDA neuron density overall or specifically in the periphery of

the cell aggregates (Fig. 18A-B).

Furthermore, we showed the formation of multiple synaptic connections in the neurons within the

aggregates, evaluating the expression of the presynaptic marker protein Synapsin 1 (Syn1) as well as

the postsynaptic density protein 95 (PSD95). While the Syn1 signal was located to neurites and

possibly cell bodies, the PSD95 staining resulted in a dotted pattern indicating a more confined

expression of the latter. In order to confirm their neuronal identity we costained the sections for the

neuronal marker microtubule associated protein 2 (MAP2) (Fig. 18B). In a 3D surface reconstruction

of synapse formation, we could illustrate the juxtaposed signals of Syn1 and PSD95 staining, which

might denote the presence of pre- and post-synaptic puncta and multiple synaptic connections (Fig.

18C).

55

FIGURE 18. Molecular analysis of long term mDA neural differentiation via immunofluorescence

staining. (A) Co-staining of TH and DAT. Magnified region shows overlapping staining. Scalebar

represents 100 µm (B) Co-staining for the presynaptic marker protein Synapsin1 (Syn1), the neuronal

marker microtubule associated protein 2 (MAP2) and the postsynaptic density protein 95 (PSD95). (C)

3D surface reconstruction of MAP2/Syn1/PSD95 staining showing formation of synaptic puncta

(white arrows).

A

B

C

56

Overall, it was confirmed that iPSC derived mDA neurons could be generated and sustained for at

least 200 days in our 3D culture system based on alginate as a biomaterial. The observed mDA

neuronal marker expression confirmed the successful maturation of the mDA neurons. The enhanced

expression of DAT, TH and GIRK2 in 3D culture may even indicate a favorable effect to the generation

of the respective cell lineage. However, the results have to be interpreted carefully, especially since

they derive from one experiment and have yet to be replicated.

57

7. DISCUSSION

7.1. Patient-specific neuronal cell models for PD research

The advent of iPSC technology enabled researchers for the first time to generate patient-specific

pluripotent cell lineages carrying virtually any genotype. These can be differentiated to potentially

any cell type to study the pathophysiology of disorders and identify novel therapeutic approaches.

However, the iPSC technology is limited by several factors. iPSC lines can retain epigenetic signatures

of their tissue of origin, and they sometimes do not gain a true pluripotent state (Cherry and Daley

2013). Thus, it is challenging to differentiate iPSCs into homogeneous neuronal populations for

disease modeling. Moreover, variations in the differentiation ability of different but also the same

cell lines have been reported (Choi et al. 2009). For this reason the correction of the disease

genotype via genetic engineering in patient’s iPSC lines is used to generate isogenic control lines

which are regarded as the best control for any disease model (Jungverdorben et al. 2017).

Furthermore, an inherent feature of iPSCs, the apparent reversal of cellular ageing occurring in the

course of reprogramming, poses a major challenge for the modeling of late onset disease. Efforts are

put into the development of techniques allowing the artificial induction of cellular ageing. One

promising approach is the ectopic expression of progeroid gene products, even though it is not clear

how closely normal ageing can be mimicked in such a way (Studer et al. 2015).

Protocols for the generation of mDA neurons have resulted in percentages of TH-positive neurons

ranging from 8% to 80%, implying that the generated cell populations remain heterogeneous (Engel

et al. 2016). One suitable possibility to produce more uniform cell populations may be the expansion

of relatively homogeneous midbrain FP progenitors, which has been reported to increase relative DA

neuron differentiation potential (Fedele et al. 2017). However, other studies reported much higher

DA neuron yields without progenitor expansion (Kriks et al. 2011; Reinhardt et al. 2013). In any case

the remaining heterogeneity gives room for further improvement in terms of DA neuron yield but

also maturity and functionality reached by different differentiation approaches. Thus, the goal of this

thesis was to present a multivariate picture of differentiation outcomes of our 3D culture system in

comparison with a 2D adherent culture system. This included the analysis of neuronal as well as early

and mature mDA neuron marker gene expression and electrophysiological characterization.

7.2. Setup of 3D system by encapsulation of hiPSCs with alginate hydrogels

The precise spatial organization and well timed cooperation and communication of cells allows for

the sustainment of complex living beings. The function of cells is profoundly influenced by their

extracellular microenvironment as well as by the cross-talk between different cell types (Alberts

58

2015). The ECM, in which cells are embedded, is an important part of this microenvironment. It helps

with tissue organization but can also be seen as a cellular extension and active participant in the

regulation of cellular function (Bissell and Barcellos-Hoff 1987). Furthermore, it has been shown that

cell size, volume and geometry constitute major cues altering gene expression (Bao et al. 2017). In

fact, several observations support the idea that cells cultured in 3D systems, where these factors can

be modelled to a certain extent, are more closely resembling their in vivo counterparts than cells

cultured in 2D on conventional rigid plastic or glass surfaces. As an example, the specification of

mesenchymal stem cells seems to be highly sensitive to three-dimensional matrix elasticity (Engler et

al. 2006). Another line of evidence is supported by the enhanced drug resistance of 3D tumor

cultures compared to more vulnerable single cell cultures of the same cells (Miller et al. 1985;

Hoffmann M.R. 1993). Also the restricted cell motility, tissue-like apoptosis and presence of cellular

ECM remodeling, observed under 3D but not 2D conditions, point towards a more in vivo like cell

behavior (Geckil et al. 2010). Furthermore, certain specific cellular phenotypes, such as the saltatory

migration of cortical interneurons, can be observed and studied in 3D cultures of brain assembloids

but not in 2D culture systems (Birey et al. 2017).

In the present thesis, we made use of ionically cross-linked alginate as a matrix for the 3D culture and

differentiation of iPSCs into mDA neurons. Alginate is a polysaccharide derived from brown algae

known for its ability to form hydrogels at physiological conditions enabling easy cell encapsulation

and retrieval. Biological functionality can be introduced to it by chemical modifications, for example

with ECM components or cell-binding peptides (DeVolder and Kong 2012; Andersen et al. 2015). It

has been reported that the stiffness of brain tissues changes significantly throughout developing

stages (Iwashita et al. 2014). Thus, during neuronal differentiation, there might be temporally

differential effects of alginate stiffness and fibronectin. However, cells embedded in a 3D matrix are

known to remodel their extracellular microenvironment, which may mask differential effects of initial

stiffness and matrix modifications at later differentiation stages (Geckil et al. 2010).

Our data show that the encapsulation of iPSCs in 1% and 2% Alg+Fn results in the formation of viable

cell aggregates. The significantly higher number of aggregates in 1% Alg+Fn suggests an increased cell

survival after encapsulation under this condition. The decreasing number of observed cell aggregates

over time may be explained by several factors: growing aggregates may fuse with others and be

counted as one or impede the sight of others underneath; furthermore, we observed that aggregates

in the periphery occasionally grew out of the alginate beads and were therefore not included in the

analysis. Some floating cell aggregates did attach to the bottom of the wells, showing a neuronal

morphology.

59

A form of dissociation-induced apoptosis referred to as anoikis is triggered upon dysplastic growth or

attachment to an inappropriate matrix of cells that normally exert their functions at well-defined

locations in vivo. The resulting poor survival of these cells in vitro represents an obstacle to research

(Taddei et al. 2012). A possibility to decrease the activation of anoikis related pathways is the

pharmacological inhibition of dissociation induced cell death. In this context, the RHO associated

kinase inhibitor Y27632 (RI) was shown to permit the survival of dissociated hESCs (Watanabe et al.

2007) and described as a crucial factor for the sustainment of hESC viability during the first three

days after encapsulation in alginate (Chayosumrit et al. 2010; Kim et al. 2013). We were able to show

the same for iPSCs encapsulated in alginate of different compositions. Very few cell aggregates were

observed after encapsulation without RI treatment under all conditions.

Based on the importance of proper cell-ECM interactions described in the literature (Bozza et al.

2014; Caliari and Burdick 2016; Andersen et al. 2015), we hypothesized that enhancing these

interactions could further support the viability and differentiation of our encapsulated iPSCs.

Therefore, we chose to add the glycoprotein fibronectin, a ubiquitous ECM component frequently

used to coat culture surfaces for diverse cell types, to the alginate matrix and examine its effect on

cell survival and growth. Our data show a highly significant increase in the observed numbers of cell

aggregates in 1% or 2% Alg+Fn. Further optimization of fibronectin concentration and testing of

other ECM components or modifications might lead to an additional increase in cell aggregate

numbers.

To examine cell viability more closely, we made use of a fluorescence staining method for the

discrimination between live and dead cells. This enabled the evaluation of cell viability within

individual aggregates. Overall, our results suggest that the treatment with RI, to inhibit disassociation

related cell-death, is crucial to sustain cell viability after encapsulation. Additionally, the introduction

of cell-ECM interactions by fibronectin further supported cell viability for the examined period.

Regarding the size of cell aggregates we observed minor differences between different conditions at

day 5 and 10. However, a highly significant increase in aggregate size for 1% and especially 2% Alg+Fn

became apparent at day 20 while aggregate size did not further increase in alginate alone. This

suggests a prolonged period of cell proliferation or a possible increase in cell volume and in cell

mobility due to cell-ECM contacts upon the addition of fibronectin.

The size of cellular aggregates, rather than the size of the microcapsules, has been reported as

limiting factor influencing the growth of encapsulated cells due to restricted nutrient and oxygen

diffusion (Huang et al. 2012). The strong increase in aggregate size in 2% Alg+Fn might therefore not

be an advantage as it could finally result in cellular necrosis in the aggregate core. It should be noted

60

that most analyzed aggregates did not have a spherical but rather a spindle- or lens-like shape. Thus,

as we used aggregate length as a proxy for size, the actual diffusion distances from aggregate surface

to aggregate core cannot be deduced from aggregate sizes.

7.3. mDA neuronal differentiation of hiPSCs in 3D alginate culture system

The cellular 3D microenvironment is an important determinant of gene expression, which in turn

influences cellular identity and function (Bissell and Barcellos-Hoff 1987; Engler et al. 2006; McKee

and Chaudhry 2017). Thus, we compared the gene expression of cells differentiated in 1% or 2%

Alg+Fn to that of cells differentiated in 2D culture. RT-qPCR showed that overall the expression of

early neuronal as well as mature mDA-neuron markers was stronger under 3D conditions. We

observed a slight advantage of 1% Alg+Fn as TH expression was highest under this condition at both

day 10 and 20 of differentiation.

The co-expression of the transcription factors LMX1A and FOXA2 is essential for the specification of

mDA neurons (Arenas et al. 2015; Engler et al. 2006). Additionally, it has been proven important for

the generation of high quality mDA neurons in vitro (Kriks et al. 2011). We showed the expression of

both LMX1A and FOXA2 under 2D and 3D conditions at day 10 via immunofluorescence staining.

Furthermore, we demonstrated, via western blot analysis, that LMX1A and FOXA2 expression levels

were significantly increased under 3D conditions at days 20 and 35 with respect to 2D culture.

The intermediate filament protein Nestin and the transcription factor Pax6 are frequently used as

markers for neural progenitors (Chambers et al. 2009; Reinhardt et al. 2013; Kim et al. 2013; Adil et

al. 2017). We were able to confirm the expression of both Nestin and Pax6 via immunofluorescence

staining at day 10 of our 3D differentiation in 1% Alg+Fn as well as in 2D culture. The numbers of

positive cells for both markers were not significantly different under 2D and 3D conditions. However,

RT-qPCR data indicated a significantly higher Pax6 expression level in 3D culture at day 10.

TH, the rate limiting enzyme in dopamine production, is crucial for mDA neuronal function and

commonly used as a marker for mature DA neurons. Thus, in order to assess the maturation state of

our cells at day 20, we analyzed the expression of TH by immunofluorescence co-staining with

different neuronal markers. We were able to show that TH and TUJ1 expression was significantly and

consistently stronger in cells cultured in 1% Alg+Fn than in 2D cultured cells, which was confirmed via

RT-qPCR at day 20 as well as at day 35. Additionally, GIRK2 expression could be shown in our

neuronal cultures at day 20 via immunofluorescence staining, with an increased expression in our 3D

model system. RT-qPCR showed increased DAT expression under 3D conditions but no significant

difference for GIRK2 between 2D and 3D culture at day 20. These findings indicate that the

61

differentiation process was already well defined at day 20. The overall low expression of GIRK2 and

DAT may be due to the relatively early time point of analysis. The consistent observation of double

positively stained cells is nevertheless supporting the notion that our cells underwent differentiation.

In order to further increase the reliability of our 3D culture system, it might be helpful to monitor the

stiffness of the alginate matrix via rheological measurement of the elastic moduli, as reported in the

literature, to ensure the consistency of this parameter found to be important for early neurogenic

differentiation (Saha et al. 2008; Keung et al. 2012).

Another realm for optimization could be the timing of exposition to differentiation factors. The

differentiation protocol we used has been developed for a 2D culture system. It might well be that

the dynamics of cellular responsiveness to these factors are different in our 3D culture. It is

conceivable that the 3D culture results in a more rapid differentiation, which is supported by the

higher expression levels of neural progenitor and mDA neuron markers at day 10 in our 3D culture

system. This would be in line with the findings of a hESC based study by Kim et al. (Kim et al. 2013). In

order to further explore these data, gene expression patterns should be analyzed at a higher

temporal resolution.

The in vitro generation of mature iPSC derived neuronal cells is challenging. iPSCs and their

derivatives usually do not reach a fully mature state due to the reversal of cellular ageing during

reprogramming. This poses a major obstacle for the modeling of late onset diseases. Many strategies

have been developed in order to face this problem (Studer et al. 2015; Jungverdorben et al. 2017).

The most basic may be the prolongation of culture periods up to several months (Ho et al. 2015),

even though post-mitotic neurons are difficult to keep in culture over long periods of time as they

are vulnerable and prone to detachment from culture surfaces.

We have shown that our iPSC derived neurons can be maintained in 3D culture for more than 200

days. Immunofluorescence analysis confirmed that these neurons co-expressed the mDA neuron

markers TH and DAT. Additionally, RT-qPCR experiments indicated an increased expression of the

mDA neuron markers TH, DAT and GIRK2 under 3D conditions. Moreover, using a 3D surface

reconstruction of synapse formation, we were able to reveal the presence of pre- and postsynaptic

puncta and multiple synaptic connections.

The problem of cell detachment upon long-term culture can be circumvented with our 3D culture

system but slow experiment turnaround remains an issue. Furthermore, we observed slow but

persistent growth of cell aggregates in this system, indicating still the presence of some proliferating,

immature cells and probably leading to insufficient nutrient and oxygen transport within aggregates.

This may alter cellular states, impede proper maturation and result in cellular necrosis.

62

The expression analysis of specific neuronal-, mature mDA- and synaptic marker- proteins is an

important way to assess the success of differentiation but the broader goal has to be the generation

of fully functional excitable mDA neurons. Thus, it is crucial to characterize the differentiated cells in

terms of electrophysiological activity. We did so by performing electrophysiological recordings in

whole-cell patch-clamp configuration in order to evaluate the presence of spontaneous APs and Ih

current. At a relatively early stage of differentiation (day 30), these characteristics were exclusively

observed in cells generated in 3D culture. The measured average firing rate of the recorded cells was

within the range reported for mDA neurons in the literature (Grace and Bunney 1984; Chu and Zhen

2010; Ramayya et al. 2014; Hartfield et al. 2014). This indicates that the 3D culture system better

supports the generation of mature functional mDA neurons, which is further emphasized by the

recorded passive electrophysiological properties.

With this thesis we aimed at providing a contribution to the improvement of a cellular model of PD

which can be used to advance the understanding of the mechanisms underlying the disease.

Oxidative stress plays a central role in the degeneration of DA neurons in PD. Several mechanisms

and sources of mROS production, including mitochondrial dysfunction, have been identified (Dias et

al. 2013; Imaizumi et al. 2012). In order to examine the utility of our 3D model system to recapitulate

known disease-related phenotypes, we chose to investigate mROS production in a 3D differentiated

α-synuclein triplication mutant cell line in comparison to a control cell line. Indeed, we were able to

demonstrate that the mROS levels were significantly higher in the mutant cell line under both 2D and

3D conditions at days 20 and 35. Intriguingly, at day 10 this difference was already significant under

3D but not under 2D conditions. This suggests that our 3D model might hold the potential for the

earlier detection of relevant disease phenotypes compared to 2D culture.

Overall, our data suggest that our alginate based 3D culture system is suitable for the directed in

vitro differentiation of mDA neurons. Several lines of evidence indicate an enhancing effect of 3D

compared to 2D culture in terms of the differentiation and maturation of the generated neurons,

although additional analyses need to be performed in order to get further insights on the quality of

the obtained neurons.

63

8. SUMMARY

In this thesis we investigated the potential of an alginate based 3D culture system to improve the

performance of an established 2D protocol for the generation of mDA neurons from iPSCs. In the first

part, we identified a biomaterial combination that supported the short and long-term survival of

iPSCs embedded in alginate beads. In the second part, we focused on the comparison of early and

late marker gene expression patterns observed for neurons cultured in 2D and 3D conditions.

Additionally, we set out to probe the functional maturity of mDA neurons differentiated in our 3D

culture system by electrophysiological characterization and demonstrated that our 3D model was

able to recapitulate a known mitochondrial dysfunctional phenotype. Furthermore, we showed that

iPSC-derived mDA neurons differentiated in 3D can be kept in culture for more than 200 days.

The advent of cellular reprogramming and iPSC technology in combination with directed

differentiation protocols made it possible to generate patient specific disease affected cell types in

vitro. However, current culture systems for neuronal differentiation show a variable reproducibility

and yield neurons with low degrees of functional maturation. In this work, we tested different

scaffold biomaterial compositions and identified 1% alginate modified with fibronectin as the most

suitable scaffold for differentiation of iPSCs in 3D culture. Furthermore, we showed that the

treatment with RI is essential for the initial survival of iPSCs embedded in this biomaterial. We

demonstrated by RT-qPCR and western blot analyses that the expression levels of neuronal

commitment as well as mature mDA neuron markers, at day 10, 20 and 35, were higher in 3D with

respect to 2D cultures. By electrophysiological analysis at day 30, we found that a significant

proportion of our mDA neurons differentiated in 3D were able to generate spontaneous action

potentials. In contrast, no spontaneous action potentials could be recorded at day 30 for mDA

neurons differentiated in 2D. Additionally, we reported that mROS levels were significantly different

between mDA neurons differentiated from a control iPSC line and those differentiated from an α-

synuclein triplication mutant iPSC line in both 2D and 3D cultures at days 20 and 35. Interestingly, at

day 10 the difference was only significant for 3D cultured mDA neurons. We kept differentiated mDA

neurons in 3D culture for more than 200 days and showed that the mDA neuron marker expression

levels were increased compared to 2D culture at this time point. Further, we confirmed the presence

of synaptic connections formed by these neurons by immunofluorescence stainings.

These data indicate that our alginate based 3D culture system represents a tool for the reliable and

rapid derivation of more mature and functional mDA neurons from iPSCs with respect to

conventional 2D culture systems. However, the system has to be tested on other, especially patient-

derived, iPSC lines, and its potential for the analysis of disease-specific cellular phenotypes has to be

further validated.

64

9. PERSPECTIVES

Most in vitro neurodegenerative disease modeling studies are so far based on conventional 2D cell

culture systems. However, 3D cell culture models are thought to better mimic the cell growth

encountered in vivo and support the expression of tissue-specific genes and proteins (Geckil et al.

2010). Therefore, we consider it important to investigate the potential of 3D culture to support the

generation of mature and functional iPSC-derived mDA neurons for PD modeling. In order to

emphasize and validate the results produced in this thesis, it will be crucial to expand and enrich the

collected data using additional technologies to further compare and characterize our 2D and 3D

neuronal cultures. A higher temporal resolution of gene expression, especially in the early phase of

differentiation, could help to optimize the differentiation schedule. Additionally, the gene expression

analysis is currently being extended to later time points in order to further characterize the mDA

neuron maturation. Moreover, RNAseq analysis is employed to generate an even more complete and

comprehensive picture of the gene expression patterns.

We have shown that cells differentiated with our 3D culture model were able to recapitulate the

increased mROS phenotype of an α-synuclein triplication mutant cell line. It will be interesting to

determine whether other major phenotypic differences between mDA neurons derived from patient

and control iPSCs differentiated with our 3D culture platform can be detected. Interesting PD-related

phenotypes include mitochondrial dysfunction characterized by altered electron transport complex I

activity, mitochondrial integrity, mitochondrial membrane potential, mitophagy, and respiration. In

this context, it would be ideal to introduce the use of isogenic iPSC controls. This would enable the

identification of more subtle phenotypic effects owed almost only to genetic differences.

Further optimization could be the functionalization of alginate with other extracellular matrix

components and factors. In order to increase the reproducibility, the use of an automated bead

formation and of 3D bioprinters might be ideal. In the long run, the use of novel rapidly evolving

technologies including microfabricated cell culture devices and microfluidics approaches will likely

provide the best way to create in vivo like microenvironments.

65

10. REFERENCES

Alberts, Bruce (2015): Molecular biology of the cell. Sixth edition. New York NY: Garland Science

Taylor and Francis Group.

Andersen, Therese; Auk-Emblem, Pia; Dornish, Michael (2015): 3D Cell Culture in Alginate Hydrogels.

In Microarrays (Basel, Switzerland) 4 (2), pp. 133–161. DOI: 10.3390/microarrays4020133.

Arenas, Ernest; Denham, Mark; Villaescusa, J. Carlos (2015): How to make a midbrain dopaminergic

neuron. In Development (Cambridge, England) 142 (11), pp. 1918–1936. DOI: 10.1242/dev.097394.

Ascherio, Alberto; Schwarzschild, Michael A. (2016): The epidemiology of Parkinson's disease. Risk

factors and prevention. In The Lancet Neurology 15 (12), pp. 1257–1272. DOI: 10.1016/S1474-

4422(16)30230-7.

Baharvand, Hossein; Hashemi, Seyed M.; Kazemi Ashtiani, Saeid; Farrokhi, Ali (2006): Differentiation

of human embryonic stem cells into hepatocytes in 2D and 3D culture systems in vitro. In The

International journal of developmental biology 50 (7), pp. 645–652. DOI: 10.1387/ijdb.052072hb.

Bao, Min; Xie, Jing; Piruska, Aigars; Huck, Wilhelm T. S. (2017): 3D microniches reveal the importance

of cell size and shape. In Nature communications 8 (1), p. 1962. DOI: 10.1038/s41467-017-02163-2.

Barone, Paolo; Antonini, Angelo; Colosimo, Carlo; Marconi, Roberto; Morgante, Letterio; Avarello,

Tania P. et al. (2009): The PRIAMO study. A multicenter assessment of nonmotor symptoms and their

impact on quality of life in Parkinson's disease. In Movement disorders : official journal of the

Movement Disorder Society 24 (11), pp. 1641–1649. DOI: 10.1002/mds.22643.

Beyer, Katrin; Domingo-Sàbat, Montserrat; Ariza, Aurelio (2009): Molecular pathology of Lewy body

diseases. In International journal of molecular sciences 10 (3), pp. 724–745. DOI:

10.3390/ijms10030724.

Birey, Fikri; Andersen, Jimena; Makinson, Christopher D.; Islam, Saiful; Wei, Wu; Huber, Nina et al.

(2017): Assembly of functionally integrated human forebrain spheroids. In Nature 545 (7652),

pp. 54–59. DOI: 10.1038/nature22330.

Bissell, Mina J.; Barcellos-Hoff, Maria Helen (1987): The influence of extracellular matrix on gene

expression_Is structure the message. In Journal of cell science. Supplement.

Björklund, Anders; Hökfelt, Tomas (1983-): Handbook of chemical neuroanatomy. Amsterdam, New

York: Elsevier.

Boergermann, J. H.; Kopf, J.; Yu, P. B.; Knaus, P. (2010): Dorsomorphin and LDN-193189 inhibit BMP-

mediated Smad, p38 and Akt signalling in C2C12 cells. In The international journal of biochemistry &

cell biology 42 (11), pp. 1802–1807. DOI: 10.1016/j.biocel.2010.07.018.

Bonaventure, J.; Kadhom, N.; Cohen-Solal, L.; Ng, K. H.; Bourguignon, J.; Lasselin, C.; Freisinger, P.

(1994): Reexpression of cartilage-specific genes by dedifferentiated human articular chondrocytes

cultured in alginate beads. In Experimental cell research 212 (1), pp. 97–104. DOI:

10.1006/excr.1994.1123.

Bozza, Angela; Coates, Emily E.; Incitti, Tania; Ferlin, Kimberly M.; Messina, Andrea; Menna,

Elisabetta et al. (2014): Neural differentiation of pluripotent cells in 3D alginate-based cultures. In

Biomaterials 35 (16), pp. 4636–4645. DOI: 10.1016/j.biomaterials.2014.02.039.

66

Braak, Heiko; Del Tredici, Kelly (2017): Neuropathological Staging of Brain Pathology in Sporadic

Parkinson's disease. Separating the Wheat from the Chaff. In Journal of Parkinson's disease 7 (s1),

S71-S85. DOI: 10.3233/JPD-179001.

Byers, Blake; Cord, Branden; Nguyen, Ha Nam; Schüle, Birgitt; Fenno, Lief; Lee, Patrick C. et al. (2011):

SNCA triplication Parkinson's patient's iPSC-derived DA neurons accumulate α-synuclein and are

susceptible to oxidative stress. In PloS one 6 (11), e26159. DOI: 10.1371/journal.pone.0026159.

Caliari, Steven R.; Burdick, Jason A. (2016): A practical guide to hydrogels for cell culture. In Nature

methods 13 (5), pp. 405–414. DOI: 10.1038/nmeth.3839.

Centeno, Eduarda G. Z.; Cimarosti, Helena; Bithell, Angela (2018): 2D versus 3D human induced

pluripotent stem cell-derived cultures for neurodegenerative disease modelling. In Molecular

neurodegeneration 13 (1), p. 27. DOI: 10.1186/s13024-018-0258-4.

Chayosumrit, Methichit; Tuch, Bernard; Sidhu, Kuldip (2010): Alginate microcapsule for propagation

and directed differentiation of hESCs to definitive endoderm. In Biomaterials 31 (3), pp. 505–514.

DOI: 10.1016/j.biomaterials.2009.09.071.

Chu, Hong-yuan; Zhen, Xuechu (2010): Hyperpolarization-activated, cyclic nucleotide-gated (HCN)

channels in the regulation of midbrain dopamine systems. In Acta pharmacologica Sinica 31 (9),

pp. 1036–1043. DOI: 10.1038/aps.2010.105.

Colognato, H.; Yurchenco, P. D. (2000): Form and function. The laminin family of heterotrimers. In

Developmental dynamics : an official publication of the American Association of Anatomists 218 (2),

pp. 213–234. DOI: 10.1002/(SICI)1097-0177(200006)218:2<213::AID-DVDY1>3.0.CO;2-R.

Corti, Olga; Lesage, Suzanne; Brice, Alexis (2011): What genetics tells us about the causes and

mechanisms of Parkinson's disease. In Physiological reviews 91 (4), pp. 1161–1218. DOI:

10.1152/physrev.00022.2010.

Dahlstroem, A.; Fuxe, K. (1964): Evidence for the Existence of Monoamine-containing Neurons in the

Central Nervous System. I. Demonstration of Monoamines in the Cell Bodies of brain stem neurons.

In Acta physiologica Scandinavica. Supplementum, SUPPL 232:1-55.

Damier, P.; Hirsch, E. C.; Agid, Y.; Graybiel, A. M. (1999): The substantia nigra of the human brain. II.

Patterns of loss of dopamine-containing neurons in Parkinson's disease. In Brain : a journal of

neurology 122 (Pt 8), pp. 1437–1448.

Deng, Hao; Wang, Peng; Jankovic, Joseph (2018): The genetics of Parkinson disease. In Ageing

research reviews 42, pp. 72–85. DOI: 10.1016/j.arr.2017.12.007.

DeVolder, Ross; Kong, Hyun-Joon (2012): Hydrogels for in vivo-like three-dimensional cellular studies.

In Wiley interdisciplinary reviews. Systems biology and medicine 4 (4), pp. 351–365. DOI:

10.1002/wsbm.1174.

Dias, Vera; Junn, Eunsung; Mouradian, M. Maral (2013): The role of oxidative stress in Parkinson's

disease. In Journal of Parkinson's disease 3 (4), pp. 461–491. DOI: 10.3233/JPD-130230.

Dutta, Ranjna C.; Dutta, Aroop K. (2018): 3D cell culture. An introductory textbook / Ranjna C. Dutta,

Aroop K. Dutta. 1st. Singapore: Pan Stanford Publishing.

Engler, Adam J.; Sen, Shamik; Sweeney, H. Lee; Discher, Dennis E. (2006): Matrix elasticity directs

stem cell lineage specification. In Cell 126 (4), pp. 677–689. DOI: 10.1016/j.cell.2006.06.044.

Farrer, Matthew James (2006): Genetics of Parkinson disease. Paradigm shifts and future prospects.

In Nature reviews. Genetics 7 (4), pp. 306–318. DOI: 10.1038/nrg1831.

67

Fennema, Eelco; Rivron, Nicolas; Rouwkema, Jeroen; van Blitterswijk, Clemens; Boer, Jan de (2013):

Spheroid culture as a tool for creating 3D complex tissues. In Trends in biotechnology 31 (2), pp. 108–

115. DOI: 10.1016/j.tibtech.2012.12.003.

Frantz, Christian; Stewart, Kathleen M.; Weaver, Valerie M. (2010): The extracellular matrix at a

glance. In Journal of cell science 123 (Pt 24), pp. 4195–4200. DOI: 10.1242/jcs.023820.

Fuchs, J.; Nilsson, C.; Kachergus, J.; Munz, M.; Larsson, E-M; Schüle, B. et al. (2007): Phenotypic

variation in a large Swedish pedigree due to SNCA duplication and triplication. In Neurology 68 (12),

pp. 916–922. DOI: 10.1212/01.wnl.0000254458.17630.c5.

Geckil, Hikmet; Xu, Feng; Zhang, Xiaohui; Moon, SangJun; Demirci, Utkan (2010): Engineering

hydrogels as extracellular matrix mimics. In Nanomedicine (London, England) 5 (3), pp. 469–484. DOI:

10.2217/nnm.10.12.

Godugu, Chandraiah; Patel, Apurva R.; Desai, Utkarsh; Andey, Terrick; Sams, Alexandria; Singh,

Mandip (2013): AlgiMatrix™ based 3D cell culture system as an in-vitro tumor model for anticancer

studies. In PloS one 8 (1), e53708. DOI: 10.1371/journal.pone.0053708.

Grace, Anthony A.; Bunney, Benjamin S. (1984): The control of firing pattern in nigral dopamine

neurons: Single spike firing. In Journal of neuroscience 4 (11), pp. 2866–2876.

Grayson, Warren L.; Ma, Teng; Bunnell, Bruce (2004): Human mesenchymal stem cells tissue

development in 3D PET matrices. In Biotechnology progress 20 (3), pp. 905–912. DOI:

10.1021/bp034296z.

Gribova, Varvara; Crouzier, Thomas; Picart, Catherine (2011): A material's point of view on recent

developments of polymeric biomaterials. Control of mechanical and biochemical properties. In

Journal of materials chemistry 21 (38), pp. 14354–14366. DOI: 10.1039/C1JM11372K.

Hartfield, Elizabeth M.; Yamasaki-Mann, Michiko; Ribeiro Fernandes, Hugo J.; Vowles, Jane; James,

William S.; Cowley, Sally A.; Wade-Martins, Richard (2014): Physiological characterisation of human

iPS-derived dopaminergic neurons. In PloS one 9 (2), e87388. DOI: 10.1371/journal.pone.0087388.

Hegarty, Shane V.; Sullivan, Aideen M.; O'Keeffe, Gerard W. (2013): Midbrain dopaminergic neurons.

A review of the molecular circuitry that regulates their development. In Developmental biology 379

(2), pp. 123–138. DOI: 10.1016/j.ydbio.2013.04.014.

Hegarty, Shane V.; Sullivan, Aideen M.; O'Keeffe, Gerard W. (2014): Roles for the TGFβ superfamily in

the development and survival of midbrain dopaminergic neurons. In Molecular neurobiology 50 (2),

pp. 559–573. DOI: 10.1007/s12035-014-8639-3.

Hu, Chenxia; Li, Lanjuan (2016): Current reprogramming systems in regenerative medicine. From

somatic cells to induced pluripotent stem cells. In Regenerative medicine 11 (1), pp. 105–132. DOI:

10.2217/rme.15.79.

Huang, Xiaobo; Zhang, Xiangyu; Wang, Xiaoguang; Wang, Chan; Tang, Bin (2012): Microenvironment

of alginate-based microcapsules for cell culture and tissue engineering. In Journal of bioscience and

bioengineering 114 (1), pp. 1–8. DOI: 10.1016/j.jbiosc.2012.02.024.

Huh, Dongeun; Hamilton, Geraldine A.; Ingber, Donald E. (2011): From 3D cell culture to organs-on-

chips. In Trends in cell biology 21 (12), pp. 745–754. DOI: 10.1016/j.tcb.2011.09.005.

Imaizumi, Yoichi; Okada, Yohei; Akamatsu, Wado; Koike, Masato; Kuzumaki, Naoko; Hayakawa,

Hideki et al. (2012): Mitochondrial dysfunction associated with increased oxidative stress and α-

68

synuclein accumulation in PARK2 iPSC-derived neurons and postmortem brain tissue. In Molecular

brain 5, p. 35. DOI: 10.1186/1756-6606-5-35.

Jackson, E. L.; Lu, H. (2016): Three-dimensional models for studying development and disease.

Moving on from organisms to organs-on-a-chip and organoids. In Integrative biology : quantitative

biosciences from nano to macro 8 (6), pp. 672–683. DOI: 10.1039/c6ib00039h.

Joksimovic, Milan; Yun, Beth A.; Kittappa, Raja; Anderegg, Angela M.; Chang, Wendy W.; Taketo,

Makoto M. et al. (2009): Wnt antagonism of Shh facilitates midbrain floor plate neurogenesis. In

Nature neuroscience 12 (2), pp. 125–131. DOI: 10.1038/nn.2243.

Josephine C. Adams; Fiona M. Watt (1993): Regulation of development and differentiation by the

extracellular matrix. In Development (Cambridge, England) (117), pp. 1183–1198.

Jovanovic, Vukasin M.; Salti, Ahmad; Tilleman, Hadas; Zega, Ksenija; Jukic, Marin M.; Zou, Hongyan et

al. (2018): BMP/SMAD Pathway Promotes Neurogenesis of Midbrain Dopaminergic Neurons In Vivo

and in Human Induced Pluripotent and Neural Stem Cells. In The Journal of neuroscience : the official

journal of the Society for Neuroscience 38 (7), pp. 1662–1676. DOI: 10.1523/JNEUROSCI.1540-

17.2018.

Jungverdorben, Johannes; Till, Andreas; Brüstle, Oliver (2017): Induced pluripotent stem cell-based

modeling of neurodegenerative diseases. A focus on autophagy. In Journal of molecular medicine

(Berlin, Germany) 95 (7), pp. 705–718. DOI: 10.1007/s00109-017-1533-5.

Keung, Albert J.; Asuri, Prashanth; Kumar, Sanjay; Schaffer, David V. (2012): Soft microenvironments

promote the early neurogenic differentiation but not self-renewal of human pluripotent stem cells.

In Integrative biology : quantitative biosciences from nano to macro 4 (9), pp. 1049–1058. DOI:

10.1039/c2ib20083j.

Kim, Jaemin; Sachdev, Perminder; Sidhu, Kuldip (2013): Alginate microcapsule as a 3D platform for

the efficient differentiation of human embryonic stem cells to dopamine neurons. In Stem cell

research 11 (3), pp. 978–989. DOI: 10.1016/j.scr.2013.06.005.

Kriks, Sonja; Shim, Jae-Won; Piao, Jinghua; Ganat, Yosif M.; Wakeman, Dustin R.; Xie, Zhong et al.

(2011): Dopamine neurons derived from human ES cells efficiently engraft in animal models of

Parkinson's disease. In Nature 480 (7378), pp. 547–551. DOI: 10.1038/nature10648.

Kuo, C. K.; Ma, P. X. (2001): Ionically crosslinked alginate hydrogels as scaffolds for tissue engineering.

Part 1. Structure, gelation rate and mechanical properties. In Biomaterials 22 (6), pp. 511–521.

LaMarca, Elizabeth A.; Powell, Samuel K.; Akbarian, Schahram; Brennand, Kristen J. (2018): Modeling

Neuropsychiatric and Neurodegenerative Diseases With Induced Pluripotent Stem Cells. In Frontiers

in pediatrics 6, p. 82. DOI: 10.3389/fped.2018.00082.

Lang, A. E.; Lozano, A. M. (1998): Parkinson's disease. First of two parts. In The New England journal

of medicine 339 (15), pp. 1044–1053. DOI: 10.1056/NEJM199810083391506.

Lee, Kuen Yong; Mooney, David J. (2012): Alginate. Properties and biomedical applications. In

Progress in polymer science 37 (1), pp. 106–126. DOI: 10.1016/j.progpolymsci.2011.06.003.

Li, Lulu; Davidovich, Alexander E.; Schloss, Jennifer M.; Chippada, Uday; Schloss, Rene R.; Langrana,

Noshir A.; Yarmush, Martin L. (2011): Neural lineage differentiation of embryonic stem cells within

alginate microbeads. In Biomaterials 32 (20), pp. 4489–4497. DOI:

10.1016/j.biomaterials.2011.03.019.

69

Lim, F.; Sun, A. M. (1980): Microencapsulated islets as bioartificial endocrine pancreas. In Science

(New York, N.Y.) 210 (4472), pp. 908–910.

Lin, Michael T.; Beal, M. Flint (2006): Mitochondrial dysfunction and oxidative stress in

neurodegenerative diseases. In Nature 443 (7113), pp. 787–795. DOI: 10.1038/nature05292.

Luckenbill-Edds, L. (1997): Laminin and the mechanism of neuronal outgrowth. In Brain research.

Brain research reviews 23 (1-2), pp. 1–27.

Ma, Xiaojie; Kong, Linghao; Zhu, Saiyong (2017): Reprogramming cell fates by small molecules. In

Protein & cell 8 (5), pp. 328–348. DOI: 10.1007/s13238-016-0362-6.

Manyam, B. V. (1990): Paralysis agitans and levodopa in "Ayurveda". Ancient Indian medical treatise.

In Movement disorders : official journal of the Movement Disorder Society 5 (1), pp. 47–48. DOI:

10.1002/mds.870050112.

McKee, Christina; Chaudhry, G. Rasul (2017): Advances and challenges in stem cell culture. In Colloids

and surfaces. B, Biointerfaces 159, pp. 62–77. DOI: 10.1016/j.colsurfb.2017.07.051.

McWilliams, Thomas G.; Muqit, Miratul Mk (2017): PINK1 and Parkin. Emerging themes in

mitochondrial homeostasis. In Current opinion in cell biology 45, pp. 83–91. DOI:

10.1016/j.ceb.2017.03.013.

Michel, P. P.; Agid, Y. (1996): Chronic activation of the cyclic AMP signaling pathway promotes

development and long-term survival of mesencephalic dopaminergic neurons. In Journal of

neurochemistry 67 (4), pp. 1633–1642. DOI: 10.1046/j.1471-4159.1996.67041633.x.

Miller, B. E.; Miller, F. R.; Heppner, G. H. (1985): Factors affecting growth and drug sensitivity of

mouse mammary tumor lines in collagen gel cultures. In Cancer research 45 (9), pp. 4200–4205.

Moustafa, Ahmed A.; Chakravarthy, Srinivasa; Phillips, Joseph R.; Gupta, Ankur; Keri, Szabolcs;

Polner, Bertalan et al. (2016): Motor symptoms in Parkinson's disease. A unified framework. In

Neuroscience and biobehavioral reviews 68, pp. 727–740. DOI: 10.1016/j.neubiorev.2016.07.010.

Narendra, Derek; Tanaka, Atsushi; Suen, Der-Fen; Youle, Richard J. (2008): Parkin is recruited

selectively to impaired mitochondria and promotes their autophagy. In The Journal of cell biology 183

(5), pp. 795–803. DOI: 10.1083/jcb.200809125.

Olderøy, Magnus Ø.; Lilledahl, Magnus B.; Beckwith, Marianne Sandvold; Melvik, Jan Egil; Reinholt,

Finn; Sikorski, Pawel; Brinchmann, Jan E. (2014): Biochemical and structural characterization of

neocartilage formed by mesenchymal stem cells in alginate hydrogels. In PloS one 9 (3), e91662. DOI:

10.1371/journal.pone.0091662.

Pașca, Sergiu P. (2018): The rise of three-dimensional human brain cultures. In Nature 553 (7689),

pp. 437–445. DOI: 10.1038/nature25032.

Pawar, Siddhesh N.; Edgar, Kevin J. (2012): Alginate derivatization. A review of chemistry, properties

and applications. In Biomaterials 33 (11), pp. 3279–3305. DOI: 10.1016/j.biomaterials.2012.01.007.

Perris, Roberto; Perissinotto, Daniela (2000): Role of the extracellular matrix during neural crest cell

migration. In Mechanisms of Development 95 (1-2), pp. 3–21. DOI: 10.1016/S0925-4773(00)00365-8.

Placzek, Marysia; Briscoe, James (2005): The floor plate. Multiple cells, multiple signals. In Nature

reviews. Neuroscience 6 (3), pp. 230–240. DOI: 10.1038/nrn1628.

Playne, Rebecca; Connor, Bronwen (2017): Understanding Parkinson's Disease through the Use of

Cell Reprogramming. In Stem cell reviews 13 (2), pp. 151–169. DOI: 10.1007/s12015-017-9717-5.

70

Poewe, Werner (2017): Parkinson's disease and the quest for preclinical diagnosis. An interview with

Professor Werner Poewe. In Neurodegenerative disease management 7 (5), pp. 273–277. DOI:

10.2217/nmt-2017-0027.

Poewe, Werner; Seppi, Klaus; Tanner, Caroline M.; Halliday, Glenda M.; Brundin, Patrik; Volkmann,

Jens et al. (2017): Parkinson disease. In Nature reviews. Disease primers 3, p. 17013. DOI:

10.1038/nrdp.2017.13.

Ramayya, Ashwin G.; Zaghloul, Kareem A.; Weidemann, Christoph T.; Baltuch, Gordon H.; Kahana,

Michael J. (2014): Electrophysiological evidence for functionally distinct neuronal populations in the

human substantia nigra. In Frontiers in human neuroscience 8, p. 655. DOI:

10.3389/fnhum.2014.00655.

Reinhardt, Peter; Glatza, Michael; Hemmer, Kathrin; Tsytsyura, Yaroslav; Thiel, Cora S.; Höing,

Susanne et al. (2013): Derivation and expansion using only small molecules of human neural

progenitors for neurodegenerative disease modeling. In PloS one 8 (3), e59252. DOI:

10.1371/journal.pone.0059252.

Rowlands, Andrew S.; George, Peter A.; Cooper-White, Justin J. (2008): Directing osteogenic and

myogenic differentiation of MSCs. Interplay of stiffness and adhesive ligand presentation. In

American journal of physiology. Cell physiology 295 (4), C1037-44. DOI: 10.1152/ajpcell.67.2008.

Rowley, J. A.; Madlambayan, G.; Mooney, D. J. (1999): Alginate hydrogels as synthetic extracellular

matrix materials. In Biomaterials 20 (1), pp. 45–53.

Saha, Krishanu; Keung, Albert J.; Irwin, Elizabeth F.; Li, Yang; Little, Lauren; Schaffer, David V.; Healy,

Kevin E. (2008): Substrate modulus directs neural stem cell behavior. In Biophysical journal 95 (9),

pp. 4426–4438. DOI: 10.1529/biophysj.108.132217.

Sakai, Shinji; Ono, Tsutomu; Ijima, Hiroyuki; Kawakami, Koei (2004): Behavior of enclosed sol- and

gel-alginates in vivo. In Biochemical Engineering Journal 22 (1), pp. 19–24. DOI:

10.1016/j.bej.2004.07.010.

Sanes, Dan H.; Reh, Thomas A.; Harris, William Anthony (2012): Development of the nervous system.

3. ed. Amsterdam [u.a.], Amsterdam [u.a.]: Elsevier, Academic Press.

Sapir, Yulia; Kryukov, Olga; Cohen, Smadar (2011): Integration of multiple cell-matrix interactions into

alginate scaffolds for promoting cardiac tissue regeneration. In Biomaterials 32 (7), pp. 1838–1847.

DOI: 10.1016/j.biomaterials.2010.11.008.

Schmidt, John J.; Rowley, Jon; Kong, Hyun Joon (2008): Hydrogels used for cell-based drug delivery. In

Journal of biomedical materials research. Part A 87 (4), pp. 1113–1122. DOI: 10.1002/jbm.a.32287.

Schwarzbauer, Jean E.; DeSimone, Douglas W. (2011): Fibronectins, their fibrillogenesis, and in vivo

functions. In Cold Spring Harbor perspectives in biology 3 (7). DOI: 10.1101/cshperspect.a005041.

Shu, Xiao Zheng; Ahmad, Shama; Liu, Yanchun; Prestwich, Glenn D. (2006): Synthesis and evaluation

of injectable, in situ crosslinkable synthetic extracellular matrices for tissue engineering. In Journal of

biomedical materials research. Part A 79 (4), pp. 902–912. DOI: 10.1002/jbm.a.30831.

Sinha, Surajit; Chen, James K. (2006): Purmorphamine activates the Hedgehog pathway by targeting

Smoothened. In Nature chemical biology 2 (1), pp. 29–30. DOI: 10.1038/nchembio753.

Soon-Shiong, P.; Feldman, E.; Nelson, R.; Heintz, R.; Yao, Q.; Yao, Z. et al. (1993): Long-term reversal

of diabetes by the injection of immunoprotected islets. In Proceedings of the National Academy of

Sciences of the United States of America 90 (12), pp. 5843–5847.

71

Sutherland, Robert M.; McCredie, John A.; Inch, W. Rodger (1971): Growth of Multicell Spheroids in

Tissue Culture as a Model of Nodular Carcinomas<xref ref-type="fn" rid="FN2">2</xref>. In JNCI:

Journal of the National Cancer Institute. DOI: 10.1093/jnci/46.1.113.

Takahashi, Kazutoshi; Yamanaka, Shinya (2006): Induction of pluripotent stem cells from mouse

embryonic and adult fibroblast cultures by defined factors. In Cell 126 (4), pp. 663–676. DOI:

10.1016/j.cell.2006.07.024.

Tischfield, Max A.; Baris, Hagit N.; Wu, Chen; Rudolph, Guenther; van Maldergem, Lionel; He, Wei et

al. (2010): Human TUBB3 mutations perturb microtubule dynamics, kinesin interactions, and axon

guidance. In Cell 140 (1), pp. 74–87. DOI: 10.1016/j.cell.2009.12.011.

Trinh, Joanne; Farrer, Matt (2013): Advances in the genetics of Parkinson disease. In Nature reviews.

Neurology 9 (8), pp. 445–454. DOI: 10.1038/nrneurol.2013.132.

Volpicelli, Floriana; Consales, Claudia; Caiazzo, Massimiliano; Colucci-D'Amato, Luca; Perrone-

Capano, Carla; Di Porzio, Umberto (2004): Enhancement of dopaminergic differentiation in

proliferating midbrain neuroblasts by sonic hedgehog and ascorbic acid. In Neural plasticity 11 (1-2),

pp. 45–57. DOI: 10.1155/NP.2004.45.

Wakabayashi, Koichi; Tanji, Kunikazu; Odagiri, Saori; Miki, Yasuo; Mori, Fumiaki; Takahashi, Hitoshi

(2013): The Lewy body in Parkinson's disease and related neurodegenerative disorders. In Molecular

neurobiology 47 (2), pp. 495–508. DOI: 10.1007/s12035-012-8280-y.

Wang, Nan; Adams, Gary; Buttery, Lee; Falcone, Franco H.; Stolnik, Snow (2009): Alginate

encapsulation technology supports embryonic stem cells differentiation into insulin-producing cells.

In Journal of biotechnology 144 (4), pp. 304–312. DOI: 10.1016/j.jbiotec.2009.08.008.

Wang, Yongzhong; Kim, Ung-Jin; Blasioli, Dominick J.; Kim, Hyeon-Joo; Kaplan, David L. (2005): In

vitro cartilage tissue engineering with 3D porous aqueous-derived silk scaffolds and mesenchymal

stem cells. In Biomaterials 26 (34), pp. 7082–7094. DOI: 10.1016/j.biomaterials.2005.05.022.

Wilson, John T.; Chaikof, Elliot L. (2008): Challenges and emerging technologies in the

immunoisolation of cells and tissues. In Advanced drug delivery reviews 60 (2), pp. 124–145. DOI:

10.1016/j.addr.2007.08.034.

Zanon, Alessandra; Kalvakuri, Sreehari; Rakovic, Aleksandar; Foco, Luisa; Guida, Marianna;

Schwienbacher, Christine et al. (2017): SLP-2 interacts with Parkin in mitochondria and prevents

mitochondrial dysfunction in Parkin-deficient human iPSC-derived neurons and Drosophila. In Human

molecular genetics 26 (13), pp. 2412–2425. DOI: 10.1093/hmg/ddx132.