University of Southern California Dissertations and Theses

Cultured neuronal cells derived from the olfactory neuroepithelium growing in three dimensions as a model system for schizophrenia

(USC Thesis Other)

Cultured neuronal cells derived from the olfactory neuroepithelium growing in three dimensions as a model system for schizophrenia

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

CUL TURED NEURONAL CELLS DERIVED FROM THE
OLFACTORY NEUROEPITHELIUM GROWING IN
THREE DIMENSIONS AS A MODEL SYSTEM FOR
SCHIZOPHRENIA

Abdurrahman Wisam Muhtaseb

A thesis submitted to the faculty of the Graduate School at the
University of Southern California
Keck School of Medicine
in partial fulfillment of the requirements for the degree of
Master of Science
in the department of
Biochemistry and Molecular Biology

August, 2016

ii

© 2016
Abdurrahman Wisam Muhtaseb
ALL RIGHTS RESERVED
iii

ABSTRACT
Schizophrenia is a devastating neurological disorder. Multiple lines of evidence
support the hypothesis that abnormal neurodevelopment is involved in the etiology of
schizophrenia. Abnormality in synaptic connectivity and/or function has also been found in
the brains of individuals with schizophrenia.
Our lab has previously developed 250 lines of neural cells derived from the olfactory
neuroepithelium of schizophrenic and control subjects. Cells growing in two-dimensional
cultures were characterized by RNA-Seq as neural progenitor cells. Most of the genes that
are differentially expressed comparing cells from schizophrenia and control are related to
neurodevelopment. Genes involved in synaptogenesis are either not expressed or expressed
at very low levels.
Here, we provide preliminary data on cultured neuronal cells derived from the
olfactory neuroepithelium (known as CNON) and maintained in conditions promoting cell
differentiation and growth in a three-dimensional fashion. We noticed dramatic increases in
the expression of some synaptic machinery genes. We, also, found that these 3D cell cultures
consistently are organized into reproducible morphological structures.
We suggest that 3D CNON cultures may serve as a model system for studying altered
neurodevelopment and neuronal connectivity in schizophrenia, allowing the identification
of genes related to abnormal structure and function of the schizophrenic brain.

iv

To my parents, who have been doing more than they could to help me pursue my dreams
to my grandmother, who keeps me in her prayers
to my uncles, without whose support I could not have come so far
to my brothers and sisters, who always showed me how proud of me they are
to my friends and family members who believed in me and supported me
to my love, who made it harder at times and more beautiful and meaningful at all times
and to my three plants, Rosie, Chloe, and Zaina, that are hanging there in my room…

…Thank you

v

ACKNOWLEDGEMENT
I would like to send my deepest appreciation to Dr. Zoltan Tokes, who helped me
come to USC, then helped me find the right lab, and staid supportive till my last day at USC.
My sincere gratitude to Dr. Robert Chow and Dr. Oleg Evgrafov, my mentors, for
letting me join their collaborative project, and for their patience and guidance, and for
everything I have learned from both of them.
I, also, want to thank Dr. James Knowles and his lab staff for welcoming me into
their lab and helping me in my project. Edder, Valeria, Hugo, Sonia from Dr. Chow’s lab,
and Adrian taught me some of the techniques I used in the project and helped with some.
Hugo, Chris W. (my lab friend), and Joe helped me in RNA preparation and sequencing.
Madison from Dr. Chow’s lab worked with me on the project and helped generously with
some experiments. Jennifer and Chris A. helped me in running the RNASeq pipeline and in
performing statistical analysis.
I extend my appreciation to Dr. Marcello Coba and his lab for helping me with
expertise and antibodies for synaptic proteins, and to Dr. Ruchi Bajpai, my thesis
committee chair, for being patient and cooperative with me.

vi

T ABLE OF CONTENTS
LIST OF TABLES…...........................................................................................................................ix
LIST OF FIGURES………………………………………………………………….…………….........x
ABBREVIATIONS....….……………………………………………………………………………...xi
INTRODUCTION………………………………………………………………………….……..........1
Background…........................................................................................................................................1
Animal models...……………………………………………………………………………………………...2
Human cell-based models………………………………………………………………………………..3
3D cell cultures……………………………………………………………………………………………….4
CNON – NPCs from the olfactory neuroepithelium…………………………………………….6
Hypothesis……………………………………………………………………………………………………..8
Methodology…………………………………………………………………………………………………..8
MATERIALS AND METHODS………………………………………………………….10
Samples…………………………….…………………………….…………………………….……………...10
CNON cell culture…………………………….…………………………….…………………………......10
vii

Cryopreservation…………………………….…………………………………………………………....11
Defrosting cells…………………………….…………………………….…………………………...........12
CNON cell passaging…………………………….…………………………….………………………….12
“CNON Sandwich” 3D cultures…………………………….…………………………….……………13
Differentiation medium…………………………….…………………………….……………………..14
Light microscopy and imaging…………………………….…………………………….……………15
RNA purification…………………………….…………………………….…………………………........15
RT-qPCR…………………………….…………………………….…………………………........................16
cDNA synthesis…………………………….…………………………….…………………………....16
Quantitative PCR…………………………….…………………………….…………………………17
Fluorescence NanoDrop…………………………….…………………………….……………………17
aRNA amplification…………………………….…………………………….…………………………...18
cDNA synthesis with Illumina TruSeq
®
…………………………….…………………………….20
Library quantification, cluster generation, and sequencing……………………………...21
Statistical analysis…………………………….…………………………….………………………….....22
Calcium imaging…………………………….…………………………….………………………….........22
RESULTS…………………………….………………………………………………………………………23
CNON cells in 2D…………………………….…………………………….………………………….........23
Morphological Characterization of the “CNON Sandwich” 3D cultures………………24
Network-like structures in CNON Sandwiches…………………………….…………………..24
Arborization and cell-cell communication…………………………….………………………...28
Functional characterization…………………………….…………………..…………………………34
viii

RNA purification and quantitation…………………………….……………………………………34
RNASeq…...…………………………….…………………..…………………………….…………………...35
Data analysis…………………………….…………………..…………………………….………………...36
Differential gene expression…………………………….…………………………………………….38
DISCUSSION…………………………………………………………..………………………………....44
CONCLUSION……………….………………………………………………...……………………......47
REFERENCES…………………………………………………………………………………………..49
ix

LIST OF T ABLES
Table 1:.………………………………………………………………………………………………………………………..15
Table 2:..……………………………………………………………………………………………………………………….35

x

LIST OF FIGURES
Figure 1: ……………………………………………………………………………………………………………..............23
Figure 2: ……………………………………………………………………………………………………………..............24
Figure 3: ……………………………………………………………………………………………………………..............25
Figure 4: ……………………………………………………………………………………………………………..............26
Figure 5: ……………………………………………………………………………………………………………..............27
Figure 6: ……………………………………………………………………………………………………………..............27
Figure 7: ……………………………………………………………………………………………………………..............28
Figure 8: ……………………………………………………………………………………………………………..............29
Figure 9: ……………………………………………………………………………………………………………..............30
Figure 10: ……………………………………………………………………………………………………………............31
Figure 11: ……………………………………………………………………………………………………………............32
Figure 12: ……………………………………………………………………………………………………………............33
Figure 13: ……………………………………………………………………………………………………………............34
Figure 14: ……………………………………………………………………………………………………………............36
Figure 15: ……………………………………………………………………………………………………………............37
Figure 16: ……………………………………………………………………………………………………………............39
Figure 17: ……………………………………………………………………………………………………………............40
Figure 18: ……………………………………………………………………………………………………………............41
Figure 19: ……………………………………………………………………………………………………………............43

xi

ABBREVIATIONS
3D: Three-dimensional
ECM: Extracellular matrix
CNON: Cultured neuronal cells derived for the olfactory neuroepithelium
CNS: Central nervous system
hESCs: Human embryonic stem cells
hiPSCs: Human induced pluripotent stem cells
ICC: Immunocytochemistry
iPSCs: Induced pluripotent stem cells
MBM: Matrigel basal membrane
NSCs: Neural stem cells
NPCs: Neural progenitor cells
ON: Olfactory neuroepithelium
PSCs: Pluripotent stem cells
SCZ: Schizophrenia
FBS: Fetal Bovine Serum
xii

DPBS: Dulbecco's phosphate-buffered saline
DMEM/F12: Dulbecco's Modified Eagle Medium/Nutrient Mixture F-12 (Ham)
UHR: Universal Human Reference RNA
PCR: Polymerase chain reaction
RT-PCR: Reverse transcriptase polymerase chain reaction
qPCR: Quantitative polymerase chain reaction
PCA: Principal component analysis
1

INTRODUCTION
Background
In a study published in 1987, E. Fuller Torrey wrote: “As tragic a disease as schizophrenia is,
it is, also, one of the greatest intellectual challenges of contemporary medicine”
(1, 2)
.
Schizophrenia (SCZ) is a devastating neurological disorder in the central nervous system
(CNS) that affects approximately 1% of the population and is considered the most common
functional psychosis around the world. It produces a lifetime of disabilities and emotional
distress for affected individuals and their families. It becomes clinically apparent in late
adolescence or early adulthood with an average that is 5 years earlier in males than in
females
(3-5)
. It is characterized by psychosis, apathy, social withdrawal, and cognitive
impairment. It is also considered one of the top causes of long-term disabilities
(6)
.
It is now clear that the etiology of SCZ involves a combination of genetic and environmental
factors
(5)
. Even though the characteristic symptoms arise late in adolescence and adulthood,
lines of evidence agree that it’s a neurodevelopmental disorder
(3, 5, and 7)
.
Early studies on the pathophysiology of SCZ were done on postmortem brains. However,
collecting and interpreting data from these brains was limited and complicated due to the
confounding effects of chronic treatment, as well as possible further changes in brain
structure after the manifestation of symptoms
(4)
. These studies show little about the
predisposition or initiation of the disease
(7)
. With more advanced imaging techniques, the
effects of SCZ on the brain can be readily detected
(6).
However, the molecular pathology
behind this disease remains unclear.
2

Studying patterns of gene expression is, probably, an adequate approach to decipher
causative aberrations within specific brain cells. Alterations in patterns of gene expression
lead to disturbance in the developmental process, which leads ultimately to the clinical
symptoms
(5)
. Because we cannot study gene expression in developing brain in SCZ, we must
instead design a model system that mimics the developing human brain.

Animal models
Due to the complexity of the disorder and ethical limitations to its exploration in humans,
researchers chose to use animal models to study SCZ. Throughout the years, a few animal
models have been developed to study certain aspects of the disease
(8)
. These models were
designed to study the hypotheses proposed to explain the molecular mechanisms that result
in the signs and symptoms of SCZ. For example, some of these models target specific
neurotransmitter systems believed to be involved in SCZ, such as GABA and dopamine, or
relevant brain regions that could be involved in the etiology of the disease such as the
prefrontal cortex or the limbic system
(8, 9)
.
There are several disadvantages and challenges that come with these animal models. The
main concern is whether less cognitively developed animals can capture and reproduce such
a sophisticated and heterogeneous disorder that is generally perceived as a cognitive
dysfunction. Heterogeneity in etiology, symptomology and the variability in individual
response to treatment makes it difficult to build a comprehensive model system
(10)
.
Consequently, the current animal models are more suitable to study specific aspects of SCZ
in isolation rather than serving as a complete equivalent of the human disorder
(8)
.
3

Another difficulty of studying animal models is the inability to accurately reproduce
schizophrenic symptoms seen in affected humans
(8)
.

Human cell-based models
The need for more adequate models lead scientists to look for feasible and reliable human-
based models for the study of neurological disorders like SCZ. There are limitations to
postmortem samples, and ethical concerns that encompass obtaining primary cells from the
brains of adults affected with SCZ. To overcome these challenges, some model systems used
engineered human cell lines. However, such cell lines have a different physiology than brain
cells
(11)
.
In order to study cells that have characteristics similar to those in the brain, models have
been designed using stem cells
(11)
. These, also, helped transcend ethical and technical
difficulties associated with primary brain cells. The challenge is still, though, how to obtain
stem cells from individuals, and how to differentiate them and set them up in vitro such that
they behave similar to their differentiated counterparts in vivo.
Human embryonic stem cells (hESCs) were the first pluripotent stem cells (PSCs) to be
isolated from the blastocyst
(12)
and have been isolated during preimplantation genetic
diagnosis ever since
(11)
. However, the supply of these cells is risky and short due to the
ethical regulations and limited numbers of accessible specimens, and embryonic cells do not
manifest the behaviors of late onset diseases such as SCZ. This requires special and
painstaking protocols to grow them while maintaining their pluripotency, and then
differentiate them into neuronal cells.
4

To utilize the great potential of stem cells for modeling human development and disease on
a larger scale, a paradigm shift in the last ten years allowed scientists to use induced
pluripotent stem cells instead of adult or embryonic stem cells
(11)
. Human induced
pluripotent stem cells (hiPSCs) are differentiated somatic cells converted into PSCs by
reprograming their genetic expression
(13)
. These cells can be generated from virtually any
individual using a variety of differentiated cells such as fibroblasts and keratinocytes
(14, 15)
.
hiPSCs became a source for neural stem cells (NSCs) and neural progenitor cells (NPCs). Even
though human NSCs should have the self-renewal, undifferentiated state like other stem
cells, NSCs derived from hiPSCs still have challenges
(13)
. These NSCs grown from hiPSCs will,
also, have NPCs and differentiated cells of limited types of neurons that usually lack
myelination
(7, 11)
, and those which aren’t differentiated yet will terminally differentiate with
passaging
(11, 16)
.
Scientists are now turning to NPCs as a resource for neural cells for their high replicative
rate, rapidity in initiating differentiation and their relatively easy maintenance and
manipulation, as opposed to NSCs
(7)
. Moreover, NPCs can be derived from patients with
disease of late onset such as SCZ
(11)
.

3D Cell cultures
Efforts were made to find the right cell types to model the development of the human CNS,
but proper conditions for the growth and differentiation of these cells are also crucial.
Culturing iPSCs and NPCs in 2D has led to the generation of a variety of neural cell types
(13)
.
5

However, Monolayer cultures had many limitations that pushed scientists to try culture cells
in 3D
(17)
.
Contrary to the 3D neuronal and synaptic growth within the extracellular matrix (ECM) in
the brain tissue, 2D cultures fail to recapitulate the brain cytoarchitecture of neuronal
network formation and arborization
(17)
. Two-dimensional cultures also show immature
differentiation and cell death due to constrained cell morphology and inability to form
complex synaptic interactions
(18)
.
Even though some NSC/ NPC-derived cells show spontaneous electrical activity in 2D
(19, 20)
,
they are not as active as when they are cultured in 3D
(18)
, and the electrophysiology does
not resemble that of 3D networks of active neurons
(17)
. Moreover, neuronal cell cultures in
2D do not show a large heterogeneity of cell types in comparison to organotypic 3D cultures
(21)
and don’t have the capacity for layer formation
(22, 23)
.
3D cultures have shown significant improvement in recapitulating the physiology of
neuronal cells in vivo with higher cell survival and more electrically active cells
(20)
. These 3D
cultures were found to show a formation of neuronal networks, more differentiation and
heterogeneity with myelination and synaptogenesis, cell-cell and cell-ECM interaction,
functional synaptic connectivity
(17, 18, 20, 13)
, and layer formation
(22)
. The ratio of
differentiation from NSCs/NPCs in 3D could reach 70% compared to 14% in 2D
(24)
. This
expands to other non-neuronal cells such as microglia
(20)
.
Furthermore, 3D cultures were able to capture the abnormal development in cells taken
from patients with neurodevelopmental disorders
(25, 23)
, which could be of great benefit for
studying SCZ.
6

Despite the advancing development in neural 3D culture protocols, they still come with
challenges and limitation. A major challenge is the reduction in consistency in 3D as
compared with 2D due to the technical difficulties in maintaining the integrity of the cultures
(18, 20, and 23)
. Some protocols
(18)
found a way around by differentiating NPCs as 2D cultures
before culturing them in 3D to better control their fate. Another limitation these protocols
face is inefficient passive migration of nutrient supply to the 3D culture cores that is not
comparable to the blood vessel network supply in brain tissues
(26)
.
Furthermore, neural 3D cultures do not grow in an environment similar to the embryonic
environment, which means they lack the important non-neural embryonic tissue cross-talk
(21)
. This increases the stochasticity in their growth and the reliance on the medium contents
and their supply to each cell in the 3D culture. Therefore, the phenotypic outcomes of any
abnormalities have to be pronounced enough in comparison to controls to be statistically
significant
(21)
.

CNON – NPCs from the olfactory neuroepithelium
As previously mentioned, NPCs were found to be a good and easier alternative for NSCs for
modeling neurological diseases like SCZ. Instead of going through the long protocols of
transforming somatic cells to collect NPCs, researchers found a shortcut by using the
olfactory neuroepithelium as a source for their neural cell cultures
(27, 28)
.
The olfactory neuroepithelium (ON) contains receptor neurons that project directly to the
olfactory bulb in the ventral forebrain
(29)
. ON cells are capable of continuous neurogenesis
(30)
. ON biopsies are an accessible source of PSCs that can be obtained postmortem or from
7

living patients relatively easily with a very high success rate in harvesting and culturing cells
from them
(27, 30)
.
Cultured neuronal cells derived from the olfactory neuroepithelium (CNONs) are cultured
from ON as the name indicates. They express neuronal markers at high levels and show NPC
characteristics
(28, 30)
. CNON cells have the potential of serving a model system for studying
the neurodevelopmental theory of neuropsychiatric disorders like SCZ
(31 - 36)
.
Our lab has previously obtained 250 nasal biopsies from SCZ patients and controls and
successfully generated cell lines of clear neuronal lineage that retain mitotic activity. The
phenotype and expression pattern of these lines correspond to that of NPCs
(28)
.
Receptor neurons in ON are neuronal cells that are part of the CNS and, arguably, the only
part of CNS in direct contact with the environment
(37)
.
Interestingly, several studies have shown that a subset of SCZ patients have an abnormal
sense of smell. Although some argue that the findings are still inconclusive
(38)
, several
studies propose that early developmental abnormalities in the structure, growth, or function
of ON would be indicative of some neurodevelopmental abnormalities such as SCZ
(29, 38-40)
.
Some even go further to indicate that, in SCZ, impairments in ON can be evident and
diagnostically specific in individuals who are symptom-free but genetically at risk, as
opposed to other major psychiatric conditions with olfactory impairments; and can be
predictive of high-risk individuals who will develop SCZ later in life.
Data from our lab as well as from other studies
(28, 40)
indicate that CNON cells from SCZ
patients show different characteristics in cell culture compared to controls. Expression
8

profiles of SCZ patient-derived CNON cells show a differential expression of many genes
compared to controls.

Hypothesis
CNON cells have significant advantages for studying SCZ over other human cell-based
models. ON biopsies can be obtained from living patients as source of NPCs and other neural
cells. These cells show consistency in gene expression between samples. Culturing and
harvesting CNONs is relatively easy compared to other protocols that involve developing
hiPSCs by de-differentiation of somatic cells followed by differentiation into neural lineage.
Growing CNON cells in 3D instead of 2D may better recapitulate neuronal development in
the brain in vitro. Specifically, we may expect that the ability to form neuronal network in 3D
cultures makes it easier to identify gene expression differences between SCZ and control
cultures relevant to axon guiding, migration, differentiation, brain connectivity and, possibly,
synaptogenesis.

Methodology
We grew the CNON cells in 3D conditions. We seeded the cells in droplets of Matrigel basal
membrane (MBM) allowing them to migrate, connect to each other and form networks
within the MBM scaffold.
We used bright-field light microscopy to see the morphology of these cells and networks. We
purified RNA from 3D cultures and studied their expression profiles using RNA-Seq. We
analyzed differential gene expression between 2D- and 3D-grown cell cultures using
9

DESeq2, and did statistical data analysis, like correlation, principal component analysis
(PCA), and hierarchical clustering, using Partek
®
Genomics Suite (GS).
For functional characterization and electrical activity of the 3D culture cells, we used
synthetic calcium indicator dyes.
10

MATERIALS AND METHODS
Samples
The CNON samples used in this study were taken from a previous NIH-funded study
conducted in the lab
(30, 28)
. Nasal biopsies were taken from patients diagnosed with SCZ and
a control group who had no history of psychiatric disorders. Neither the SCZ patients nor the
controls had any active allergies or history of sinus disease.

CNON cell culture
The technique was published by Evgrafov et al. in 2011
(28)
. Tissue pieces from ON were put
in 60 mm tissue culture dishes coated with 25 µl of 50% MBM (BD Bioscience, USA, cat. no.
356234/ 356234) reconstituted in Coon's medium (1:2) (Sigma-Aldrich, USA)
(41, 42)
, and
then tissue pieces were each covered by 25 µl of 100% MBM. After incubation at 37°C for at
least 20 minutes to allow MBM to polymerize, the immobilized pieces were covered with 5
ml of medium 4506
(36)
, which is based on Coon's medium. Medium is supplemented with
6% fetal bovine serum (FBS) (Invitrogen, US origin, qualified, 26140079), 5 µg/ml human
transferrin (Sigma-Aldrich), 1 µg/ml insulin (Sigma-Aldrich), 10 nmol/l hydrocortisone
(Sigma-Aldrich), 2.5 ng/ml sodium selenite (Sigma-Aldrich), 40 pg/ml thyroxine (Sigma-
Aldrich), 1% antibiotic–antimycotic (Invitrogen), 150 µg/ml bovine hypothalamus extract
(Sigma-Aldrich), and 50 µg/ml endothelial cell growth supplement (Millipore, USA).
Within 1–4 weeks of culturing, CNON cells were observed to grow out of the embedded
pieces of tissue. Outgrown colonies of cells with a neuronal phenotype were then physically
11

isolated using cloning cylinders, and released by digestion using Dispase (BD Bioscience,
USA, cat. no. 354235).
Cells collected from inside the cloning cylinders were washed twice with Coon's medium,
plated onto 35 mm tissue culture dish coated with 50% MBM (incubated at 37°C for 20-35
minutes to polymerize), and supplied with medium 4506.
After these cells grew to confluence they were passaged to 60 mm tissue culture dishes
coated with 50% MBM, grown to confluence again, and passaged into three 60 mm dishes
coated in 50 MBM supplied with medium 4506.

Cryopreservation
When Cells reached confluence, they were harvested from two 60 mm plates of the same cell
line. Medium was removed from plates. Cells were harvested from plates by adding dispase
into each plate and incubating them at least 30 mins (and no more than an hour). Cells were,
then, collected in 15 ml conical tubes. Dispase was diluted with Coon’s medium up to a total
of 10 ml then mixed. Cell suspension was centrifuge 7 minutes at 2000 rpm, supernatant was
decanted. Freezing medium was made of 75% medium 4506, 20% FBS, and 5% DMSO
(Sigma-Aldrich, USA cat. no. D2650-5X10ML), and 3 ml was added onto the cell pellet in the
conical tube. Cell pellet was suspended in the freezing medium, and 1 ml of cell suspension
was stored in a cryogenic tube making 3 of them for each plate of two for each cell line (total
of 6).
12

Cryogenic tubes were placed in a special freezing chamber (Nalgene Cryo-Freezing
Container) containing Isopropyl Alcohol, frozen at -80 overnight, and then stored in liquid
nitrogen.

Defrosting cells
To use samples stored in cryopreservation, cryogenic tubes were taken out of liquid nitrogen
and put in a 37°C water bath for 5 mins. Once thawed, cells were transferred into a 15 mL
conical tube, and 2 mL of pre-warmed medium 4506 were slowly added to them (while
shaking the), followed by 2ml of pre-warmed Coon’s medium added the same way. Cells were
mixed and centrifuged at 2000rpm for 7 mins, medium was decanted, cells were re-
suspended in medium 4506 and plated on two 60 mm petri dishes coated with 50% MBM
(pre-incubated at 37°C to polymerize MBM).
Next day, medium is changed, and cells are incubated until confluence for the next passage
or seeding of 3D cultures.

CNON cell passaging
When the cell cultured were 85-100% confluent, plates were taken out and medium was
removed. 1 ml of Sterile DPBS 1X was added to the 60 mm plate for washing followed by
complete aspiration. This step was repeated one more time. Cells were then treated with 1ml
trypsin-EDTA, (0.05% and 0.02% respectively) (USC Norris Comprehensive Cancer center
Bioreagent & Cell Culture Core, Los Angeles, CA, USA) per 60mm dish, and plate was
incubated at 37°C for up to 5 minutes.
13

Plates were tapped several times before and after incubation to distribute the trypsin-EDTA
and detach cells. Once cells were detached, trypsin-EDTA was deactivated by adding 1
volume of medium 4506 (contains 6% FBS). Cell suspension was then transferred to a 15 ml
conical tube and centrifuged at 2000 rpm for 5 minutes. Supernatant was discarded and
pellet was re-suspended in Coon’s medium for a second wash then centrifuged again, and
supernatant was discarded.
For passaging, pellet was re-suspended in 2 ml medium 4506 and mixed well. 500 µl of the
homogeneous cell suspension was transferred to a 60 mm Petri dish pre-coated with 25 µl
of 50% MBM (Corning, USA, product #: #356234). This is a 1:4 dilution of the cells. The new
cell cultures were, then, incubated in a 5% CO2 incubator at 37°C. For 3D cultures, the pellet
is re-suspended in 100 µl for a higher density of cells.

“CNON Sandwich” 3D cultures
The protocol was based on the paper published by Lancaster and Knoblich
(21)
, but several
and significant modifications have been made.
Cells growing confluently were harvested from 60 mm plates, and were suspended in 100 µl
medium 4506. 1 volume of cell suspension was mixed with 4 volumes of liquefied 100%
MBM (on wet ice, liquid at 0-4°C) and vortexed. Using p20 pipettes, the mix was then
distributed in 5 µl droplets on dimples made with a Parafilm M
®
membrane (sterilized with
70% ethanol) on an empty p200 pipette tip rack to hold the droplets. Rack was then covered
and droplets were incubated at 37°C for 10 minutes to polymerize. Cells were kept in
humidity by adding droplets of medium to empty dimples.
14

Each semi-solid core had uniformly distributed cells within the MBM scaffold. droplets were
coated with 15 µl of liquefied 100% MBM added on top of each core droplet and let surround
the majority of the core’s surface by cohesion forces. The coated droplets were then returned
to incubator for 15 minutes under the same conditions so the outer layer polymerizes
together with the core. Cultures should not stay without medium for more than 30 minutes.
After the second incubation, sandwiches were dislodged from the parafilm dimples by
carefully detaching the parafilm using forceps, and popping the parafilm dimples that
contain 3D cultures outward (without stretching the parafilm). With gentle pipetting with
medium around the sandwich cultures, each CNON sandwich was carefully pushed into a
well of a 24-well plate. 800 µl of medium 4506 was added to each sandwich-containing well.
The culture plate was incubated at 37°C and 5% CO2 for a week without changing the media.

Differentiation medium
After a week of incubating the CNON sandwiches in medium 4506, cultures were put into a
differentiation medium
(21)
(see Table 1), and plate was put onto a shaker inside the
incubator.
When CNON sandwiches were moved to differentiation media, 2% FBS (gibco, US origin, cat.
no. 26140-079) was added to medium. Media were changed every 3-4 days with FBS
concentration being reduced stepwise to 1%, 0.5% and, finally, 0.1% at each media change.
After keeping cultures on 0.1% FBS for 3-4 days, media was removed and CNON sandwiches
were put in serum-free differentiation media that was changed every 3-4 days. Media was
prepared every 2 weeks.
15

Ingredients Quantities
DMEM/ F12 (1:1) (1X) (gibco
®
, USA, REF. 11330-032) 50 ml
Neurobasal
®
Medium (gibco
®
, USA, REF. 21103-049) 50 ml
GlutaMAX
TM
supplement (100X) (gibco
®
, USA, REF. 35050-061) 1000 µl
Penicillin-Streptomycin (5000 units/ml & 5000 mcg/ml respectively)
A
1000 µl
B27 + vitamin A supplement (50X) (gibco
®
, USA, REF. 17504-044) 1000 µl
N-2 supplement (100X) (gibco
®
, USA, REF. 17502-048) 500 µl
MEM Non-essential amino acids (100X) (gibco
®
, USA, REF. 11140-050) 500 µl
≥98% pure 2-mercaptoethanol (Bio-Rad, USA, cat. no. 161-0710) (1:100)
B
35 µl
Insulin solution human (Sigma-Aldrich, USA, cat. no. I9278-5ML) 25 µl

Light microscopy and imaging
Bright field images were taken with the EVOS
TM
FL Cell Imaging System (Life Technologies).
2-photon microscopy was used to take higher resolution images. For cells in 3D cultures.
Image stacks were 3D rendered using ImageJ software.

RNA purification
When cells had shown good growth and viability throughout the CNON sandwich, RNA was
isolated from CNON sandwiches by placing these sandwich into a 1.5 ml Eppendorf tube.
Then, 60-200 µl of QIAzol
®
Lysis Reagent
(43)
(Qiagen, Maryland, USA, cat no. 79306) was
added to the CNON sandwich in the Eppendorf tube (QIAzol reagent must be at least 3 times
the sample volume, 20 µl, but not more than 10 times). The sample was mixed with the lysis
Table 1: Differentiation medium of the CNON sandwich neural cells is based on the “Cerebral organoids”
differentiation medium mentioned in the protocol by Lancaster and Knoblich
(21)
. Quantities are for ~100
ml batch size. Medium is filtered using a vacuum-driven 0.2-µm filter unit and stored at 2–8 °C for up to 2
weeks. A: Penicillin-Streptomycin is supplied by USC Norris Comprehensive Cancer center Bioreagent &
Cell Culture Core, Los Angeles, CA, USA. B: 1:100 dilution of 2-mercaptoethanol is prepared in the same
DMEM-F12 medium used in the recipe.
16

reagent by pipetting and left for 5 minutes with frequent shaking. Samples were stored
frozen at -80°C after each step until the next step.
Direct-zol™ RNA MicroPrep kit and protocol (Zymo Research, CA, USA, cat. no. R2060/
R2062) were used for RNA purification. The optional DNase I treatment step in the protocol
was also performed to remove trace DNA before proceeding to cDNA synthesis.

RT-qPCR
cDNA synthesis
cDNA synthesis was performed on purified RNA using a modified SuperScript
®
VILO
™
cDNA
Synthesis Kit (Invitrogen, CA, USA, cat. no. 11754050/ 11754250) in order to maximize the
production of cDNA. cDNA standards were prepared using dilutions of the Universal Human
Reference RNA (UHR) (Agilent Technologies, CA, USA, cat. no. 740000) that underwent cDNA
synthesis alongside the RNA samples.
UHR standards were prepared in dilutions of 2000 pg/µl, 1000 pg/µl, 500 pg/µl, and 250
pg/µl (a series of 1:2 dilutions to ensure accuracy). Buffer EB elution buffer from Qiagen
(Hilden, Germany, mat. no. 1014608) was used for all the steps in UHR dilution and sample
preparation for cDNA synthesis. Incubation protocol for reverse transcriptase PCR (RT-PCR)
was as follows:
1. 25°C for 10 m
2. 42°C for 2 hr.
3. 85°C for 5 m
4. 4°C forever.
17

Quantitative PCR
The quantitative PCR (qPCR) assay was performed using the PerfeCTa
®
SYBR
®
Green
FastMix
®
kit and protocol in 20 ml volume reaction on an MJ Research Chromo4™ System.
For Real-Time PCR Detection (Bio-Rad, Cat. no. CFB-3240) with Opticon Monitor™ Software
Forward and reverse18s ribosomal RNA primers
(44)
(Invitrogen, cat. no. 10336022) were
used in the reaction. This was to normalize data to 18S ribosomal RNA to account for
differences in reverse transcription efficiencies
(44, 45)
. Primer sequences are:
18S forward: 5’-GTAACCCGTTGAACCCCATT-3’
18S reverse: 5’-CCATCCAATCGGTAGTAGCG-3’
qPCR was run using Opticon Monitor™ Software. The protocol is as follows:
1. Incubate at 95.0 C for 00:00:30
2. Incubate at 95.0 C for 00:00:15
3. Incubate at 56.0 C for 00:00:15
4. Incubate at 72.0 C for 00:00:10
5. Plate Read
6. Go to line 2 for 40 more times
7. Melting Curve from 55.0 C to 90.0 C
8. END

Fluorescence NanoDrop
To validate the concentration of the cDNA products, NanoDrop 3300 Fluorospectrometer
(Thermo Scientific, DE, USA) was used to perform the RiboGreen
®
Assay for RNA
Quantitation protocol provided in the NanoDrop user manual. Quant-iT™ RiboGreen® RNA
18

Reagent (Invitrogen Molecular Probes
®
, cat. no. R11491) was used in the protocol. UHR was
used as a standard with a series of 1:2 dilutions (2000 pg/µl, 1000 pg/µl, 500 pg/µl, 250
pg/µl, and 125 pg/µl) to ensure accuracy for a high RNA concentration range. A final 1:5
dilution (25 pg/µl) was also added to detect smaller concentrations. cDNA products were
diluted 1:2 before measurement.
1X TE buffer (diluted from an RNase-free 20X stock of 200mM Tris-HCL and 20mM EDTA,
pH 7.5) (Life Technologies, USA, REF. R11490) was used for dilution of standards and
samples, and as blanks.

aRNA amplification
A modified version of the Eberwine aRNA amplification method
(46, 47)
was used for this
process.
First, RNA was denatured with 70°C in 9.35 µl of initial mix composed of 5X first strand
buffer, Oligo dT-T7 primers, 10mM dNTPs (Thermo scientific #R0193) and Dithiothreitol
(DTT). The Oligo dT-T7 primer sequence is:
5’-GGAGGCCGGAGAATTGTAATACGACTCACTATAGGGAGACGCGTGTTTTTTTTTTTTTTTTTTTTTTTTV-3’
Immediately after the denaturation, 1.75 µl of reaction mix composed of RNasin
®

Ribonuclease Inhibitor, SuperScript® III Reverse Transcriptase (Invitrogen, Cat. no.
18080085), and Diethylpyrocarbonate water (DEPC water) was added to the sample mix for
reverse transcription. The reverse transcription was performed in 42°C for 30 minutes, then
deactivated for 15 minutes at 70°C. At this point, sample can be stored at –20°C or -80°C or
used in next steps.
19

After the reverse transcription 13.81 µl second strand reaction mix was prepared. The
second strand synthesis was performed using RNase H nicking. Therefore, the mix contains
5X Second strand buffer (Invitrogen, cat. no. 10812014), additional dNTPs, DNA polymerase
I (Invitrogen, cat. no. 18010025), Ribonuclease H (RNase H) (Invitrogen, cat. no. 18021071),
and RNase Free Water.
Second strand synthesis was performed with 2-hour incubation at 16°C. After exactly 2
hours, 1μl of T4 DNA polymerase (5 U/ml, Invitrogen, cat. no. 18005025) was added for end
repair of the RNA primer generated by RNase H. This end repair was performed by
incubation at 16°C for additional 10 minutes.
The sample mix was purified using Agencourt RNAClean beads (Beckman Coulter, IN, USA,
cat. no. A63987). 52 µl of bead suspension was added to the sample mix. After 2 rounds of
purification using 70% ethanol, samples were eluted in 4 µl of nuclease free water.
The eluted samples were linearly amplified using MEGAscript
®
T7 Transcription Kit
(Invitrogen, cat. no. AMB13345). The amplification mix was composed of 1 µl of each NTP, 1
µl of 10X reaction buffer, and 10X enzyme mix. The linear amplification was performed in
37C for 14 hrs.
After the overnight incubation, the reaction was purified using Agencourt RNAClean beads.
18 µl of bead suspension was added to the sample mix. After 2 purifications, samples were
eluted in 10 µl of nuclease free water. The concentration was measured with RNA tapes on
the 2200 TapeStation (Agilent Technologies). 100 ng of the product were suspended and
dried down for cDNA synthesis.

20

cDNA synthesis with Illumina TruSeq
®

17 µl of the “Fragment, Prime, Finish” Mix from the TruSeq
®
Stranded mRNA Lib Prep kit
(Illumina, CA, USA, cat. no. RS-122-2101) was added to each well of the RNA Bead Plate (RBP)
that dried cDNA was in. The entire volume was gently pipetted up and down 6 times to mix
thoroughly.
“First Strand Synthesis Act D” Mix tube was thawed and centrifuged to 600 × g for 5 seconds.
50 µl of SuperScript II was added to the entire First Strand Synthesis Act D Mix tube, or at a
ratio of 1 µl SuperScript II for each 9 µl First Strand Synthesis Act D Mix. The new mix was
mixed gently, but thoroughly, and centrifuged briefly. 8 µl of First Strand Synthesis Act D
Mix-SuperScript II mix was added to each well of the cDNA plate (CDP) with a gentle
pipetting for the entire volume up and down 6 times for a thorough mixing.
CDP plate was sealed with a Microseal® 'B' Adhesive Seal (Bio-Rad, USA, cat. no. MSB1001)
and then centrifuged briefly. Sealed CDP plate was placed on a pre-programmed thermal
cycler and run with the following instructions:
a. pre-heat lid to 100°C
b. 25°C for 10 minutes
c. 42°C for 15 minutes
d. 70°C for 15 minutes
e. Hold at 4°C
After finishing the PCR run, CDP plate was moved immediately for the next step. Second
strand DNA synthesis, adenylate 3' ends, adapters’ ligation, DNA fragments enrichment,
21

library validation, and pool Libraries normalization were all proceeded according to the
TruSeq
®
Stranded mRNA Sample Preparation Guide provided by Illumina
®
.

Library quantification, cluster generation, and sequencing
KAPA Library Quantification Kit for Illumina
®
platforms was used to quantify the cDNA
libraries following the kit protocol.
For clustering, cDNA libraries were denatured and loaded onto the automated cBot 2 Cluster
Generation System (Illumina
®
, CA, USA) to hybridize templates onto the flow cell creating
clonal clusters. Cluster preparation was done according to the PhiX Spike-In Protocol
provided by Illumina
®
for Sequencing Runs on the HiSeq systems. Template loading
concentration was 13mM.
Flow cell was then transferred to the Illumina
®
HiSeq 2500 system for bridge amplification
and sequencing. Samples were run for single reads using the HiSeq Rapid SBS Kit v2. Final
templates were 101 bp + 7 bp Index.

Differential gene expression analysis using DESeq2
The samples FASTQ sequencing data were run through Pegasus GT-FAR (on the USC
Genomics could) to map the reads and create exon-junction gene counts files.
The individual counts files were then combined into a single file and formatted for input into
DESeq2. An annotations file was created listing each sample ID and its group (2D samples
and 3D samples). A pre-written R-script was used to run a simple 2-group comparison
22

between the 2D cultures data that was obtained from a previous work in the lab
(28)
and the
new data from RNASeq of the 3D sandwiches.

Statistical data analysis
Partek
®
Genomics Suite 6.6 was used for statistical data analysis. Read counts of sequenced
technical replicates were normalized to the 3
rd
quantile, then transformed to logarithms to
the base 2 with an off-set of +1. Further analysis like similarity measurements and
hierarchical clustering. Principal component analysis (PCA) was also performed between
technical replicates altogether, 3D sandwiches versus their corresponding 2D original data,
and for 2D technical replicates and 3D technical replicates separately. Same analysis was
performed on combined technical replicates of the same biological replicate.

Calcium imaging
Evoked calcium oscillation in CNON sandwiches was measured by using fluorescent calcium
indicators. Cells were depolarized with 100 mM potassium.
23

RESUL TS
CNON cells in 2D
Olfactory NPCs were successfully isolated subjects (98.5%). Morphology of cells in
monolayers changed, but still showed elongated branches in many cells. Phenotypically, they
are very distinct from epithelial cells. These changes in cell morphology are very consistent
from sample to sample (Fig. 1).

Figure 1: CNON cells in 2D. (a) Unattached cells during passaging. (b) CNON cells after getting embedded
in the MBM monolayer. (c, d) CNON cells start growing along the MBM bed.
24

Further morphological characterization and analysis details are found in the Evgrafov et al
paper previously mentioned
(28)
.
Morphological characterization of the ‘CNON Sandwich’ 3D cultures
We used bright field microscopy to identify the morphology of CNON cells grown in 3D. Our
3D cultures were named CNON sandwiches following the modifications on the protocol that
show a core of cells in the center surrounded by an outer layer of MBM (Fig. 2)

Network-like structures in CNON Sandwiches
When cells grow in these sandwiches they form structurally connected networks of
processes and arbors (Fig. 3-5), and grow out of their initial seeding regions.
Figure 2: seeding of cells in a 3D fashion using an MBM matrix. This method is called ‘Sandwich, which is
based on a cell containing core of MBM and outer layer of clear MBM that allows cells to grow and migrate
if they were migratory. The small seeded cells in the center are mixed uniformly with MBM and serve as
cores for the 3D cultures (c, d).
25

Figure 3: cells that grow in 3D in the CNON sandwiches grow processes and arbors that form network-like
structures. The cells look structurally connected.
26

Figure 4: a closer look on processing and interconnected cells. Cells tend to form structure like patterns
consistently.
27

Cells tend to grow outside the center towards the source of nutrients (the medium).
Once they reach the limits of the
MBM they start grow densely on
the boundaries and sometimes
release some processes out of the
MBM matrix (Fig. 6b). Cells that
grow out of the core show a
similar behavior until they
escape it and grow into the outer
layer (Fig. 6c).

Figure 5: A high resolution stack of images was taken with a 2-photon microscope and rendered on the
ImageJ software to show a rather closer view on the cells that form networks. On the right is a cell body that
projects it’s arbors in a fashion similar to pyramidal cells.
Figure 6: Cells grow away from the
center (a) and form ribbons of more
condensed cells on the boundaries (b).
Once core cells reach the outer layer
they, sometimes, form a similar
behavior until they escape towards the
outer layer of the MBM matrix (c).
28

Arborization and cell-cell communication
When cells send their processes they branch like trees and go in the direction of either
nutrients (away from the center) or other neighboring or distant cells (Fig. 7).

These branching, especially between cells, are often bilateral. Cells form hubs of sorts, where
a group of cells connect to another group via a small number of neurites between the two
relatively distant cell networks that look like bridges (Fig. 8 & 9). Sometimes, a cell network
may reach out to smaller and distal clumps of cells via large neurites (Fig. 10).
Figure 7: branching and Arborization of cells growing in the CNON sandwiches. The direction of these
processes is either away from the center (toward the source of nutrients) or toward neighboring or distant
cells.
29

Figure 8: Neurite formation that resembles unilateral and bilateral cell-cell communication.
30

Figure 9: Bridge like neurites that connect small networks of cells making them connected like stations.
31

Figure 10: a network of cells (a) sends a neurite-like projection (b, c) toward a distant clump of cells.
32

These structures are consistent in all stages in most of the sandwiches we grew in the lab.
This consistency is also present in cells from patients with SCZ and controls as well. Figure
11 shows structures and patterns found in SCZ cells and control cells. Figure 12 shows
similar findings in the very CNON sandwiches we used for RNASeq.

Figure 11: similarities in the general structure of CNON sandwiches can be found in SCZ cell and control
cells.
33

Figure 12: Similarities in the general structure of CNON sandwiches were also found in the samples we
used for RNASeq
34

Functional Characterization
Calcium imaging was performed on 4 CNON-sandwich samples (Fig. 13). The data clearly
shows the cells response to depolarizing as compared with spontaneous calcium oscillation
activity.

RNA purification and quantitation
RNA was purified from 6 samples that represent 2 sets of 3 technical replicates. One set was
from a SCZ cells line (denoted by 3DS075), and the other was from a control cell line
(3DS074). In the sample notations, 3DS stands for 3D sandwich, and the number refers to
the cell line.
After purifying RNA, concentration was measured by RT-qPCR and validated by fluorescence
NanoDrop. aRNA amplification was planned based on the initial RNA concentration.
Libraries were prepared for the samples to be run for RNASeq.
Figure 13: Evoked calcium oscillation was measured in CNONs in 3D sandwiches using a fluorescent
calcium indicator. On the left is the relative fluorescence of spontaneous calcium oscillation in 4 color-
coordinated samples, and on the right is the relative fluorescence of calcium oscillation in cells that had
been depolarized with 100 mM KCL.
35

RNASeq
Each biological replicate had three technical replicates. We had previous data from Evgrafov
et al
(28)
for 2D cultures. Each biological replicate has also 3 technical replicates. Notation for
the 2D cultures was SEPL followed by the cell line number. Hence, we obtained 3 technical
replicates for SEPL045, and 3 technical replicates for SEPL074 (Table 2).
Cell line CNON045 CNON074
Biological replicates 2D (SEPL045) 3D (3DS045) 2D (SEPL074) 3D (3DS074)
Technical replicates 3 3 3 3
RNA from the six 3D samples was purified and prepared for RNASeq. Reads from the
sequencing data were mapped to transcriptome (GenCode v.22). For 2D samples, raw data
from the previous study by Evgrafov et al
(28)
were rerun on the same version of the software
and GenCode annotation for proper comparison. An annotations file was created listing each
sample ID and its group (2D samples and 3D samples) on DESeq2. A simple 2-group
comparison between the 2D technical replicates and their 3D counterparts. The same
comparison was done for the biological replicates by combining the read counts of the
technical replicates that belong to the same group or biological replicate.

Data analysis
Principal component analysis (PCA) of the 12 technical replicates show a high correlation
between technical replicates that belong to the same group or biological replicate, and those
replicates cluster together in a scatterplot (Fig. 14).
Table 2: the technical replicates RNASeq data
36

Correlation of the expression profile for technical replicates of each group was measured
on scatterplots. Technical replicates for each of the two 2D samples (SEPL) show a high
correlation (Fig. 15 A). On the other hand, technical replicates for each of the two 3D
samples (3DS) show a correlation that is not as high as that of 2D samples (Fig. 15 B).
A list of 88 genes that were expressed in all 12 technical replicates, and are involved in
synaptic connectivity, were chosen based on a recommendation from Dr. Marcello Coba,
from USC’s Zilkha Neurogenetic Institute, as references for the consistency of gene
expression in synaptic machinery. These genes show a high correlation even in technical
replicates of 3D samples (were shown as purple dots in the scatterplots).

Figure 14: PCA of the 12 different technical replicates. Replicates here are color-coded based on their
biological replicate origin. The balls with the two purple shades are 3D samples and the blue ones are 2D.
The lighter shade of either color is for CNON045 samples, and the darker shade of either color is for
CNON074 samples.
37

A.

B.

Figure 15: scatter plots in three dimensions show the correlation between the 3 technical replicates of
each group (biological replicate). Each dimension of the scatter plot corresponds to one of the 3 technical
replicates of each group. (A) shows correlation in 2D samples (a and b for SEPL045, and c and d for SEPL074
technical replicates). (B) shows correlation in 3D samples (a and b for 3DS045, and c and d for 3DS074
technical replicates. Purple dots that label synaptic genes are aligned along the R line (not shown) with no
outliers in all 4 groups.
38

Differential gene expression
Technical replicates for each sample show a correlation in gene expression. We combined
the read counts in the expression profiles of the 3 technical replicates of each sample. We
used the combined expression profiles to study differential gene expression of the two 3D
samples in respect their 2D counterparts.
We wanted to show the correlation between 3D samples and between 2D samples. So, we
plotted the two 3D samples against each other, and we did the same for the two 2D samples
(Fig. 16). Correlation between the two SEPL045 and SEPL074 samples, which were grown in
2D conditions, was high with some outliers. To a lesser extent, correlation between the two
3DS045 and 3DS074 samples, which were grown in 3D conditions (sandwiches), can also be
seen.
We also wanted to see the differences in the gene expression between growing cells in 2D
and growing them in 3D. We plotted SEPL045 against 3DS045, and SEPL074 against 3DS074
(Fig. 17). Scatterplots of 2D vs. 3D show weak correlation in both cell line, which means that
there was a significant change in the expression profile.
We normalized combined read counts to the 3
rd
quantile and transformed the values to their
logarithm to the base 2 with a +1 off-set. We filtered the genes in their original read counts
to the ones with a base mean > 10. Number of genes that were filtered in was 17,859 genes.
Normalized combined read counts from these genes were used to perform a hierarchical
clustering for the 4 biological replicates: two monolayer and two sandwich cultures. Fig. 18
shows a heat map for the differential expression of these genes in the four samples.
39

Figure 16: Correlation of gene expression in 2D cultures and 3D cultures. (a) Scatterplot of gene expression
for SEPL045 and SEPL074, both grown in monolayers (i.e. 2D). (b) Scatterplot of gene expression for
3DS045 and 3DS074, both grown in sandwiches (i.e. 3D)
40

Hierarchical clustering of genes from the 4 samples (Fig. 18) shows distinctive patterns of
differentially expressed genes in in the two 3D samples.
Figure 17: Differential gene expression in 2D vs. 3D. (a) Scatterplot of gene expression for SEPL045 (2D)
against 3DS045 (3D), both of which are from the same CNON cell line. (b) Scatterplot of gene expression
for SEPL074 (2D) against 3DS074 (3D), both of which are from the same CNON cell line.
41

Figure 18: Hierarchical clustering of genes with read counts that have a base mean >10. The two lanes on the left are for the two 2D samples. The two
lanes on the right are for the two 3D samples. Heat map shows differentially expressed genes with a color code on a spectrum between blue (low-
expressed) to red (highly expressed). 2D samples show a more uniform pattern as opposed to 3D samples.
42

Monolayer samples show a more uniform pattern of differentially expressed genes.
Sandwich samples show a less uniform but a distinctive pattern of some differentially
expressed genes in both combined samples.
Of these 17,859 genes in the hierarchical clustering, 5,123 genes are differentially expressed
in 3D cultures with a Log2 fold change of 2 or higher, and an adjusted p-value > 0.05. 543 of
the aforementioned differentially expressed genes have a Log2 fold change of 5 or higher.
We used DAVID Bioinformatics Resources
(48, 49)
to study the functional annotation of the top
543 differentially expressed genes in CNON sandwiches in respect to monolayer CNON
cultures. Genes were divided into 2 lists: Genes with higher expression in 3D cultures as
compared to 2D cultures, and genes with lower expression in 3D cultures. We used GOTERM
categories to list the functional annotation. Functional annotations associated with low
expressed genes were mostly involved in mitosis and cell cycle with p-value < 0.05 for most
of them. Functional annotations associated with highly expressed genes didn’t have p-values
as low as those for low expressed genes. However, functional annotations with normal
(0.01< p-value <0.05) and low (<0.01) p-values are mostly involved in cell metabolism, cell
signaling, and synaptic transmission. SNAP25 gene, an important gene involved in synaptic
activity, is heavily present in these functions and is highly expressed in the 2 CNON
sandwiches.
We asked Dr. Marcello Coba to help us choose the expressed synaptic genes that are highly
or exclusively involved in synapses. He used the Psychiatric Pathway Protein Resource
(PsyPPRes) online software to see expressed genes that are involved in protein-protein
43

interactions in synapses. A list of 88 genes that are highly or exclusively involved in synaptic
connectivity were chosen.
We performed hierarchical clustering for these genes in the four combined samples and
found that few of these genes are highly expressed in the two 3D samples but not in 2D
samples (Fig. 19), most notably SNAP25, DLG4, and GRIN2B. Few others were
underexpressed like CASKIN2.
Figure 19: Hierarchical clustering of 88 differentially expressed genes that are highly or exclusively
involved in synaptic structure or function. Some of the genes are expressed in 3D samples such as SNAP25,
DLG, and GRIN4. Few are underexpressed, such as CASKIN2.
44

DISCUSSION
The main goal of developing CNON 3D cultures is to improve the precious 2D CNON model
our lab developed to study brain cells in SCZ, in order to better capture potential molecular
differences related to cell migration, connectivity and, possibly, synaptic function.
Experimental data support the idea that certain brain circuits are affected in SCZ, and that
the synaptic structure and/or function in these circuits is/are impaired. Alterations in
synaptic function would result from changes in the expression of genes whose protein
products are involved in synaptic function and vice versa
(5)
.
Studying SCZ in animals turned out to be ‘problematic’ and ‘daunting’ as it cannot
recapitulate the complexity of the human brain

(10)
. On the other hand, human cell-based
models preserve species-specific properties.
NPCs provide the most adequate source of cells to study neurodevelopment, and modeling
them in a 3D fashion makes them more relevant to brain cells. So far, the significant
physiological relevance of neural tissue models gives them the priority over simplicity,
control and reproducibility of 2D cultures
(17)
, but we are trying to overcome some of the
challenges by providing a better protocol.
ON proved to be an easy source for NPCs in the form of CNON
(28)
. CNON cells are highly
replicative and relatively easy to maintain.
Growing CNON cells in 3D in MBM sandwiches changed their behavior. In order for the cells
to grow and migrate, cells were mixed in liquid MBM, droplets of this matrix-cells mix were
solidified on parafilm, and then coated with 3 times their volume of MBM. Cells grew within
45

the core and migrated toward and inside the outer MBM layer. They formed networks with
each other and with distal cells as well. Similar neuronal cell networks are the basis of
communication in the brain. If these connections turn out to be synaptically functional, these
cells may give a very good representation of what is going on between cells in the brain.
For that reason, we wanted to study the expression profile of cells in these CNON networks.
Data show that CNON cells in 3D cultures are different from the CNON cells from the same
sample when they were grown in 2D. Their expression profiles indicate that they are closer
to differentiated neuronal cells.
The differentiation medium we used was designed to support a developmental progression
of both NPCs and their progeny similar to that of the neural ectoderm in vivo, and data from
our work and others

(21)
show that it somewhat does, but further experiments on the medium
contents and the conditions in which cells are grown need to be done.
Our CNON sandwiches showed consistency in certain morphological characteristics across
different samples and cultures. However, these cultures were also sensitive to changes in
media constituents, structure of the MBM matrix, or distribution of cells in the core matrix.
This was also reflected in the expression profile of these cultures. Unlike CNON 2D cultures,
gene expression in CNON sandwiches was more heterogeneous, however, some genes and
groups of genes were consistently differentially expressed.
Different CNON sandwiches show more correlation than the same CNON cell line grown in
sandwiches and in monolayers. This means that CNON sandwiches share a common feature
46

in gene expression that is different from gene expression patterns in 2D cultures. This could
reflect (a) different differentiation stage(s) in neuronal cells in the brain.
In order to support these findings, more samples need to be studied, and more aspects of
these cells should be explored. Protein-protein interactions in potential synaptic locations
also need to be explored. The 3D model we are have developed should be a suitable platform
for further biochemical and pharmacological experiments, and we expect the model to be
refined.
Successful human neural 3D cultures provide a very helpful tool in a myriad of fields such as
drug testing, development, and characterization of neurological disorders

(19,20)
, and our
model should augment the approaches available .
47

CONCLUSION
In conclusion, we developed a novel method of growing CNON cells in 3D culture based on a
“sandwich” structure of MBM. This method is more consistent compared to other
approaches, and is a convenient approach for studying cell migration and cell connectivity.
Microscopy studies confirmed that culturing CNON cells in such sandwich-like MBM
structures allowed the cells to grow, migrate, and project neurite-like processes similar to
neuronal cells in the brain. They also appeared to be communicating with each other.
The gene expression profile in CNON cells growing in 3D changed significantly, as compared
to cells from the same sample growing in 2D. For example, the genes involved in mitosis and
cell division were underexpressed in 3D cultures, and many important synaptic genes in cells
grown in 3D were expressed, some of which are significantly overexpressed as compared to
cells in 2D.
It is not clear if synaptic machinery in CNON cells growing in 3D is fully functional, but these
dramatic changes in synaptic gene expression indicate that our 3D model may be helpful in
studying neuronal connectivity.
Our study demonstrates the potential of our novel model system to advance studies of SCZ.
However, some characteristics of our 3D cultures will need further characterization, such as
electrophysiology of the cells, the nature and extent of protein-protein interactions, and
specific gene expression profiles of cells, especially as correlated with the appearance of
particular morphological structures in our 3D cultures.
48

CNON sandwiches are a promising base on which to develop a human cell-based model to
study the development of neuronal cells in SCZ compared to controls. They can, also, provide
a model system to study drugs and treatment options and their effects on brain cells in SCZ.
Further improvements on the ‘CNON sandwich’ model system may provide an easier
alternative to model the complexity of the human brain.
49

REFERENCES
1- Torrey EF. Prevalence studies in schizophrenia. Br J Psychiatry. 1987 May; 150:598-
608.
2- Goldner EM, Hsu L, Waraich P, and Somers JM. Prevalence and incidence studies of
schizophrenic disorders: a systematic review of the literature. Can J Psychiatry. 2002
Nov; 47(9):833-43.
3- Weinberger DR. Implications of normal brain development for the pathogenesis of
schizophrenia. Arch Gen Psychiatry. 1987 Jul; 44(7):660-9.
4- Raedler TJ, Knable MB, and Weinberger DR. Schizophrenia as a developmental
disorder of the cerebral cortex. Curr Opin Neurobiol. 1998 Feb; 8(1):157-61.
5- Lewis DA and Levitt P. Schizophrenia as a disorder of neurodevelopment. Annu Rev
Neurosci. 2002; 25:409-32.
6- Mueser KT and McGurk SR. Schizophrenia. Lancet. 2004 Jun 19; 363(9426):2063-72.
7- Brennand K, Savas JN, Kim Y, Tran N, Simone A, Hashimoto-Torii K, Beaumont KG,
Kim HJ, Topol A, Ladran I, Abdelrahim M, Matikainen-Ankney B, Chao SH, Mrksich M,
Rakic P, Fang G, Zhang B, Yates JR , and Gage FH. Phenotypic differences in hiPSC NPCs
derived from patients with schizophrenia. Mol Psychiatry. 2015 Mar; 20(3):361-8.
8- Eric R. Marcotte, Debra M. Pearson, and Lalit K. Srivastava. Animal models of
schizophrenia: a critical review. J Psychiatry Neurosci. 2001 Nov; 26(5): 395–410.
9- Kamiya A. Animal models for schizophrenia via in utero gene transfer: understanding
roles for genetic susceptibility factors in brain development. Prog Brain Res. 2009;
179:9-15.
50

10- Young JW, Zhou X, and Geyer MA. Animal models of schizophrenia. Curr Top Behav
Neurosci. 2010; 4:391-433.
11- Sterneckert JL, Reinhardt P, and Schöler HR. Investigating human disease using stem
cell models. Nat Rev Genet. 2014 Sep; 15(9):625-39.
12- Thomson JA, Itskovitz-Eldor J, Shapiro SS, Waknitz MA, Swiergiel JJ, Marshall VS, and
Jones JM. Embryonic stem cell lines derived from human blastocysts. Science. 1998
Nov 6; 282(5391):1145-7.
13- Paşca AM, Sloan SA, Clarke LE, Tian Y, Makinson CD, Huber N, Kim CH, Park JY,
O'Rourke NA, Nguyen KD, Smith SJ, Huguenard JR, Geschwind DH, Barres BA, and
Paşca SP. Functional cortical neurons and astrocytes from human pluripotent stem
cells in 3D culture. Nat Methods. 2015 Jul; 12(7):671-8.
14- Maherali N, Ahfeldt T, Rigamonti A, Utikal J, Cowan C, and Hochedlinger K. A high-
efficiency system for the generation and study of human induced pluripotent stem
cells. Cell Stem Cell. 2008 Sep 11; 3(3):340-5.
15- Aasen T, Raya A, Barrero MJ, Garreta E, Consiglio A, Gonzalez F, Vassena R, Bilić J,
Pekarik V, Tiscornia G, Edel M, Boué S, and Izpisúa Belmonte JC. Efficient and rapid
generation of induced pluripotent stem cells from human keratinocytes. Nat
Biotechnol. 2008 Nov; 26(11):1276-84.
16- Brennand KJ, Simone A, Tran N, and Gage FH. Modeling psychiatric disorders at the
cellular and network levels. Mol Psychiatry. 2012 Dec; 17(12):1239-53.
17- Irons HR, Cullen DK, Shapiro NP, Lambert NA, Lee RH, and Laplaca MC. Three-
dimensional neural constructs: a novel platform for neurophysiological investigation.
J Neural Eng. 2008 Sep; 5(3):333-41.
51

18- Helena T Hogberg, Joseph Bressler, Kimberly M Christian, Georgina Harris, Georgia
Makri, Cliona O'Driscoll, David Pamies, Lena Smirnova, Zhexing Wen, and Thomas
Hartung. Toward a 3D model of human brain development for studying
gene/environment interactions. Stem Cell Res Ther. 2013; 4(Suppl 1): S4.
19- Hogberg HT1, Sobanski T, Novellino A, Whelan M, Weiss DG, and Bal-Price AK.
Application of micro-electrode arrays (MEAs) as an emerging technology for
developmental neurotoxicity: evaluation of domoic acid-induced effects in primary
cultures of rat cortical neurons. Neurotoxicology. 2011 Jan; 32(1):158-68.
20- Pamies D, Hartung T, and Hogberg HT. Biological and medical applications of a brain-
on-a-chip. Exp Biol Med (Maywood). 2014 Sep; 239(9):1096-107.
21- Lancaster MA and Knoblich JA. Generation of cerebral organoids from human
pluripotent stem cells. Nat Protoc. 2014 Oct; 9(10):2329-40.
22- Eiraku M and Sasai Y. Self-formation of layered neural structures in three-
dimensional culture of ES cells. Curr Opin Neurobiol. 2012 Oct; 22(5):768-77.
23- Karus M, Blaess S, and Brüstle O. Self-organization of neural tissue architectures from
pluripotent stem cells. J Comp Neurol. 2014 Aug 15; 522(12):2831-44.
24- Brännvall K, Bergman K, Wallenquist U, Svahn S, Bowden T, Hilborn J, and Forsberg-
Nilsson K. Enhanced neuronal differentiation in a three-dimensional collagen-
hyaluronan matrix. J Neurosci Res. 2007 Aug 1; 85(10):2138-46.
25- Lancaster MA, Renner M, Martin CA, Wenzel D, Bicknell LS, Hurles ME, Homfray T,
Penninger JM, Jackson AP, and Knoblich JA. Cerebral organoids model human brain
development and microcephaly. Nature. 2013 Sep 19; 501(7467):373-9.
52

26- Lancaster MA and Knoblich JA. Organogenesis in a dish: modeling development and
disease using organoid technologies. Science. 2014 Jul 18; 345(6194):1247125.
27- Perry C, Mackay-Sim A, Feron F, and McGrath J. Olfactory neural cells: an untapped
diagnostic and therapeutic resource. The 2000 Ogura Lecture. Laryngoscope. 2002
Apr; 112(4):603-7.
28- Evgrafov OV, Wrobel BB, Kang X, Simpson G, Malaspina D, Knowles JA. Olfactory
neuroepithelium-derived neural progenitor cells as a model system for investigating
the molecular mechanisms of neuropsychiatric disorders. Psychiatr Genet. 2011 Oct;
21(5):217-28.
29- Bruce I. Turetsky, Chang-Gyu Hahn, Karin Borgmann-Winter, and Paul J. Moberg.
Scents and Nonsense: Olfactory Dysfunction in Schizophrenia. Schizophr Bull. 2009
Nov; 35(6): 1117–1131.
30- Bozena B. Wrobel, Jill M. Mazza, Oleg V. Evgrafov, and James A. Knowles. Assessing
the Efficacy of Endoscopic Office Olfactory Biopsy Sites to Produce Neural Progenitor
Cell Cultures for the Study of Neuropsychiatric Disorders. Int Forum Allergy Rhinol.
2013 Feb; 3(2): 133–138.
31- Féron F, Perry C, Hirning MH, McGrath J, and Mackay-Sim A. Altered adhesion,
proliferation and death in neural cultures from adults with schizophrenia. Schizophr
Res. 1999 Dec 21; 40(3):211-8.
32- Zhang X, Cai J, Klueber KM, Guo Z, Lu C, Winstead WI, Qiu M, and Roisen FJ. Role of
transcription factors in motoneuron differentiation of adult human olfactory
neuroepithelial-derived progenitors. Stem Cells. 2006 Feb; 24(2):434-42.
53

33- Cascella NG, Takaki M, Lin S, and Sawa A. Neurodevelopmental involvement in
schizophrenia: the olfactory epithelium as an alternative model for research. J
Neurochem. 2007 Aug; 102(3):587-94.
34- Wolozin B, Lesch P, Lebovics R, and Sunderland T. A.E. Bennett Research Award
1993. Olfactory neuroblasts from Alzheimer donors: studies on APP processing and
cell regulation. Biol Psychiatry. 1993 Dec 15; 34(12):824-38.
35- Johnson GS, Basaric-Keys J, Ghanbari HA, Lebovics RS, Lesch KP, Merril CR,
Sunderland T, and Wolozin B. Protein alterations in olfactory neuroblasts from
Alzheimer donors. Neurobiol Aging. 1994 Nov-Dec; 15(6):675-80.
36- Ghanbari HA, Ghanbari K, Harris PL, Jones PK, Kubat Z, Castellani RJ, Wolozin BL,
Smith MA, and Perry G. Oxidative damage in cultured human olfactory neurons from
Alzheimer's disease patients. Aging Cell. 2004 Feb; 3(1):41-4.
37- Gizurarson S. Anatomical and histological factors affecting intranasal drug and
vaccine delivery. Curr Drug Deliv. 2012 Nov; 9(6):566-82.
38- Hüfner K, Frajo-Apor B, and Hofer A. Neurology issues in schizophrenia. Curr
Psychiatry Rep. 2015 May; 17(5):32.
39- Nguyen AD, Shenton ME, and Levitt JJ. Olfactory dysfunction in schizophrenia: a
review of neuroanatomy and psychophysiological measurements. Harv Rev
Psychiatry. 2010 Sep-Oct; 18(5):279-92.
40- Mackay-Sim A. Concise review: Patient-derived olfactory stem cells: new models for
brain diseases. Stem Cells. 2012 Nov; 30(11):2361-5.
41- Ambesi-Impiombato FS, Parks LA, and Coon HG. Culture of hormone-dependent functional
epithelial cells from rat thyroids. Proc Natl Acad Sci U S A. 1980 Jun; 77(6):3455-9.
54

42- H G Coon, F Curcio, K Sakaguchi, M L Brandi, and R D Swerdlow. Cell cultures of neuroblasts
from rat olfactory epithelium that show odorant responses. Proc Natl Acad Sci U S A. 1989
Mar; 86(5): 1703–1707.
43- Chomczynski P and Sacchi N. Single-step method of RNA isolation by acid guanidinium
thiocyanate-phenol-chloroform extraction. Anal Biochem. 1987 Apr; 162(1):156-9.
44- Schmittgen TD and Zakrajsek BA. Effect of experimental treatment on housekeeping gene
expression: validation by real-time, quantitative RT-PCR. J Biochem Biophys Methods. 2000
Nov 20; 46(1-2):69-81.
45- Xin X1, Khan ZA, Chen S, Chakrabarti S. Extracellular signal-regulated kinase (ERK) in
glucose-induced and endothelin-mediated fibronectin synthesis. Lab Invest. 2004 Nov;
84(11):1451-9.
46- Morris J, Singh JM, and Eberwine JH. Transcriptome analysis of single cells. J Vis Exp. 2011
Apr 25 ;( 50). pii: 2634. Doi: 10.3791/2634.
47- Morris J, Bell TJ, Buckley PT, and Eberwine JH. Antisense RNA amplification for target
assessment of total mRNA from a single cell. Cold Spring Harb Protoc. 2014 Nov 3;
2014(11):1149-60. doi: 10.1101/pdb.prot072454.
48- Huang da W, Sherman BT, and Lempicki RA. Systematic and integrative analysis of large gene
lists using DAVID bioinformatics resources. Nat Protoc. 2009; 4(1):44-57. doi:
10.1038/nprot.2008.211.
49- Huang da W, Sherman BT, and Lempicki RA. Bioinformatics enrichment tools: paths toward
the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 2009 Jan;
37(1):1-13. doi: 10.1093/nar/gkn923. Epub 2008 Nov 25.

Abstract (if available)

Abstract Schizophrenia is a devastating neurological disorder. Multiple lines of evidence support the hypothesis that abnormal neurodevelopment is involved in the etiology of schizophrenia. Abnormality in synaptic connectivity and/or function has also been found in the brains of individuals with schizophrenia. ❧ Our lab has previously developed 250 lines of neural cells derived from the olfactory neuroepithelium of schizophrenic and control subjects. Cells growing in two-dimensional cultures were characterized by RNA-Seq as neural progenitor cells. Most of the genes that are differentially expressed comparing cells from schizophrenia and control are related to neurodevelopment. Genes involved in synaptogenesis are either not expressed or expressed at very low levels. ❧ Here, we provide preliminary data on cultured neuronal cells derived from the olfactory neuroepithelium (known as CNON) and maintained in conditions promoting cell differentiation and growth in a three-dimensional fashion. We noticed dramatic increases in the expression of some synaptic machinery genes. We, also, found that these 3D cell cultures consistently are organized into reproducible morphological structures. ❧ We suggest that 3D CNON cultures may serve as a model system for studying altered neurodevelopment and neuronal connectivity in schizophrenia, allowing the identification of genes related to abnormal structure and function of the schizophrenic brain.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Best practice development for RNA-Seq analysis of complex disorders, with applications in schizophrenia

PDF

An induced neuron cell model derived from human olfactory epithelial cells

PDF

Neurogenic placodes provide migratory enteric sensory neural progenitors in response to endothelin signaling pathway

PDF

Modeling SynGAP1 truncating mutations in neurodevelopmental disease using iPSC-derived neurons

PDF

Development of ECM for the preservation of adult mouse pancreatic islet function in vitro

PDF

A functional genomic approach based on shRNA-mediated gene silencing to delineate the role of NF-κB and cell death proteins in the survival and proliferation of KSHV associated primary effusion l...

PDF

Do ZFX and ZNF711 regulate the same genes in HEK293T cells?

PDF

A synthetic lethal screen for NF-κB-dependent plasma cell disorders

PDF

Identification and characterization of the enhancer elements for lymphatic-specific expression of Prox1

PDF

Novel roles for Maf1 in embryonic stem cell differentiation and adipogenesis

PDF

SLIT3 gene expression by 1,25(OH)₂D₃ in an endometriosis stromal cell line

PDF

LINC00261 induces a G2/M cell cycle arrest and activation of the DNA damage response in lung adenocarcinoma

PDF

DNA methylation markers for blood-based detection of small cell lung cancer in mouse models

PDF

Development of immunotherapy for small cell lung cancer using novel modified antigens

PDF

Role of the bone marrow niche components in B cell malignancies

PDF

Induction of hypersignaling as a therapeutic approach for treatment of BCR-ABL1 positive Acute Lymphoblastic Leukemia (ALL) cells

PDF

Using novel small molecule modulators as a tool to elucidate the role of the Myocyte Enhancer Factor 2 (MEF2) family of transcription factors in leukemia

Asset Metadata

Creator Muhtaseb, Abdurrahman Wisam (author)

Core Title Cultured neuronal cells derived from the olfactory neuroepithelium growing in three dimensions as a model system for schizophrenia

School Keck School of Medicine

Degree Master of Science

Degree Program Biochemistry and Molecular Biology

Publication Date 07/28/2018

Defense Date 05/18/2016

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag 3D cultures,expression profile,OAI-PMH Harvest,RNA-seq,schizophrenia,transcriptome

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Bajpai, Ruchi (committee chair), Chow, Robert (committee member), Evgrafov, Oleg (committee member), Knowles, James (committee member), Tokes, Zoltan (committee member)

Creator Email awmuhtaseb@live.com,muhtaseb@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-291642

Unique identifier UC11281464

Identifier etd-MuhtasebAb-4708.pdf (filename),usctheses-c40-291642 (legacy record id)

Legacy Identifier etd-MuhtasebAb-4708.pdf

Dmrecord 291642

Document Type Thesis

Format application/pdf (imt)

Rights Muhtaseb, Abdurrahman Wisam

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA