University of Southern California Dissertations and Theses

An iPSC-based biomarker strategy to identify neuroregenerative responders to allopregnanolone

(USC Thesis Other)

An iPSC-based biomarker strategy to identify neuroregenerative responders to allopregnanolone

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

AN IPSC-BASED BIOMARKER STRATEGY TO IDENTIFY
NEUROREGENERATIVE RESPONDERS TO ALLOPREGNANOLONE

by

Christine Marie Solinsky

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(CLINICAL AND EXPERIMENTAL THERAPEUTICS)

December 2017

Copyright 2017 Christine Marie Solinsky
ii
Dedication

To all my cell babysitters, without whose help I would have missed
Thanksgivings and Christmases at home, weddings, birthdays, and various other
key life events… and to my family who recognized that I’ve had billions of
surrogate children over the last five years and been understanding.

iii
Acknowledgements

I have a distinct memory from middle school telling my Mom that there was no
way I was going to get a doctorate degree… here we are an unmentionable
number of years later – I promise, I’m done! Most importantly, thank you from
the bottom of my heart to my family (genetic and otherwise) who were here for
me during all the ups and the downs, providing love, encouragement, countless
prayers, laughs, cocktails, TV episodes, and other distractions. I don’t regret a
minute and wouldn’t have made it without you.

To my advisor, Roberta Diaz Brinton, thank you for your mentorship, not only in
science but for pushing me to be comfortable outside my comfort zone. I’ve had
a unique, valuable PhD experience that I couldn’t have had with anyone else.
And to all the other women in science who’ve inspired me over the years
including Kathleen Rodgers, Diane Stephenson, Deirdre Corrigan, and Aisling
O’Leary to name but a few. You’ve all played a part in the scientist I am today
and where I want to go.

To Brian Kirby – for my first independent research project, for encouraging me to
look past negative results, and to persevere in research.

To the members of the Brinton Lab, past and present - I learned something from
each of you and am incredibly lucky to count you as colleagues and friends.
iv
Thank you to the members of the Ichida lab for sharing your space and time as I
learned all about reprogramming, differentiation, and stem cell culture. Many
thanks also to Naoko Kono for conducting the correlational analyses presented
here.

Last, but most certainty not least, I want to thank all the participants in the
Allopregnanolone clinical trial who donated blood samples for the studies
described here. Your struggle and bravery to participating in scientific research
was the motivation to keep striving forward.
v
Abbreviations Index
3xTgAD Triple transgenic Alzheimer’s Disease Mouse Model
AD Alzheimer’s Disease
Allo Allopregnanolone
ADAS-Cog 13 Alzheimer’s Disease Assessment Scale-Cognitive-Plus 13 item
scale
ADRC Alzheimer’s Disease Research Center
APP Amyloid precursor protein
ARIA Amyloid-relating imaging abnormalities
ALS Amyotrophic lateral sclerosis
Aβ β-amyloid
DSMB Data Safety Monitoring Board
DMN Default mode network
GCMLDG Granule cell and molecular layers of the dentate gyrus
CA Cornu Amonis
DTI Diffusion tensor imaging
DMSO Dimethyl sulphoxide
EdU 5-ethynyl-2´-deoxyuridine
ER Endoplasmic reticulum
EtOH Ethyl alcohol
EMA European Medicines Agency
fAD Familial AD
FBS Fetal Bovine Serum
FCCP Carbonylcyanide p-trifluoromethoxyphenylhydrazone
FDA US Food and Drug Administration
FXTAS Fragile-X associated tremor/ataxia syndrome
GABA Gamma aminobutyric acid
cGMP Current good manufacturing practice
HATA Hippocampal amygdala transition area
hESC Human embryonic stem cells
iPSCs induced pluripotent stem cells
IEM Inherited erythromelalgia
IND Investigational new drug
iPS-CMs iPSC-derived cardiomyocytes
IV Intravenous
LSM Lymphocyte Separation Medium
MRI Magnetic resonance imaging
MTD Maximally tolerated dose
MCI Mild cognitive impairment
MMSE Mini-mental state exam
MOCA Montreal Cognitive Assessment
vi
MEFs Mouse embryonic fibroblasts
MS Multiple sclerosis
NSCs Neural Stem Cells
OCR Oxygen consumption rate
PD Parkinson’s Disease
PBMCs Peripheral blood mononuclear cells
PMDA Pharmaceuticals and Medical Devices Agency
PK Pharmacokinetics
PBS Phosphate buffered saline
PSEN1 Presenilin-1
ROS Reactive oxygen species
RI Rock inhibitor
SMA Spinal muscular atrophy
sAD Sporadic AD
SBECD Sulfabutylether-beta-cyclodextrin
SRSE Super refractory status epilepticus
USC IRB University of Southern California Health Sciences Institutional
Review Board
vii

Table of Contents
DEDICATION II
ACKNOWLEDGEMENTS III
ABBREVIATIONS INDEX V
TABLE OF CONTENTS VII
LIST OF TABLES IX
LIST OF FIGURES X
ABSTRACT XII
CHAPTER 1: BACKGROUND 1
ALZHEIMER’S DISEASE: BACKGROUND 1
ALZHEIMER’S DISEASE: DRUG DEVELOPMENT 1
ALLOPREGNANOLONE: A NOVEL REGENERATIVE THERAPEUTIC 2
ALLOPREGNANOLONE: PRE-CLINICAL EFFICACY 4
ALLOPREGNANOLONE: CLINICAL SAFETY 6
ALLOPREGNANOLONE: PROOF-OF-CONCEPT 6
INDUCED PLURIPOTENT STEM CELLS: BACKGROUND 9
IPSC APPLICATION: CELL TRANSPLANTATION 16
IPSC APPLICATION: PRECISION MEDICINE 18
STUDY HYPOTHESIS 23
CHAPTER 2: ALLOPREGNANOLONE PHASE 1 CLINICAL TRIAL 25
OVERVIEW: 25
CLINICAL TRIAL OUTCOMES: 27
CLINICAL STUDY PARTICIPANTS: 30
FORUMLATION 34
PHARMACOKINETICS 34
STATISTICS 34
STUDY ETHICS 35
STUDY CLINICAL TRIAL STATUS 35
MATERIALS AND METHODS 36
RESULTS 37
DISCUSSION 39

viii
CHAPTER 3: ISPC GENERATION & DIFFERENTIATION TO NSC AND BASELINE EVALUATION OF
PROLIFERATION AND MITOCHONDRIAL RESPIRATION 40
INTRODUCTION 40
MATERIALS AND METHODS 41
RESULTS 58
DISCUSSION 69
CHAPTER 4: NSC PROLIFERATION & ALLOPREGNANOLONE 71
INTRODUCTION: 71
MATERIALS & METHODS 71
RESULTS 72
DISCUSSION 79
CHAPTER 5: NSC MITOCHONDRIAL FUNCTION & ALLOPREGNANOLONE 81
INTRODUCTION 81
MATERIALS & METHODS 81
RESULTS 83
DISCUSSION 94
CHAPTER 6: IN VITRO AND CLINICAL CORRELATIONS 97
INTRODUCTION 97
MATERIALS AND METHODS 98
RESULTS 99
CHAPTER 7: DISCUSSION 115
IPSC MODEL OF SAD AND ALLOPREGNANOLONE RESPONSE 115
IN VITRO - CLINICAL CORRELATIONS 120
REFERENCES 131
ix
List of Tables

Table 1.1. Synopsis of AD studies utilizing iPSC-derived neural cells .................... 14
Table 1.2. Number of AD donor cell lines evaluated in previous iPSC-based AD
studies. ................................................................................................................... 15
Table 2.1. Allopregnanolone Phase 1 Clinical Trial Participant Demographics ...... 33
Table 3.1. Non-Integrating Episomal Plasmid Mixture ............................................... 46
Table 3.2. Summary of iPSC Karyotyping ................................................................... 63
Table 6.1. Clinical Volumetric MRI and Cognitive Test Changes after 12 Weeks of
Allo. ....................................................................................................................... 102
Table 6.2. Spearman Correlations for Allo Participants Only ................................. 104
Table 6.3. Spearman Correlations for All Clinical Trial Participants. ..................... 105
x
List of Figures

Figure 1.1. Allopregnanolone Neurogenic mechanims of action in NSCs ................. 3
Figure 1.2. Sedation as a Biomarker of Target Engagement ...................................... 7
Figure 1.3. Illustration of the iPSC approach applied to clinical trials. .................... 19
Figure 1.4. Project Flow Diagram ................................................................................. 23
Figure 2.1. Allopregnanolone Phase 1 Clinical Trial Active Arm Dosing Protocol . 27
Figure 2.2. Allopregnanolone Phase 1 Clinical Trial Consort Statement ................. 32
Figure 2.3. Participant ApoE Genotyping by Dosing Cohort .................................... 38
Figure 2.4. Participant ApoE Genotyping by Sex ....................................................... 38
Figure 3.1. PBMC Reprogramming Protocol ............................................................... 47
Figure 3.2. NSC Directed Differentiation: Dual SMAD inhibition monolayer protocol
................................................................................................................................. 50
Figure 3.3. iPSC induction from lymphocytes using nonviral, integration-free
plasmids ................................................................................................................. 59
Figure 3.4. Reprogramming Colony Count ................................................................. 60
Figure 3.5. Effect of Storage Time on Reprogramming Efficiency ........................... 60
Figure 3.6. Effect of Blood Sample Storage on Reprogramming Efficiency. ........... 61
Figure 3.7. iPSC Immunocytochemistry Tra-1-60 ....................................................... 62
Figure 3.8. Representative iPSC Karyotyping ............................................................ 63
Figure 3.9. NSC Immunocytochemistry ...................................................................... 64
Figure 3.10. NSC baseline proliferation for 22 clinical trial participant cell lines. .. 65
Figure 3.11. NSC mitochondrial respiration profile for 22 participant cell lines.. ... 66
Figure 3.12. Parameters of Mitochondrial Respirational Profile subgrouped by sex
and ApoE genotype ............................................................................................... 68
Figure 4.1. Inter-line NSC Proliferation with Allo treatment (10nM and 100nM). ..... 75
Figure 4.2. Inter-line treatment proliferation data from NSC cell lines classed as
negative, non-, or positive responders ............................................................... 76
Figure 4.3. Association between Baseline Proliferation and Allo response ........... 79
Figure 5.1. Inter-line NSC Mitochondrial Respiration: Maximal Respiration and
Reserve Respirational Capacity with Allo treatment (10nM and 100nM). ........ 85
Figure 5.2. Inter-line Treatment Mitochondrial Respiration from NSC cell lines
classed as negative, non-, or positive responders ............................................ 89
Figure 5.3. Association between Baseline and Vehicle Mitochondrial Function and
Allo Response. ....................................................................................................... 93
Figure 5.4. Effect of Allo on extracellular amyloid-b levels ....................................... 94
Figure 6.1. Correlations between NSC Proliferation and Left Hippocampal Volume
............................................................................................................................... 103
Figure 6.2. Correlations between NSC Mitochondrial Function and Left
Hippocampal Volume .......................................................................................... 107
Figure 6.3. Spearman Correlations Between NSC Mitochondrial Function and Left
Hippocampal Volume by Dosing Cohort. .......................................................... 110
Figure 6.4. Spearman Correlations Between NSC Mitochondrial Function and Left
Hippocampal Volume by ApoE Genotype. ........................................................ 112
Figure 6.5. Spearman Correlation Spearman Correlation between NSC
Mitochondrial Function and Left Hippocampal Volume by Sex ...................... 114
xi
Figure 7.1. Potential Errors in NSC Model ................................................................ 122
Figure 7.2. Correlational Analysis Interpretation ..................................................... 123
Figure 7.3. Recommended Clinical Trial Design ...................................................... 126
Figure 7.4. Alternative Clinical Trial Design ............................................................. 128

xii
Abstract

Introduction. Alzheimer’s disease (AD) is a national and global epidemic with complex
pathoetiology including compromised brain metabolic activity and decreased
regenerative capacity. Allopregnanolone (Allo) is an investigational neuroregenerative
therapeutic, currently in Phase 1b clinical trial for AD (NCT02221622,
https://clinicaltrials.gov/ct2/show/NCT02221622?term=NCT02221622&rank=1). In rodent
preclinical models, Allo promotes neural stem cell (NSC) proliferation and neural
differentiation, improves mitochondrial function, and reduces amyloid-beta pathology. To
develop biomarkers predictive of a clinical regenerative response to Allo, this thesis
investigates the impact of Allo on human induced pluripotent stem cells (iPSCs) and
iPSC-derived NSCs as well as the feasibility of implementing such an approach in a
large clinical trial setting. To evaluate the translatability of in vitro data from clinical trial
participant cell lines, this data was then correlated with clinical outcomes in the same
study participants.
Methods. Peripheral blood mononuclear cells were isolated from whole blood samples
of Allo clinical trial participants and t-lymphocytes reprogrammed to iPSCs via a non-
integrating, non-viral, episomal plasmid method. Using dual inhibition of SMAD signaling,
iPSCs were differentiated to NSCs. Assays were conducted to assess NSC proliferation
and mitochondrial function by flow cytometry and live cell extracellular flux respectively.
Resulting data were analyzed to determine the regenerative and bioenergetics effect of
Allo on clinical trial participant iPSC-NSCs. In vitro data from the first two dosing cohorts
was correlated with volumetric MRI of the hippocampus and cognitive testing clinical
outcomes.
xiii
Results. In vitro data indicated that Allo treatment was associated with an increase in
proliferation in 3 out of the 19 participant NSC lines, while contrary to previous in vivo
and in vitro studies, in 6 participants, treatment was associated with a decrease in NSC
proliferation. However, the mitochondrial function parameters maximal respiration and
spare respirational capacity were significantly increased in 8 NSC lines with Allo
treatment (mean increase of 30% and 40% versus vehicle respectively), and maximal
respiration was nearly significant in an additional 4 participants (mean increase of 20%
versus vehicle). These participant cell lines have been labeled ‘responders’, while those
that did not have increased NSC proliferation or metabolic capacity are ‘non-
responders’. These data demonstrate that Allo promotes regeneration and mitochondrial
function of iPSC-derived NSCs in a subset of clinical trial participants. Analyses with
clinical data indicated that mitochondrial parameters have positive correlation with
participants’ change in left hippocampal volume after 12 weeks of Allo treatment.
Correlations increase when limited to the subgroup of participants with the ApoE3/4
genotype, indicating that an iPSC-approach may be a better model of their therapeutic
response to Allo compared to modeling the ApoE 3/3 genotype.
Conclusions. Allo treatment of iPSC-derived NSCs resulted in an increase in
mitochondrial function that positively correlates with a clinical increase in left
hippocampal volume after 12-weeks of Allo infusions. This study is not only one of the
largest investigations linking an in vitro iPSC-derivative response with clinical changes
but also serve to translate previous discoveries from mouse models into a human AD
population. Future work will advance both our basic understanding of Allo in AD by
investigating the mechanisms for improved mitochondrial functioning, and also
expanding and refining our understanding of in vitro mitochondrial function as a
xiv
component of a larger drug discovery program. By proposing mitochondrial respiration
as possible novel method of patient stratification, these data form the foundation for
developing the first biomarker of regenerative potential in brain to determine and monitor
response to Allo as a regenerative therapeutic.

1
Chapter 1: Background

Alzheimer’s Disease: Background
In the United States, Alzheimer’s disease (AD) is the most common cause of dementia,
affecting an estimated 5.5 million people. With ~500,000 new cases of late onset AD to
be diagnosed in 2017 alone, it is one of the few major causes of death to have an
increasing incidence
1
. A complex, progressive disease, AD is often idiopathic, but in
some cases in known to be a result of genetic mutations, increased genetic risk, and
environmental factors
1-3
. Having the ApoE4 genotype is the most common genetic
cause of increased AD risk and approximately 27% of the US population has at least
one allele
1
. To date, there are four FDA approved medications indicated for AD, the most
recent, memantine, being approved in 2003
4
. Only about 50% of treated patients see
improvement and even then it is usually of limited magnitude and duration. These
medications do not prevent, slow, or cure the disease
5
.

Alzheimer’s Disease: Drug Development
In AD drug development, the figures are disheartening. From 2002-2012, there were 83
Phase III clinical trials of which, only one (memantine) was successful, resulting in an
overall failure rate of 99.6%
6
. Just this week, an additional high profile failure was
announced by Axovant, when their selective serotonin antagonist, intepirdine, failed to
meet primary efficacy endpoints
7
. There are a number of challenges to be overcome in
order to maximize the chances of successful development in the future. These include
2
poor translation from preclinical models to the clinic, lack of early diagnosis, appropriate
clinical trial outcome measures, and the need for biomarkers
8
. As an objective tool in the
drug development toolbox, a biomarker could be used to monitor therapeutic
interventions. In clinical research, biomarkers can be used to stratify patients, enriching
the recruited patient population for those most likely to demonstrate efficacy, thereby
improving the response rate and shortening trial duration
8
. Inappropriate patient
selection has been cited among the causes of more than one high-profile therapeutic
failure, as evidenced in Pfizer’s phase III clinical trial of bapineuzumab
9
. This β-amyloid
(Aβ) antibody failed to improve clinical outcomes in mild to moderate AD, and post-hoc
analyses showed that in non-ApoE4 carriers, 36% of participants did not even have Aβ
plaques at baseline
10
.

Allopregnanolone: A Novel Regenerative Therapeutic
Allopregnanolone (Allo), a neurosteroid, promises a novel therapeutic approach,
targeting the endogenous regenerative capacity of the brain to sustain neurological
function and to prevent, delay, or treat neurodegenerative diseases
11,12
. Allo is an
endogenous molecule, which when given exogenously, penetrates the blood-brain
barrier to bind to the GABA
A
receptor. Along with GABA
A
receptors, neural stem cells
(NSCs) also express the SLC12A2 transporter, a membrane protein regulating chloride
and sodium transport and is responsible for the high intracellular Cl
-
concentration. Upon
binding in the a-subunit transmembrane domains, Allo activates the GABA
A
complex,
resulting in Cl
-
efflux and subsequent depolarization of the NSC. This in turn activates
the voltage-dependent L-Type Ca
2+
channel. In response, a signaling cascade is initiated
3
which actives cyclin AMP-responsive element-binding protein 1 and increased
transcription of pro-mitotic with concurrent down-word dultiple regulation of anti-mitotic
genes
13
(Figure 1.1.). Allo demonstrates greatest potentiation when bound to GABA
A

receptors containing a d subunit
14
.

Due to the low aqueous solubility of Allo (logP 5.042), clinical formulations have been
manufactured using albumin as a carrier or a cyclodextrin solubilizing agent. In pre-
clinical in vivo studies, the amount of the later has proven critical as the complexation
ratio of Allo with a cyclodextrin, such as sulfabutylether-beta-cyclodextrin (SBECD), is a
major determinant of drug release into the blood and brain. Using acute motor
Allo
From: Brinton, R. D. (2013) Nat. Rev. Endocrinol.
Figure 1.1. Allopregnanolone Neurogenic mechanims of action in NSCs. 1) Binding
initiates potentiation of GABA complex; 2) Channel activation results in Cl- efflux and
subsequent membrane depolarization; 3) Depolarization activates the L-type Ca2+
channel; 4) Increased intracellular Ca2+ activates signalling cascade, including CREB1
phosphorylation and up-regulation of cell cycle genes required for transition from G0 to S
and M phases of the cell cycle and down-regulation of cell division repressors; 5) NSC
proliferation in the subgranular zone and oligodendrocyte precursors in white matter.
4
impairment as a surrogate marker for CNS target engagement, rats dosed with a
SBECD:Allo formulation with a molar ratio of 5.89, there was rapid and prolonged
sedation to a maximally tolerated dose (MTD) of 8mg/kg. When that molar ratio is
increased to 23.56 or decreased to 1.47, the rate of Allo release to the brain is reduced,
as indicated by a lack of sedation
15
.

Allopregnanolone: Pre-Clinical efficacy
Previously, the Brinton research group has demonstrated that Allo promotes proliferation
of rodent and human neural progenitor cells in vitro
16
. In vivo, we’ve demonstrated that
in the triple transgenic Alzheimer’s disease (3xTgAD) mouse, chronic Allo treatment
increases neurogenesis within the hippocampus and restores learning and memory
function to normal
16-18
. As an allosteric modulator of the GABA
A
receptor, Allo induces
chloride efflux from the NSC and a concomitant influx of calcium, which together lead to
transcription of pro-mitotic and downregulation of anti-mitotic genes
11
. Increasing
evidence has linked mitochondrial dysfunction with aging and multiple
neurodegenerative disorders, such as AD
19-29
. The Brinton research group has
previously demonstrated that mitochondrial bioenergetic deficits precede AD pathology
in the female 3xTgAD mouse model
28
suggesting mitochondrial bioenergetic deficiency
may cause AD pathogenesis
30
. Clinically, AD pathology is accompanied by a decrease
in the expression and activity of enzymes involved in mitochondrial bioenergetics, which
can be expected to lead to compromised electron transport chain complex activity,
abnormal aerobic glycolysis, and reduced ATP synthesis
20
. Unpublished data from the
Brinton research group demonstrates that Allo treatment also improves mitochondrial
5
function in 3xTgAD mice, restoring it to levels equivalent to non-transgenic, healthy
mice. Additionally, Allo increases markers of white matter regeneration and cholesterol
homeostasis while simultaneously reducing Aβ burden and microglia inflammatory
markers
17
. The neurotoxic effects of Aβ on mitochondria are well understood and known
to propagate the degenerative process
31
. When Aβ accumulates in the mitochondria it
blocks the transport of mitochondria proteins, uncouples the electron transport chain,
increases reactive oxygen species (ROS) production, impairs mitochondrial dynamics
and ATP production, and decreases the mitochondrial membrane potential, resulting in
mitochondrial damage and eventually cell death
30,31
. The γ-secretase complex is present
in cellular compartments other than the plasma membrane, including the endoplasmic
reticulum (ER), ER-mitochondria associated membranes, and the mitochondria
32
. The
complex appears to be on the outer mitochondrial membrane, with the same orientation
as seen in the plasma membrane, where it is active and can cleave amyloid precursor
protein (APP)
33,34
. The effect(s) of familial AD (fAD) mutations on mitochondria-related
events are as yet unknown
35
. Prior to cell death, the decline in mitochondrial
bioenergetics and increase in oxidative stress induces a hypermetabolic state as the
mitochondria attempts to meet the cell’s energy requirements
36
. This drives the
production of additional Aβ, creating a vicious cycle in which excessive Aβ accumulation
and sustained mitochondrial dysfunction synergize to active a cascade of
neurodegenerative pathways
28,37
. Oxidative damage due to increased ROS levels
causes further damage to the mitochondrial by inducing mitochondrial DNA mutations,
altering APP cleavage down the amyloidogenic pathway, enhancing lipid peroxidation,
and activating mitophagy
30
.

6
Allopregnanolone: Clinical Safety
As an endogenous molecule, there is a substantial body of work on the inherent safety
and disease implications of Allo. It has been given to a wide-variety of populations in a
number of clinical studies, the results of which were comprehensively reviewed in Irwin,
Solinsky & Brinton
38
. In brief, while pre-clinical studies have utilized a range of routes of
administration, clinically, Allo has been given exclusively by intravenous (IV)
administration. This has allowed for the characterization of well-defined pharmacokinetic
parameters and corresponding pharmacodynamic responses within the given dosage
ranges. At dosage ranges of 0.05mg/kg-0.09mg/kg (which at a 70kg body weight is a
3.5mg-6.3mg dose), administered over one to ninety minutes, mean Cmax values in
healthy study participants fell between 51nM/L and 150 nM/L
39-43
. In studies where self-
rated sedation was recorded, 1/3

– 2/3rds of participants reported feelings of mild
intoxication or sleepiness
40-43
. Even with these reports of sedation, overall the drug is
well-tolerated with limited reports of flushing and nausea
42
. This positive clinical safety
data, combined with pre-clinical safety and efficacy, supported further development of
Allo for use in neurodegenerative and neuropsychiatric conditions
38
.

Allopregnanolone: Proof-of-Concept
To date, clinical trials of Allo are ongoing in Alzheimer’s Disease (See Chapter 2), have
been completed for traumatic brain injury
44
, super refractory status epilepticus (SRSE)
45-
47
, post-partum depression
48,49
, essential tremor
50
, and fragile-X associated tremor/ataxia
syndrome (FXTAS)
51
, and future studies are planned for Parkinson’s Disease (PD),
7
multiple sclerosis (MS), Neimann- Pick Type C, and chemotherapy-induced neuropathy.
It is important for researchers designing these studies to consider that the plasma levels
required for therapeutic effect and the dosage regimen for same will depend on the
disease target
38
. Overall, the pre-clinical data
from the disease states mentioned here indicates that a constant infusion of Allo is not
biologically relevant in neurological disorders and therefore should not be a therapeutic
option. However, in epileptic seizures and traumatic injury, continuous infusion
paradigms have shown significant benefit (Figure 1.2.)
38
. The clinical trial protocols for
AD and FXTAS have utilized a once-per-week infusion, with both studies aiming to
Figure 1.2. Sedation as a Biomarker of Target Engagement. The sedative
component of GABA
A
receptor activation in the brain can be used a biomarker outcome
of Allo delivery and tolerability. Through activation of the regenerative system in
intermittent pulses, Allo is beneficial for many neurological disorders. In status
epilepticus and traumatic brain injury it is important to deliver sedative doses of
continuous Allo to protect the brain from further excitotoxicity.
No
sedation
Mild-
sedation
Deep-
sedation
• Alzheimer’s disease
• Parkinson’s disease
• Diabetic neuropathy
• Catamenial epilepsy

• Multiple sclerosis
• Neimann Pick type C
• Sleep aid
• Status epilepticus

• Traumatic brain injury
• General anesthetic

States of Sedation
Beta wave
13-30 Hz

Alpha wave
8-13 Hz

Theta wave
4-8 Hz

Delta wave
0-4 Hz

Awake

Moderate
sedation

General
anesthesia

Deep
anesthesia
0-4 Hz
Electroencephalogram
Neurological Disorder
Allopregnanolone Dose
Dosing Regimen
Intermittent exposure
Continuous exposure
Continuous exposure
Intermittent exposure
Regenerative potential
Protection from excitotoxicity Regeneration
Adapted from: Irwin, Solinsky, & Brinton (2014) Front Cell. Neurosci.
8
establish the maximally tolerated dose in their respective populations. The FXTAS study
had limited pharmacokinetic (PK) analysis but indicated that a 6mg IV infusion over 30
minutes resulted in an Allo Cmax of 120nmol/L with no signs of sedation
52
. The final
results of the AD study have not yet been finalized (See Chapter 2), but analysis
between dosing cohorts indicates that a 4mg IV infusion over 30 minutes results in a
similar Cmax of 125.48nM ±19.13, also with no signs of sedation
53
. Clinical usage in
SRSE has opted for an infusion paradigm in order to suppress cortical hyperexcitability
resistant to benzodiazepines and barbiturates. Resistance has been attributed to
internalization and desensitization of synaptic GABA
A
receptors, but as Allo can bind
intracellularly, makes it a potential therapeutic alternative
46
. When Allo was utilized in
two female paediatric cases and two males in their 20’s with SRSE, it was infused to a
steady state plasma level of 150nM/L, the maximal dose allowed by the US Food and
Drug Administration (FDA) (as described in Chapter 2), although in one case, there was
a reported peak of 404.7nmol at 24 hours
45,46
. This steady state level was maintained for
72-96 hours before Allo was withdrawn. During the Allo treatment period, in all cases,
clinicians were able to successfully withdraw pentobarbital and/or midazolam without
seizure recurrence, serving as a positive proof-of-concept for this neurological indication.
No adverse drug effects were reported in these cases and it should be noted that as
these patients were sedated on a complicated regimen of antiepileptics, general
anesthetics, and benzodiazepines etc, sedation is not a definition of therapeutic
intolerance in this patient population
45,46
. In the FXTAS study, although of limited size
(N=6), additional proof-of-concept signs of efficacy were noted in participants with
preserved hippocampi and corpus collosum, who had improved executive function,
learning, and memory
51
. Data in an AD population will be analyzed at the conclusion of
9
the ongoing clinical trial, and some initial analyses from the first two cohorts of
participants is presented and discussed here (Chapters 3-6).

Induced Pluripotent Stem Cells: Background
In 2007, Takahashi and Yamanaka described the first human induced pluripotent stem
cells (iPSCs)
5
. With this process, somatic cells transduced with 4 transcription factors
(Oct3/4, Sox2, Klf4, and c-Myc) can be reprogrammed back to a pluripotent state and
differentiated into a myriad of cells types from the three germ layers
54
. Used to model a
variety of neurodegenerative diseases, iPSC research has focused on monogenic
disease forms, validating that iPSC derivatives recapitulate disease
55-57
. By using familial
diseases where diagnosis is certain combined with isogenic controls (when patient DNA
is genetically engineered using CRISPR/Cas-9 or similar, to the healthy, wildtype
sequence), the function of a gene can be linked to the in vitro phenotype
58
. While
isogenic controls are not yet ubiquitous, the overall iPSC approach to fAD has proven
successful with a number of studies demonstrating altered cellular physiology in iPSC-
derived NSC, neurons, and astrocytes
55,59-68
. Table 1.1 presents a comprehensive
compilation of studies completed using iPSC-derivatives from AD donors (or health
donors genetically engineered to have AD inducing mutations). The models described
cover 12 x presenilin-1 (PSEN1) mutations, 1 x presenilin-2 mutation, 7 x APP
mutations, 6 x risk-variant genes, and 5 examples of sporadic AD (sAD). An overall
limitation to the work done in AD using iPSCs is the limited number of donors evaluated
in each study, with many using lines from the same donors (Table 1.2.). With few
exceptions, the fAD models describe a consistent AD phenotype of altered Aβ
10
generation, resulting in an increased Aβ42/40 ratio. The generation of phosphorylated
tau is also reported, albeit at a lower frequency, but may be due to age of the cell culture
studied, and require an ‘aged’ culture
69
. Studies using iPSC-neural cells from ApoE3/4 or
4/4 donors and other sAD donors indicate that in these cases, the presence of a
pathological phenotype is sporadic. This may relate to the age of symptom onset, as in
one study only those sAD cases classed as ‘early onset’ (donors 35-45 years of age at
onset) displayed a phenotype
63
. Unfortunately, other studies including sAD donors do
not indicate the age of clinical onset. These and all mutations associated with AD and
other dementias are described on the Alzforum Mutations database and includes
associated clinical and neuropathological features as well as reported functional
effects
70
.

While nearly all the fAD iPSC studies have examined Aβ production, only one has
reported on aspects of a metabolic phenotype. Researchers describe increased
mitochondrial dysfunction compared to isogenic neurons, as measured by mitochondrial
membrane potential
67
. One study on familial PD has looked directly at the mitochondrial
respiration of iPSC-derived neural cells, in a method similar to that proposed for this
study. While not the same disease or mutations as the ones proposed for this body of
research, the data demonstrated that iPSC-derived neural cells from three different
familial PD donors had three different oxidative phosphorylation phenotypes when
compared to that of a healthy subject
71
. This highlights the possible variability that could
be anticipated in an equivalent AD study.

11
In total, 50% of the AD iPSC researchers detail the use of a therapeutic against the
observed phenotype, generally with success. These examples of positive in vitro
response demonstrate the potential application for iPSC-based therapeutic screening
and precision medicine. The experimental therapeutics were usually directed against Ab
generation (g-secretase and b-secretase inhibitors, Ab antibodies) and none targeted the
brain’s endogenous regenerative potential.
12

Mutation Cell Type(s) Phenotype Therapeutic Intervention? Reference
PSEN1
3D neural organoids (Raja) No effect of mutation on A! or pTau (Raja) Raja et al, 2016
69
Neurons (Brownjohn) Selamectin increased ratio of A!38/A!42 (Brownjohn) Brownjohn et al, 2017
73
NPCs (Jones, Sproul, Liu)
No difference in proliferation with mutation (Jones)
Increased A!42/A!40 ratio (Sproul, Liu)
Neurons (Sproul, Liu) Increased A!42:A!40 ratio (Sproul, Liu)
Astrocytes (Jones)
Reduced morphological heterogeneity, increased cellular atrophy
and release of soluble inflammatory mediators (Jones).
3D neural organoids (Raja) Increased A! particles and pTau (Raja)
NPCs (Yang, Sproul, Liu)
Neuronal differentiation: premature neuronal differentiation,
decreased proliferation, increased apoptosis and neurite
framentation (Yang)
Increased A!42/A!40 ratio (Sproul, Liu)
GSI (DAPT) blocked A! production (Sproul)
GSI (Semagacestat) lowered A! levels (ratio and total);
GSM IV lowered A! ratio, not total A! (Liu)
Neurons (Yang, Armijo, Sproul, Duan,
Mahairaki, Liu, Yagi)
Increased A!42 (Yang, Armijo)
Increased A!42/A!40 ratio (Yang, Armijo, Sproul, Mahairaki, Liu,
Yagi)
Increased pTau (Yang)
Increased susceptibility to A! toxicity, as measured by cell viability
and LDH release (Armijo)
No significant effect of effect of mutation on A!42:A!40 ratio,
(Duan)
No abnormal Tau (Yagi)
GSI (DAPT) blocked A! production (Sproul)
No effect of GSI (Compound E) (Duan)
GSI (Semagacestat) lowered A! levels (ratio and total),GSM IV
lowered A! ratio, not total A! (Liu)
GSI (compound E) and GSM (Compound W) lowered A!40, A!42,
and the A!40:A!42 ratio. (Yagi)
NPCs (Woodruff, 2013)
Increased A!42:A!40 ratio. No effect on pTau or total Tau.
(woodruff)
Neurons (Woodruff 2013, 2016)
Increased A!42:A!40 ratio. No effect on pTau or total Tau.
(woodruff 2013)
Increased soma APP, decreased axon APP; decreased
endocytosis and transcytosis of APP and LDL (Woodruff 2016)
Low dose GSI (Compound E): increased A!42:A!40 ratio in
heterozygous neurons.
High dose inhibited A! production. (woodruff)
BSM IV and GSI (Compound E) rescued LDL endocytosis and
transcytosis (Woodruff).
A79V Neurons Increased A!42:40 ratio
Treatment with high dose NSAIDS (Indomethacin, ibuprofen,
diclofenac, flurbiprofen) reduced the ratio. However, at a
therapeutically relevant dose, no alteration in A! production was
observed.
Mertens et al, 2013
66
Pires et al., 2016
77
NSCs Decreased proliferation. Decreased A!40.
Neurons Decreased A!40
NSCs Decreased NSC proliferation. Increased A!42:A!40 ratio.
Neurons Increased A!42:A!40 ratio.
NSCs Increased A!42:A!40 ratio
Neurons Increased A!42:40 ratio
NPCs
Neuronal differentiation: premature neuronal differentiation,
decreased proliferation, increased apoptosis and neurite
framentation.
Neurons Increased A!42, A!42:A!40 ratio, and pTau (Yang)
M146I
GSI (DAPT) blocked A! production (Sproul)
GSI (Semagacestat) lowered A! levels (ratio and total), GSM IV
lowered A! ratio, not total A! (Liu)
Jones et al, 2017
75
Sproul et al, 2014
83
Liu et al, 2014
59
M146L
A246E
Raja et al 2016
69
Yang et al, 2017
87
Armijo et al, 2017
72
Sproul et al, 2014
83
Duan et al, 2014
63
Mahairaki et al, 2014
76
Liu et al, 2014
59
Yagi et al, 2011
86
Woodruff et al , 2013
84
Woodruff et al., 2016
85
(delta)E9*
D385N* Koch et al., 2012
65
NSAIDs (ibuprofen, indomethacin) had no effect on A! production Koch et al., 2012
65
L166P*
H163R
GSI (Semagacestat) lowered A! levels (ratio and total);
GSM IV lowered A! ratio, not total A!
Liu et al, 2014
59
S169del Yang et al, 2017
87
13

Mutation Cell Type(s) Phenotype Therapeutic Intervention? Reference
PSEN1 (continued…)
P117L* Neurons
Mitochondrial dysfunction, DNA damage, high intracellular A!,
and oxidative stress.
Oka et al., 2016
67
L150P iPSC Not evaluated
Poon et al., 2016
78
L282F iPSC Not evaluated Poon et al., 2016 (b)
79
PSEN2
N141I Neurons Increased A!42:A!40 ratio. No abnormal tau pathology detected
GSI (compound E) and GSM (Compound W) lowered A!40, A!42,
and the A!42:A!40 ratio.
Yagi et al., 2011
86
APP
3D neural organoids (Raja)
Increased A!40, A!42, and A!42:A!40 ratio. Increased pTau.
Abnormal endosome morphology and recycling (Raja)
GSI (Compound E), BSI (BACE-1) decreased AB!and pTau (Raja)
Neurons (Brownjohn, Israel)
Increased A!40, and pTau/total Tau ratio. Increased number and
size of early endosomes in neural soma (Israel)
Selamectin increased ratio of A!38:A!42 (Brownjohn)
Treatment with GSI (Compound E; DAPT) and BSI (BetaSi-II;
OM99-2) reduced A!40 to control levels. BSIs also rescued
pTau/tTau ratio.
V717F* Neurons Decreased LDL uptake BSI IV rescued LDL endocytosis and transcytosis. Woodruff et al., 2016
85
V717I Neurons (Brownjohn, Muratore) Increased A!42:A!40 ratio, APP, and pTau. (Muratore)
Selamectin increased ratio of A!38:A!42 (Brownjohn)
AB antibody treatment (3D6, AW7) reduced total tau.
GSI (DAPT) decreased A!. (Muratore)
Brownjohn et al., 2017
73
Muratore et al., 2014
64
V717L Neurons Increased extracellular A!42:40 ratio BSI IV reduced A!40 and A!42 Kondo et al., 2013
61
KM670/671NL* Neurons Decreased LDL uptake BSI IV had no effect on LDL uptake Woodruff et al., 2016
85
E693(del) Neurons
Reduced A!40 and A!42 production. Increased intracelluar A!
and resulted in ER and oxidative stress. Neuronal survival rate
decreased
BSI IV: reduced extracellular A!40 and A!42.
BSI IV and DHA: decreased oxidative stress. DHA rescued neuron
survival rate.
Kondo et al., 2013
61
K724N Neurons Increased A!42:A!40 ratio
Treatment with high dose NSAIDS (Indomethacin, ibuprofen,
diclofenac, flurbiprofen) reduced A!42:!40 ratio. However, at a
therpaeutically relevant dose, no alteration in A! production was
observed.
Mertens et al., 2013
66
Risk Variants
Trem2 R47H iPSC Not evaluated
Schröter et al., 2015
80
Schröter et al., 2016
81
CR-1 iPSC Not evaluated Schröter et al., 2016 (b)
82
NSCSs Variable levels of SORL1
Neurons Variable levels of SORL1
Neurons with protective alleles treated with BDNF had reduced
A!40.
Neurons with risk alleles showed no response with BDNF
treatment.
All neurons treated with cAMP, regardless of genotype, had
reduced AB40.
Raja et al., 2016
69
Brownjohn et al., 2017
73
Israel et al., 2012
55
Duplication
SORL1 (risk & protective variants) Young et al., 2015
88
14

Table 1.1. Synopsis of AD studies utilizing iPSC-derived neural cells. Studies categorized by AD-associated mutation
investigated (if present). GSM: gamma-secretase modulator; GSI: gamma-secretase inhibitor; *Genetically engineered.
Mutation Cell Type(s) Phenotype Therapeutic Intervention? Reference
Risk Variants (continued…)
ApoE 3/3 Neurons
1 of 2 ApoE3/3 sAD lines had increased A!40 and pTau/tTau. This
line also an increased number and size of early endosomes in
neural soma.
Treatment with GSI (Compound E; DAPT) and BSI (BetaSi-II;
OM99-2) reduced A!40 to control levels. BSI also rescued
pTau/tTau ratio.
Israel et al., 2012
55
ApoE 3/4 Neurons
2 of 3 ApoE3/4 sAD lines had increased A!42:A!42 ratio. The
same two lines showed increased vulnerability to glutamate-
mediated cell death.
GSI (DAPT) treatment reduced A!40 secretion in lines with
phenotype.
Duan et al., 2014
63
NPCs (Jones)
NPCs: No difference in proliferation with mutation.
Astrocytes (Jones, Zhao)
Reduced morphological heterogeneity, increased cellular atrophy
and release of soluble inflammatory mediators. (Jones)
Secreted less lipidated lipoprotein particles and were less
supportive of neuronal viability in co-culture, compared to
ApoE3/3. (Zhao)
Sporadic AD (risk variants unknown)
A! levels were not increased over healthy controls. However, sAD
neurons were more sensitive to A! toxicity, as measured by cell
viability and LDH release (Armijo).
Armijo et al., 2017
72
TRPC6 mRNA levels decreased, correlated with peripheral blood
analyses (Lu).
Lu et al., 2017
68
Increased transcriptome expression of pTau vs H9 neurons
(Hossini)
GSI downregulated pTau (Hossini). Hossini et al., 2015
74
Extracellular A! levels were not increased over controls. However
1 of the 2 sAD lines had increased intracellular A!, which resulted
in ER and oxidative stress (Kondo).
BSI IV: reduced extracellular A!40 and A!42 and decreased
oxidative stress (Kondo).
Kondo et al., 2013
61
iPSC Not evaluated Zhang et al., 2016
89
Jones et al., 2017
75
Zhao et al., 2017
90
ApoE 4/4
Neurons (Armijo, Lu, Hossini, Kondo, Zhang)
15

Reference Number of AD Donors:
Armijo et al., 2017
72
2
Brownjohn et al, 2017
73
3
Duan et al., 2014
63
5
Hossini et al., 2015
74
1
Israel et al., 2012
55
4
Jones et al, 2017
75
2
Koch et al., 2012
65
1
Kondo et al., 2013
61
7
Liu et al, 2014
59
4
Mahairaki et al, 2014
76
3
Mertens et al, 2013
66
4
Muratore et al., 2014
64
2
Pires et al., 2016
77
1
Poon et al., 2016
78
1
Poon et al., 2016 (b)
79
1
Raja et al, 2016.
69
4
Schröter et al., 2015
80
1
Schröter et al., 2016
81
1
Schröter et al., 2016 (b)
82
2
Sproul et al, 2014
83
3
Woodruff et al, 2013
84
1
Woodruff et al., 2016
85
1
Yagi et al., 2011
86
2
Yang et al, 2017
87
2
Young et al., 2015
88
7
Zhang et al., 2016
89
1
Zhao et al., 2017
90
6

Table 1.2. Number of AD donor cell lines evaluated in previous iPSC-based AD
studies.

16

iPSC Application: Cell Transplantation
When thinking about the intersection of iPSC with regenerative medicine, the primary
instinct is to consider the possibility for autologous cell and organ transplants. As a novel
cell, and one-day possible tissue source, iPSC products offer a number of advantages
over those using human embryonic stem cells (hESC). Besides overcoming the ethical
concerns regarding cell source, iPSC products could be made for each patient using
their own somatic cells, i.e. the epitome of precision medicine, negating the need for
immunosuppression and risks of rejection. However, there are a number of critical
considerations and hurdles that will hamper widespread, and quick, adoption of such
clinical advances. These include the well-known problem with lack of consistency and
reproducibility in differentiation protocols. The regulatory landscape with regards to such
protocols as well as characterization and manufacturing, is still developing as the FDA,
European Medicines Agency (EMA), and Pharmaceuticals and Medical Devices Agency
in Japan adapt to the rapidly advancing field. The risk for tumourgenesis, be it benign or
malignant, is still relevant with iPSC-derived products and more chronic studies
investigating this concern specifically are required. One final, but increasingly relevant
concern, is that of cost. Increasing costs of healthcare are at the forefront on the national
stage and iPSC-products are not cheap. It’s been estimated that it can cost up to
$100,000 to develop a characterized, validated, line that meets current Good
Manufacturing Practice (cGMP) standards, and this increases up to $800,000 for a
clinically-useable iPSC-derived tissue product
91
. Payers in all heath care systems should
17
be engaged from the beginning of the development process to ensure that any approved
products are available to the relevant patient populations, ensuring distributive justice.

To date there has been only one trial which has used iPSC-based cell transplants
utilizing autologous iPSC-derived retinal pigment epithelium for age-related macular
degeneration
92
. Due to changes in Japanese regulations and concerns about genetic
mutations, the trial was put on hold and recently resumed using allogenic iPSCs
92-96
.
While this is a viable clinical option in countries with a relatively homogeneous
population (in Japan, 50 iPSC lines would haplotype match to 91% of the population and
150 iPSC lines would cover 93% of the United Kingdom), this proves more difficult in
more the diverse North American population (where screening could match up to 78% of
European descendants but only 45% of African Americans)
91
. A number of hESC
transplant studies have been conducted and are reviewed by Kimbrel & Lanza
93
. Along
with a comprehensive overview of the history of clinical hESC transplantation, they
describe clinical trials for macular degeneration, heart failure, type-1 diabetes, spinal
cord injury, and PD
93
.

As an alternative to direct cell replacement, some studies are investigating the
supporting effects of iPSC-derived neural cells. That is, rather than replacing neurons
lost in neurodegenerative disease, in models of AD and amyotrophic lateral sclerosis
(ALS) macrophages
97
and NSCs
98
are infused and these supportive cells break down Ab
and produce neuroprotective factors respectively, conserving the remaining neurons.

18
iPSC Application: Precision Medicine
Coined ‘macromedicine’, iPSC technology has been proposed since its advent as a
method of patient stratification based on cellular and molecular analyses of clinical trial
participants or other patient cohorts
99
(Figure 1.3.). The following case studies serve to
illustrate successful uses of iPSC-derivatives to advance precision medicine in familial
disease, idiopathic toxicity, validation of off-label indications, patient stratification, and
bench/clinical correlation.

Studies across the spectrum of disease have investigated ways of using iPSC models to
further the drug development efforts
100,101
. The earliest indicators of the success of these
efforts came in cardiology, where there is great hope for using iPSC-derived
cardiomyocytes (iPS-CMs) to predict molecular pharmacology and cardiactoxicity. When
iPS-CMs, derived from a long QT syndrome family with and without the disease were
treated with mexiletine, the mutation induced Na+ channel inactivation, was corrected in
the proband cell line. The drug/Na+ channel interaction was unique in the proband cell
line compared to those of the parents, giving a pharmacological explanation for the
positive clinical arrhythmia management, albeit with a limited therapeutic index. Overall,
the study supported using iPSC models for therapy optimization
57
.
19

From: Haruhisa Inoue et al. (2014) EMBO
Figure 1.3. Illustration of the iPSC approach applied to clinical trials. In this
component of the Allo development program, our focus is on Phase 1.5, to determine if
Allo ”responder” and “non-responder” cell lines can be identified. This data will then
inform the design of a subsequent Phase 2 clinical trial.

20
Doxorubicin is a mainstay of breast cancer chemotherapy but its dose-dependent
cardiotoxicity seemingly strikes at random with rates of up to 15% of patients
demonstrating clinical signs and 50% with subclinical cardiac damage, which can be
severe enough to warrant heart transplantation
102-104
. In 2016, researchers made
advances towards being able to predict clinical toxicity when they demonstrated that
iPS-CMs derived from cancer patients with clinical doxorubicin cardiotoxicity
recapitulated this adverse event. When outcomes such as cell viability, metabolic
function, and reactive oxygen species production were compared after doxorubicin
treatment to the in vitro response of cells derived from patients who did not react
negatively to doxorubicin treatment, the former were significantly worse. Besides
proposing a potential mechanism for susceptibility to toxicity, the study added to the
growing body of evidence supporting iPSC-derived cells as a suitable platform to identify
and characterize patient susceptibility to toxic events
105
.

Neuroscience has not been left behind. Spinal muscular atrophy (SMA), caused by
functional loss of SMN1, like AD has no cure. The anticonvulant valproic acid has shown
clinical promise, but only in about one-third of patients demonstrate increased SMN2
transcripts. Although of limited size, one study demonstrated that in iPSC-derived
GABAergic neurons from a clinical responder and one non-responder patient, valproic
acid treatment response in vitro is concordant with the donors’ clinical response.
Subsequent determination of a mechanism for non-responsiveness is relevant not only
in and selecting positive therapeutic SMA responders, but possibly other disease
indications treated with valproic acid
106
.

21
The first-line treatment for Manic Type I Bipolar Disorder is lithium, however, as in the
above SMA example, only a subset of patients show a robust clinical response with
strong symptom control. iPSC-neurons from bipolar donors, treated in vitro with lithium,
were utilized in a study investigating possible mechanism for a response, or lack thereof.
Researchers observed that while all donor cell lines demonstrated mitochondrial
abnormalities and hyperexcitability, only neurons derived from clinical responders had
decreased excitability after in vitro lithium treatment. Through this in vitro recapitulating
of clinical outcomes, this study advanced the understanding of both the disease and
mechanisms of therapeutic response
107
.

In applying an iPSC approach to developing a novel therapeutic to treat patients with
inherited erythromelalgia (IEM), researchers investigated to what extent an in vitro
therapeutic response translates clinically. As part of small Phase 2 clinical trial on the
investigational Nav1.7 blocker PF-05089771 (NCT 01769274) Pfizer researchers
clinically treated SCN9A mutation participants, as well as their iPSC derived sensory
neurons. Selective Nav1.7 blockade reduced the neuronal hyperexcitability in all patient
lines. There seemed to be a relationship between a participant’s clinical severity and the
severity of the iPSC-derived neuronal hyperexcitability, although due to small study size
(N=4) the authors warn against assuming cause & effect in this instance
108
. So while this
study serves as a proof-of-concept, the limited participant number hinders positing it as
strong evidence for the use of iPSC-derivatives to serve as clinical trial surrogates.

A novel and distinct use for iPSC based models of patients and their disease is the
approach Kevin Eggan and his team applied in repurposing the antiepileptic drug
22
ezogabine for use in ALS
109
. Recognizing the inherent challenges in validating drug
targets in neurodegenerative genetic mouse models, they derived iPSC motor neurons
from ALS patients and observed reduced neuronal excitability and improved cell survival
after ezogabine treatment. As a Kv7.2/3 potassium channel agonist, the positive effect of
ezogabine provided a novel druggable target researchers were able to validate through
CRISPR-Cas9 gene editing to correct the ALS causative SOD1 mutation. As in AD,
familial ALS is relatively rare, so to ensure the data were relevant, the study was
replicated in a larger, sporadic ALS cohort, firmly establishing the relevance of the target
and drug mechanism in the disease. This iPSC-based data, demonstrating both a novel
target phenotype and a drug response in a widely-relevant patient population, provided
the buy-in regulators needed to advance drug development without requiring any studies
involving ALS animal models. As noted by the drug development team, also critical to
their translational success was the ability to use in vitro assays that had a correlate in
the clinic (in the case of ALS, in vitro motor neuron excitability could be compared to
clinical motor neuron excitability as measured by transcranial magnetic stimulation), and
known, validated, in-patient biomarkers of disease. This rare positioning allowed for the
program of research to advance from discovery to a FDA approved Phase 2 clinical trial
(NCT 02450552) in less than two years
109
.

23
Study Hypothesis
Overall, the above indicates that the iPSC field is just getting started and researchers
are increasingly coming to a better understanding of its potential to advance drug
development. The use of such an approach has been relatively understudied in AD as
yet, where iPSC-neural cells have been used to validate the appropriateness of the
model and initial drug responsiveness to compounds not in clinical development,
focusing on a limited diversity of targets. Critically, this previous work advances our
understanding of the disease and mechanisms of therapeutic response. As the initial
clinical trial in AD comes to completion, the Allo drug development program is maturing
and determining how to more efficiently and effectively marry current clinical approaches
in AD to these promising and exciting advances in cell technology. To this end, we set
out to answer the question, using
iPSCs derived from sAD donors
enrolled in the Allo clinical trial,
can we develop a method of
identify clinical responders and
non-responders? Here, I detail
the results of the ensuing study
which included establishing iPSC
and NSC lines for all clinical trial
participants, phenotyping the
proliferative and mitochondrial
profile of each cell line,
Aim 3: In Vitro – Clinical Correlations
Correlate clinical imaging and cognitive data with in vitro response to determine if
iPSC-derived NSC recapitulate therapeutic response
Aim 2: In Vitro Allopregnanolone
Investigate the effects of in vitro Allopregnanolone treatment on the participant
derived NSCs on both cellular proliferation and mitochondrial function.
Aim 1: Generate iPSC and NSC for AD Clinical Trial Participants
Reprogram somatic cells to iPSC, differentiate to NSC and the characterize the
NSC proliferative & mitochondrial phenotypes
Figure 1.4. Project Flow Diagram
24
determining the in vitro effect of Allo on these phenotypes, and finally, correlating the
effects of Allo with clinical outcomes (Figure 1.4.).
25
Chapter 2: Allopregnanolone Phase 1 Clinical Trial

Overview:
In August 2014, the initial participant was enrolled in the Phase Ia/IIb clinical trial for Allo,
the first proposed regenerative therapeutic for AD
110
. The objective of this clinical trial is
to establish a safe and tolerated dose of Allo that can be advanced into a Phase II
efficacy study. Considering previous pre-clinical and clinical studies as described in
Chapter 1, as well as an established endogenous level of 157nmol/L (50ng/ml) in
pregnant women during the third trimester of pregnancy, the FDA supported the
proposal for a multiple ascending dose study in participants diagnosed with mild
cognitive impairment (MCI) due to AD, and early AD. The preclinical data are detailed
fully in Chapter 1, but overall indicated that a once-per-week treatment regimen at a non-
sedative dose promoted neurogenesis, restored cognitive function, and reduced the Ab
load
17
.

To this end, the clinical trial was designed to meet the following objectives: 1- complete
an ascending dose analysis of Allo over 3 doses and placebo, administered IV once-per-
week for 12 weeks; 2- complete PK analysis of Allo at the start and end of the 12-week
exposure; 3- determine the maximally tolerated (i.e. non-sedative) dose of Allo; and 4-
conduct magnetic resonance imaging (MRI) analyses to determine if the potential effect
of Allo on Ab pathology is associated with amyloid-relating imaging abnormalities
(ARIA), specifically microhemorrhages, as required by the FDA during a 2011 pre-
investigational new drug (IND) application meeting
111,112
.
26

Based on a standard double-blinded, placebo-controlled trial design, 24 participants
were randomized into three cohorts (6:2 allocation ratio of active:placebo) where the
dose was sequentially increased with each cohort. This design was carefully selected to
ensure the safety of the older target population, as well as to evaluate the cumulative
impact of multiple exposures to Allo. Initially, three doses of 2mg, 4mg, and 6mg were
chosen as it had been determined that these would result in plasma levels meeting the
FDA’s requirement to not exceed 156nmol/L (50ng/ml). However, at the completion of
the second cohort (4mg), no signs of sedation had been observed and there was
concern that a MTD would not be reached in the final cohort. In response, a protocol
change was instituted allowing for individual participant dosing. Over the 12-week
exposure period, participants randomized to the active study arm would receive
increasing Allo doses up to their maximum tolerated dose, defined as detectable mild
sedation assessed through participant reports on the Mood Rating Scale and clinical
assessment with the Stanford Sleepiness Scale. The Mood Rating Scale is administered
before the start of infusion and at its completion. The Stanford Sleepiness Scale is
administered immediately prior to the start of the infusion and every 15 minutes
thereafter until the participant returns to their baseline. Starting with 6mg for the first
infusion, the dose would be increased each week over the initial 4-week dosing period
by 4mg, up to a maximum dose of 18mg (see Figure 2.1). If sedation was detected at a
given dose, at the next infusion the dose was reduced to the highest dose previously
administered which did not induce sedation. This dose was continued for the remaining
weekly infusions. The increased dose was supported by the FDA after the submission of
data from a chronic toxicology study in rodents dosed with Allo for 6 months at levels
27
which resulted in plasma levels at least two-fold higher than those anticipated with the
increased dosing strategy. At the end of the final cohort, the Data Safety Monitoring
Board (DSMB) will evaluate all sedation data to determine what the overall MTD is for
both female and male participants. The MTD will have been exceeded if the proportion
of subjects within a cohort that develop dose-limiting sedation is equal to 2/2, 2/3, 2/4,
2/5 and 2/6 subjects.

Clinical Trial Outcomes:
PRIMARY OUTCOMES:
The primary study endpoint was to determine the safety and tolerability of the IV
Allo infusion and to identify the maximally tolerated dose(s) that can be given
repeatedly without significant sedation. To address these safety objectives
investigators evaluated 1- sedation during and after each dose of Allo; 2- the
Start Weekly IV Infusion (6:2 Active:Placebo)
DSMB Meeting
Wk 1 Wk 2 Wk 3 Wk 12
Cohort 2: 4mg
Wk 1 Wk 2 Wk 3 Wk 12
Cohort 1: 2mg
Wk 1 Wk 2 Wk 3 Wk 12
Cohort 3: 6-18mg
* Increase dose only if no significant sedation observed.
If significant sedation observed, lower dose to previously tolerated dose
Wk 1 Wk 2 Wk 3 Wk 4 Wk 5 Wk 6 Wk 12
6mg 10mg 14mg 18mg
Weekly Dosing: Cohort 3
* 18mg 18mg 18mg * *
Figure 2.1. Allopregnanolone Phase 1 Clinical Trial Active Arm Dosing Protocol
28
incidence and severity of treatment emergent adverse events; 3- designated
medical events; 4- clinically important changes in safety assessments (including
vital signs, weight, clinical laboratory tests, electrocardiograms, MRIs, cognitive
and physical, and neurological exams; and 5- ARIA by both MRI and
symptomatic reports.

SECONDARY OUTCOMES:
While the clinical trial is powered for the above safety analyses, there were a
number of secondary, exploratory endpoints, designed to evaluate clinical trends
towards efficacy as well as a diverse variety of potential biomarkers. The
clinically evaluated outcomes included three tests of cognitive function: the
Alzheimer’s Disease Assessment Scale-Cognitive-Plus 13 item scale (ADAS-
Cog 13; administered every 4 weeks), the Montreal Cognitive Assessment
(MOCA; administered every 4 weeks), and the 12-minute CogState Battery
(administered every other week). This panel of cognitive tests allowed for
functional assessment across multiple domains while facilitating biweekly
surveillance of cognition.

MRI was initially included in the study protocol at baseline and end of study in
order for safety evaluations of ARIA, required at the time by the FDA for all
experimental therapeutics that pre-clinically reduced brain Aβ, when
administered for 3 months or more
113
. Exploiting this requirement, the
Alzheimer’s Disease Neuroimaging Initiative 3T MRI protocol was expanded to
encompass evaluations of brain structure (including total brain volume, total
29
ventricular volume, and hippocampal volume), diffusion tensor imaging (DTI),
and the default mode network (DMN). DTI is an evaluation of white matter
integrity while DMN gives a measure of the average intrinsic connectivity for a
given network. These additional measures provide a foundation for designing an
efficacy trial which uses MRI-based measures reflective of disease severity and
drug effects as an approach to predict clinical trial outcomes
112,114-119
. The
overarching goal of these and cognitive analyses is to not only assess the
potential short-term effect of Allo on these outcomes, but to use this data to
inform a Phase 2 proof-of-concept clinical trial as well as the development of
clinical biomarkers of regenerative efficacy.

Additionally, a number of blood samples were collected for laboratory evaluation
of potential blood-based biomarkers. Whole blood samples were collected for
generation of iPSCs (see Chapter 3), ApoE genotyping, RNA, and metabolomics.
Serum was collected for lipid profiling and plasma for lipidomics analysis.

Overall, if successful, these trial outcomes were chosen to provide 1) an
estimated safe and well-tolerated IV dose of Allo, 2) parameter estimates for MRI
biomarkers of regeneration, 3) parameter estimates for cognitive measures of
therapeutic efficacy, 4) the feasibility as well as the accuracy and sensitivity of an
iPSC-based biomarker of therapeutic efficacy, and 5) parameter estimates for
blood-based biomarkers of therapeutic efficacy to advance to a Phase 2 trial of
Allo efficacy.

30
Clinical Study Participants:
To date, the trial has enrolled 24 participants (12 post-menopausal women, 12 men),
who met the following inclusion and exclusion criteria:

INCLUSION CRITERIA:
• 55 years of age or older
• Diagnosed with MCI or early AD
• Residing in the community with a caretaker willing and capable of
accompanying the participant to a majority of clinic visits
• No medical contraindications to participation
• Willingness to comply with study procedures

EXCLUSION CRITERIA:
• Use of benzodiazepines, sedative/hypnotics, anticonvulsants,
antipsychotics, and other drugs that might interact with the GABA
A

receptor complex,
• Seizure disorder, history of stroke, focal brain lesion, traumatic
brain injury, substance abuse, malignancy,
• Clinically significant laboratory or ECG abnormality,
• MRI indicative of any other significant abnormality, including but
not limited to multiple microhemorrhages (>4), edema, evidence of
a single prior hemorrhage or infarct >1 cm3, multiple lacunar
infarcts (>1), evidence of a cerebral contusion, encephalomalacia,
31
aneurysms, vascular malformations, subdural hematoma, or
space occupying lesions (e.g. abscess or tumor),
• Any condition that would contraindicate an MRI such as the
presence of metallic objects in the eyes, skin, heart, or body
hypersensitivity to any of the excipients contained in the study
drug formulation.

At conclusion of the study 18 participants will have received the active drug and 6 will
have received the placebo. Each 8-participant cohort was gender balanced and
randomization was stratified on gender. Recruitment was conducted throughout Los
Angeles county and southern California through the use of radio ads, collaboration with
the USC Alzheimer’s Disease Research Center (ADRC), a clinical recruitment contract
research organization, and physician referrals. Potential participants were pre-screened
over the phone for initial eligibility and then referred on if appropriate to the study
coordinator for additional screening, medical record examination, and eventual
screening appointment at the USC Neurology Clinic. The number of patients who pre-
screened, screened, randomized, and completed the study, can be seen in Figure 2.2. A
complete description of the participants’ demographics, including age, sex, ethnicity,
race, baseline mini-mental state exam (MMSE), years of education, and ApoE status
can be seen in Table 2.1.

32

Assessed for eligibility (n= 264) as of
May 30, 2017
· KNX radio – 175
· KUSC radio – 29
· WAVE radio – 4
· KOST radio – 1
· WCCT – 11
· USC Memory and Aging Center – 40
· Other – 4

Excluded (n= 243)
¨ Not meeting inclusion criteria (n= 180)
¨ Declined to participate (n= 38)
¨ Other reasons (n= 25)
¨ Analysed (n=12)
¨ Excluded from analysis (n=5)
Ø Cohort 3 in progress
¨ Lost to follow-up (n= 0)
¨ Discontinued intervention (n= 0)
¨ Received allocated intervention (n=17)
Ø Cohort 1 ALLO 2mg (n=6)
Ø Cohort 2 ALLO 4mg (n=6)
Ø Cohort 3 ALLO 6-18mg (n=5)
¨ Did not receive allocated intervention (n= 0)
¨ Lost to follow-up (n= 0)
¨Discontinued intervention (n= 1)
Ø lost interest in trial

¨ Received allocated intervention (n=6)
¨ Did not receive allocated intervention (n= 0)
¨ Analysed (n=4)
¨ Excluded from analysis (n=1)
Ø Cohort 3 in progress
Allocation Cohorts
Analysis
Follow-Up
Randomized (n= 23)
Cohort 1 (n=8); Coh 2 (n=9); Coh 3 (n=6)

Enrollment
ACTIVE PLACEBO
Figure 2.2. Allopregnanolone Phase 1 Clinical Trial Consort Statement
33

No. Study ID Age Sex Diagnosis ApoE Ethnicity Race MMSE Formal Education (yrs) Source
1 102 74 M MCI 3/ 4 NH Af American 23 14 Allo Clinical Trial - Cohort 1
2 103 60 F MCI 3/ 4 NH White 22 14 Allo Clinical Trial - Cohort 1
3 104 78 F MCI 3/ 4 NH White 30 16 Allo Clinical Trial - Cohort 1
4 105 72 F MCI 3/ 4 NH Asian 22 16 Allo Clinical Trial - Cohort 1
5 109 89 M MCI 3/ 3 NH White 24 20 Allo Clinical Trial - Cohort 1
6 110 71 M MCI 3/ 4 NH White 24 20 Allo Clinical Trial - Cohort 1
7 111 80 M MCI 3/ 3 NH White 19 18 Allo Clinical Trial - Cohort 1
8 112 67 F MCI 3/ 4 NH White 26 18 Allo Clinical Trial - Cohort 1
9 201 79 F MCI 3/ 4 NH Asian 24 14 Allo Clinical Trial - Cohort 2
10 203 75 M MCI 4/ 4 NH White 24 16 Allo Clinical Trial - Cohort 2
11 204 74 M MCI 3/ 4 NH White 23 15 Allo Clinical Trial - Cohort 2
12 207 73 F MCI 3/ 4 NH White 25 14 Allo Clinical Trial - Cohort 2
13 215 78 F MCI 3/ 4 NH Asian 22 14 Allo Clinical Trial - Cohort 2
14 217 87 M MCI 3/ 3 NH White 25 17 Allo Clinical Trial - Cohort 2
15 219 74 M MCI 3/ 3 NH White 23 16 Allo Clinical Trial - Cohort 2
16 221 81 F MCI 3/ 4 NH Af American 20 14 Allo Clinical Trial - Cohort 2
17 301 79 F MCI − H More than 1 25 16 Allo Clinical Trial - Cohort 3
18 303 67 F MCI − NH White 21 12 Allo Clinical Trial - Cohort 3
19 305 70 F MCI − NH White 22 18 Allo Clinical Trial - Cohort 3
20 307 79 F MCI − NH White 28 20 Allo Clinical Trial - Cohort 3
21 310 75 M MCI − H White 28 12 Allo Clinical Trial - Cohort 3
22 311 83 M MCI − NH White 22 12 Allo Clinical Trial - Cohort 3
23 312 67 M MCI − H White 25 16 Allo Clinical Trial - Cohort 3
Table 2.1. Allopregnanolone Phase 1 Clinical Trial Participant Demographics
34

Forumlation
Allopregnanolone is formulated as a 1.5mg/ml solution in 6% SBECD (Dexolve â) and
0.9% sodium chloride. Placebo is 0.9% sodium chloride. Both products are
manufactured by the University of California Davis GMP facility as described in IND
111085.

Pharmacokinetics
PK analyses were done on the plasma samples from the first and last treatment visits.
Samples were obtained at 0, 0.25, 0.5, 1.0, 2.0, 4.0, and 6.0 hours after start of infusion.
The Allo plasma concentration will be measured using tandem quadrupole mass
spectrometry Waters Acquity ultra-high performance liquid chromatography conducted
by the University of California Dais Bioanalytics laboratory. PK parameters are
calculated by the clinical trial biostatistics team using Microsoft Excel Add-In PKSolver.

Statistics
To date, preliminary analyses have been conducted by the biostatistics team on cohort 1
and cohort 2, while the rest of the clinical trial team remains blinded to the data.
Changes in MRI biomarkers, as well as performance on ADAS-Cog, MOCA, and
35
CogState cognitive measures have been calculated as a change from baseline (Chapter
6) to provide initial estimates of treatment efficacy at the two lowest doses.

Study Ethics
All patients provided written informed consent prior to their recruitment to the clinical
trial, and with the approval of the University of Southern California Health Sciences
Institutional Review Board (USC IRB). As part of the inclusion criteria, all participants
had the cognitive capacity to provide their own consent and proxies were not used. The
study was conducted according to the principles expressed in the Declaration of
Helsinki. All methods were performed in accordance with the relevant guidelines and
were approved by the United States Food and Drug Administration (IND 113772) and
USC IRB. The study was listed on Clinicaltrials.gov (NCT02221622) in advance of study
recruitment.

Study Clinical Trial Status
To date, all but one participant has completed the study. Database lock is expected in
the first quarter of 2018. In anticipation of final safety analyses from the biostatics team,
it is estimated that the final dose will be 10mg for females and 6mg for males, due to
observed differences in sedation in final cohort. Results from secondary outcomes from
the first two cohorts are briefly described in Chapter 6. Data from the final cohort will be
analyzed and results are expected in 2018.

36
Materials and Methods
BIOMARKER SAMPLE COLLECTION
ApoE Genotype: 4-8ml whole blood was collected at study baseline into BD
Vaccutainer tubes coated with K2 EDTA (367863, BD Medical, Franklin Lakes,
NJ). Sample was frozen at -20
o
C the same day as collection.

iPSC: 6-10ml whole blood was collected at study baseline into BD Vaccutainer
tubes coated with K2 EDTA (367863, BD Medical, Franklin Lakes, NJ). Samples
were processed as described in full in Chapter 3.

BIOMARKER SAMPLE PROCESSING:
ApoE Genotype: Whole blood sample was thawed and DNA isolated using
Qiagen DNA Mini Kit (51104, Qiagen, Hilden, Germany) as per the
manufacturer’s instructions. 25ng of DNA was mixed with for ApoE primers
(Forward 5'- TAA GCT TGG CAC GGC TGT CCA AGG A -3'; Reverse 5'- ACA
GAA TTC GCC CCG GCC TGG TAC ACT GCC -3') and RT² SYBR Green ROX
qPCR Mastermix (330503, Qiagen, Hilden, Germany), and PCR was run [95
o
C
10mins; (94
o
C 30 sec, 58
o
C 30 sec, 72
o
C 1 min) x 40 cycles; 72
o
C 7 mins]. The
PCR product was cleaned using the DNA Clean & Concentrator Kit (D4003,
Zymo Research Corp, Irvine, CA). Cleaned PCR product was digested with Hha1
(R0139s, New England BioLabs, Ipswich, MA) at 37
o
C for 1.5 – 2 hours. The
digest product with 100bp DNA ladder (15628019, Thermo Fisher Scientific,
Waltham, MA) was visualized by agarose gel electrophoresis (4% gel, 100V 50-
60mins) and ChemiDoc Imaging System (Bio-Rad, Irvine, CA). ApoE genotype
37
was assigned after restriction fragment length polymorphism analysis. ApoE 3/3
and ApoE 4/4 controls were included using DNA isolated from humanized ApoE
3/3 and ApoE 4/4 mice.

iPSC: Reprogramming and differentiation protocols are described in full in
Chapter 3.

Results
APOE GENOTYPING
Whole blood samples from the first 16 participants (Cohorts 1 & 2) who completed the
study were genotyped for ApoE2, ApoE3, and ApoE4 alleles. The results were
consistent across cohorts with 25% of participants ApoE 3/3 and 63-75% of participants
ApoE3/4. Only 1 participant (13%) was ApoE4/4 (Figure 2.3.). When analyzed by sex,
the genotype breakdown was not consistent with 100% of females ApoE3/4, 50% of
males ApoE 3/3, 38% of males ApoE 3/4, and 13% of males ApoE 4/4 (Figure 2.4.).

38

25%
69%
6%
25%
75%
0%
25%
63%
13%
0
2
4
6
8
10
12
ApoE 3/3 ApoE 3/4 ApoE 4/4
Number of Participants
ApoE Genotype: By Cohort
ALL
(N=16)
Cohort 1
(N=8)
Cohort 2
(N=8)
0%
100%
0%
50%
38%
13%
0
1
2
3
4
5
6
7
8
9
ApoE 3/3 ApoE 3/4 ApoE 4/4
Number of Participants
ApoE Genotype: By Sex
Female
(N=8)
Male
(N=8)
Figure 2.3. Participant ApoE Genotyping by Dosing Cohort
Figure 2.3. Participant ApoE Genotyping by Sex
39
Discussion
It has been estimated that of AD patients in United States, 31% are ApoE 3/3, 56-66%
are ApoE 3/4, 11-14% are ApoE 4 homozygous, and 7% carry a protective ApoE 2
allele
1,120
. Reports of ApoE4 carriers in clinical trials indicate that of those patients who
enroll in such studies, ~62% are ApoE4 carriers
121,122
. Based on this, the ApoE
breakdown of participants enrolled in the Allo clinical trial is consistent with the overall
AD population. This will be important going forward in interpreting the data from iPSC-
NSCs, due to the variability reported in cells from sAD donors. With the increase in
donors with this clinical trial, we can evaluate the cellular outcomes to do a preliminary
evaluation of the effect of ApoE genotype on NSC proliferation, mitochondrial function
and response to Allo.

40
Chapter 3: iSPC Generation & Differentiation to NSC and
Baseline Evaluation of Proliferation and Mitochondrial
Respiration

Introduction
Upon deciding to incorporate iPSC reprogramming into the clinical trial protocol, it was
determined that PBMC’s were the most appropriate somatic cell source in this instance
due to relative lack of invasiveness, increased patient acceptance, and ease of sample
collection compared to a skin biopsy for fibroblasts. An established protocol of same-day
PBMC isolation was used for all participant whole blood samples which were
subsequently reprogrammed into iPSCs. During Cohort 3, a second isolation method
was also used and its effect on reprogramming efficiency investigated. This was
determined to be important because if iPSCs will be generated in future clinical trials,
such efforts would likely involve a much larger number of patient samples (up to several
hundred) collected at clinical trial sites across the country, which would then be shipped
for processing at a central lab. Samples would no longer be able to be processed on the
same day and, likely, not by a single laboratory technician. As a result, a study was
conducted to determine the effect of transport conditions and a more time-efficient
PBMC isolation method, with reduced inter-operator variability, on reprogramming
efficiency.

41
PBMC reprograming and differentiation to NSCs were identical for all clinical trial
participant samples. Both iPSC and NSCs were successfully generated and
characterized. For the first time, we describe the proliferative and mitochondrial
phenotypes of iPSC derived NSCs from a significant cohort of AD donors. This serves to
characterize the cells at baseline, prior to treatment with Allo. If, as we hypothesize, only
a subset of cell lines has a change in functional outcome as a result of Allo treatment,
these initial analyses act as a foundation upon which to compare responders and non-
responders, and determine a potential basis for such effects.

Materials and Methods
STUDY ETHICS
As described fully in Chapter 2, all participants in the clinical trial provided written
informed consent, prior to recruitment. As part of the written informed consent,
they consent first to participation in the clinical trial (including samples for
biomarker research) and then separately to providing the blood sample that
would be used to establish induced pluripotent stem cell line(s). The
establishment of participant iPSC lines was approved by the USC IRB.

BLOOD COLLECTION
As part of the clinical study protocol, at the randomization visit, about 8-10mls of
whole blood were drawn into BD Vaccutainer tubes coated with K2 EDTA
(367863, BD Medical, Franklin Lakes, NJ). All blood draws for peripheral blood
mononuclear cells (PBMC) isolation were done prior to infusion, ensuring that

42
data generated from derivatives of these cells were agnostic to the clinical study
arm the participant was subsequently randomized to. Tubes were then stored at
4C until transferred to the Brinton or Ichida laboratory for processing to isolate
the PBMCs. PBMC’s were isolated and frozen as soon as possible after blood
draw, always on the same day. For the final 6 clinical trial participants, an
additional two tubes of blood (6-8mls total) were collected in an identical manner
to test alternative isolation protocols that could be used in future, multi-site
clinical trials for increased processing efficiency and to reduce inter-operator
variation a known limitation of the density gradient centrifugation isolation
method
123
.

PBMC ISOLATION
Density Gradient Centrifugation
Sample Handling:
After collection, blood sample was stored at 4
o
C until PBMC isolation. PBMC
isolation was done as soon as possible after blood collection, always on the
same day.

Sample Processing:
PBMCs were purified by density gradient centrifugation with Lymphocyte
Separation Medium (LSM) (25-072-CV, Corning, Manassas, VA) using the
following protocol. Blood was removed from the BD Vaccutainer and transferred
to a sterile tube and diluted 1:1 with phosphate buffered saline (PBS). LSM was
added to a sterile 50ml tube at a 1:2 ratio of LSM to blood/PBS by volume. The

43
diluted blood was carefully layered on top of the LSM, ensuring there was no
disruption to the surface interface. The sample was centrifuged at 800 x g at
room temperature for 15 minutes with no brake. At spin completion, a number of
layers can be seen: serum, buffy coat (containing PBMCs), LSM, and red blood
cells. Using a P200 pipette, the buffy coat was transferred to a sterile 15ml tube,
taking care to remove as little LSM as possible. Depending on the original
sample volume and size of buffy coat, usually 2-3mls were removed to the new
tube. The isolated buffy coat was then diluted to 10mls with sterile PBS and
centrifuged at 700 x g for 10 minutes. At spin completion, a PBMC pellet can be
seen. The supernatant is aspirated and the pellet washed with an additional
10mls sterile PBS, and spun at 700 x g for 5 minutes. After this final spin, the
supernatant was aspirated and the PBMC pellet was resuspended in 3mls of
freeze media, which consisted of 90% Fetal Bovine Serum (FBS) (302020,
ATCC, Manassasa, VA) and 10% dimethyl sulphoxide (DMSO) (D2438, Sigma-
Aldrich, St. Louis, MO). 1ml aliquots of the cell suspension were transferred to
cryovials (10018-758, VWR, Radnor, PA, VWR), and slowly frozen overnight at -
80
o
C, before transfer to longer-term liquid nitrogen storage.

SepMate
TM
Isolation
Sample Handling
After collection, blood was stored at 4
o
C for 1-2 hours, until it was transferred to
the Brinton laboratory. Upon receipt, one tube was held at room temperature and
one tube was placed in a small Styrofoam box with reusable ice packs (as would

44
be used for shipping) and both were held overnight. Samples were processed the
next day, 22 - 36 hours after blood collection.

Sample Processing:
PBMCs were purified using SepMate
TM
-15 or Sepmate
TM
-50 tubes (85415,
85450, STEMCELL Technologies, Vancouver, BC, Canada) as appropriate for
the blood volume to be processed, according to the manufacturer’s instructions.
In brief: LSM was added to the SepMate
TM
tube through the insert in the tube.
Whole blood was diluted with PBS/2% FBS and added quickly to the SepMate
TM

tube, which was then spun at 1200 x g for 10 minutes with the brake, or 20
minutes if it had been more than 24 hours since sample collection. The resulting
layer above the insert, consisting of the serum and buffy coat, was poured off into
a 50ml tube, and PBS/2%FBS added to a total of 35mls. The tube was spun at
300 x g for 8 minutes with the brake on. The wash was repeated and the
resulting PBMC pellet re-suspended in 3ml of freezing media (90% FBS (302020,
ATCC, Manassas, VA) and 10% DMSO (D2438, Sigma-Aldrich, St. Louis, MO))
and 1ml aliquots were slowly frozen at -80
o
C overnight before transfer to liquid
nitrogen for longer-term storage.

REPROGRAMMING FROM PBMCS TO IPSCS
iPSCs were generated for all participants who were randomized in the Allo Phase
Ib/IIa Clinical Trial, using PBMCs isolated as described above. Please see Table
2.1 for an overview of all participant demographics. PBMCs were reprogrammed
using the episomal vectors pCXLE-hOCT3/4-shp53, pCXLE-hSK, pCXLE-hUL,

45
and pCXWB-EBNA1 (Addgene, Cambridge, MA,
https://www.addgene.org/browse/article/6095/), as described in Okita et al
124
. In
brief, CF-1 irradiated mouse embryonic fibroblasts (MEFs) (GSC 6301G, MTI-
Global Stem, Gaithersburg, MD) are thawed and seeded on a six-well plate (25-
105, Genesee, San Diego, CA), at a density of 0.25*10
6
/well and cultured
overnight in 2m/well MEF media (Media recipes below), the day prior to
reprogramming. Each reprogramming reaction requires two wells of prepared
MEFs. On day of reprogramming (Day 1) aspirate MEF media and prepare each
well with 3ml X-VIVO 10 media (04-380Q, Lonza, Cologne, Germany) warmed to
37
o
C, 30units/ml of human IL-2 (prepared as per manufacturer’s instructions)
(200-02, Peprotech, Rocky Hill, NJ), and 5ul of DynabeadsTM Human T-
Activator CD3/CD28 (washed as per manufacturer’s instructions) (11161D,
Thermo Fisher Scientific, Waltham, MA). The plasmid mixture is prepared as
described in Table 3.1. Transfection Solution (Amaxa NHDF Nucleofector Kit,
VPD-1001, Lonza, Cologne, Germany) was prepared (Per reaction: 82ul
Nucleofector solution, 18ul Supplement 1). Vial of frozen PBMCs was quickly
thawed in a 37
o
C water bath. Cell suspension was transferred to sterile 15ml
tube, 37
o
C X-Vivo media was added to 10mls, and then spun at 180 x g for 5
minutes with brake. The cell pellet was then resuspended in 1ml of warm X-Vivo
media and counted via haemocytometer (DHC-N01-5, iNCyto, Korea). Cell
suspension with 1*10^6 cells is removed to a sterile 15ml tube and spun at 200 x
g for 5 minutes. All the supernatant is aspirated (critical to avoid arcing during
electroporation) and, working quickly, 100ul of transfection solution is added and
the cell pellet gently resuspended. Plasmid mixture is added, in the volume

46
required to deliver the amounts given in Table 3.1, and the cell suspension is
then transferred to the transfection cuvette. The Amaxa Nucleofector 2b (AAB-
1001, Lonza, Cologne, Germany), program V-24, was used to electroporate the
plasmid mixture into the cells. Using the provided transfer pipette, the cell
suspension is divided dropwise between two previously prepared wells of MEFs
and the plate is then incubated at 37
o
C/5% CO
2
. Two days later (Day 3), 3ml of
human embryonic cell media (hES) and 10µM of Y-27632, also known as rock
inhibitor (RI) (S1049, Selleck Chemicals, Houston, TX) is added to each well. An
additional two days later (Day 5), the media is aspirated and 3ml fresh hESC
media with rock inhibitor is added and changed daily (See Figure 3.1). If
nucleofection is successful, GFP expression should be visible from Day 2,
reaching maximal expression on approximately Day 4. Although line dependent,
colony formation is visible on average from Day 10, and are large enough to
manually isolate and expand from Day 21 (approximately). Clones should be
isolated before colonies have coalesced. Original MEF’s are good for about 7
days, and additional MEF’s can be thawed and added with daily hESC/RI media
change as required (0.1*10^6 cells/well, approximately Day 7 and 14).

Table 3.1. Non-Integrating Episomal Plasmid Mixture
Plasmid Identifier Amount (ug) per reaction Genes
pCXLE-hOCT3/4-shp53-F 0.83 OCT3/4, TP53 shRNA
pCXLE-hSK 0.83 SOX2, KLF4
pCXLE-hUL 0.83 L-MYC, LIN28
pCXWB-EBNA1 0.5 CAG-EBNA1
pCE-GFP 0.83 EGFP

47

iPSC induction efficiency was calculated using the following equation: (# of iPSC
colonies/number of electroporated cells)*100%. Twelve iPSC clones were isolated for
each clinical trial participant, where reprogramming efficiency allowed. Clones were
grown on Matrigel hESC-qualified matrix (354277, Corning, Corning, NY) and cultured in
mTeSR
TM
1 (85850, STEMCELL Technologies, Vancouver, BC, Canada) in a 24-well
plate (3526, Corning, Corning, NY). Based on growth, six clones were passaged with
Accutase (07920, STEMCELL Technologies, Vancouver, BC, Canada) and expanded,
each to a single well in a 6-well plate, before one was selected for further expansion and
characterization. mTeSR media was changed daily. Cells were passaged approximately
~4 days, when cells were about 80% confluent. Cells were passaged using Accutase
onto fresh matrigel in mTeSR-1 with rock inhibitor, at a 1:25 ratio approximately.
Remaining cells were frozen, as described above.

Day 0 Day 2 Day 4 Day 5 ~Day 24
X-Vivo Media
+ IL-2
+ Dynabeads
hES media
+ bFGF
+ Rock Inhibitor Plasmid Transduction
Pick colonies
Figure 3.1. PBMC Reprogramming Protocol

48
IPSC CHARACTERIZATION
Immunocytochemistry:
The iPSC clone selected for expansion, characterization, and differentiation was
stained with Tra-1-60 dye for live cells (A25618, Thermo Fisher Scientific,
Waltham, MA). Cells were then fixed with 4% paraformaldehyde (163-20145,
Wako Chemicals USA, Richmond, VA) at room temperature for 30 mins, and
stained with nanog 1:400 (AF1997, R&D Systems, Minneapolis, MN), Alexa Fluor
647 donkey anti-goat 1:500 (A-21447, Thermo Fisher Scientific, Waltham, MA),
and Tra-1-81 1:250 (4745S, Cell Signaling Technology, Danvers, MA), Alexa
Fluor 488 donkey anti-mouse 1:500 (A-21202, Thermo Fisher Scientific,
Waltham, MA).

Karyotyping:
Concurrent to culturing for expansion and immunocytochemistry, iPSCs were
cultured in a T25 flask (15708-130, VWR International, Radnor, PA), between
passage 2 and passage 6. These were used to assess genomic integrity after
reprogramming via G-Banded karyotyping by the Children’s Hospital Los Angeles
Cytogenetic Laboratory. For each cell line, 20 cells were analyzed and 5
karyotyped at a resolution of 350-550.

IPSC DIFFERENTIATION TO NSCS
A schematic summary of the differentiation procedure is shown in Figure 3.2.
iPSCs were differentiated to NSCs via monolayer dual-SMAD inhibition as
described in Chambers et al
125
and by Tomishima
126
. In brief, iPSCs are

49
passaged with Accutase to single cells and cultured in a 12-well plate (3513,
Corning, Corning, NY) at a density of 1.75*10^6 cells/well, on Matrigel hESC-
qualified matrix (354277, Corning, Corning, NY) with mTeSR-1 and 10µM of RI.
After 24 hours, a confluent monolayer of iPSC is ready for neural induction. 90-
95% confluency is required for CNS neural cell induction. mTeSR-1 media is
aspirated and SRM media, containing 10µM SB431542 (S4317, Sigma-Aldrich,
St. Louis, MO) and 100nM LDN193189 (11802, Cayman Chemical, Ann Arbor,
MI) is added (Day 0). On Day 1, aspirate the SRM and replace with fresh SRM
containing SB431542 (10µM) and LDN193189 (100nM). Repeat on Day 2. On
Day 4, aspirate the SRM and add SRM and N2 media (3:1 ratio) with SB431542
(10µM) and LDN193189 (100nM). On Day 6, aspirate media and add SRM and
N2 media (1:1 ratio) with SB431542 (10µM) and LDN193189 (100nM). On Day 8,
aspirate media and add SRM and N2 media (1:3 ratio) with SB431542 (10µM)
and LDN193189 (100nM). On Day 10, a thick layer of neuroectoderm will have
formed and is incubated with Accutase for 30 mins at 37
o
C, to disassociate layer
into single cells. Using a P1000 pipette, cells are gently triturated with
N2/Accutase and then the cell suspension is filtered through a 40µM nylon filter
(431750, Corning, Corning, NY) to a sterile 50ml tube. Cells are washed in N2
media and centrifuged at 180 x g for 5 mins, twice. The cell pellet is then
resuspended in N2 containing all growth factors and counted using a
hemacytometer (DHC-N01-5, iNCyto, Korea). 5*10^6 cells are removed and
made up to 1ml volume with N2 media with growth factors and 10µM rock
inhibitor. During centrifugation, a 6-well plate coated with Matrigel hESC-qualified
matrix (354277, Corning, Corning, NY) was prepared by aspirating all the liquid

50
from the well and allowed to dry for 15 mins. The well is spotted with 20µl drops
of the cell suspension and allowed to sit for 20 minutes before slowly adding an
additional 1ml of N2 media, growth factors, and rock inhibitor before being
carefully transferred to the incubator. Media is changed the following day to N2
and growth factors without the rock inhibitor. Thereafter media is changed every
other day. Cells are usually ready to passage two to three days later and
expanded 1:3. Cells were expanded 1:6 at passage 3 and seeded for
characterization. RI is included until passage 4, to increased cell survival. NSCs
were used in assays between passage 4 to passage 6.
Day -1
Passage & seed iPSC to confluency
mTeSR-1 media + Rock Inhibitor
Day 0
SRM Media
10μM SB431542, 100nM LDN193189
Day 1
SRM Media
10μM SB431542, 100nM LDN193189
Day 2
SRM Media
10μM SB431542, 100nM LDN193189
Day 4
SRM Media
10μM SB431542, 100nM LDN193189
N2 Media
Day 6
SRM Media
10μM SB431542, 100nM LDN193189
N2 Media
Day 8
SRM
10μM SB431542, 100nM LDN193189
N2 Media
Day 10
Passage NSC
N2 + bFGF + EGF + heparin
Figure 3.2. NSC Directed Differentiation: Dual SMAD
inhibition monolayer protocol

51
NSC CHARACTERIZATION
Immunocytochemistry
NSCs were fixed in 4% paraformaldehyde at room temperature for 30 minutes,
washed, blocked, and then stained with Nestin 1:1000 (A302-212A, Bethyl
Laboratories, Montgomery, TX), Alexa Fluor goat anti-rabbit 488 (A11008,
Thermo Fisher Scientific, Waltham, MA) 1:500, and Sox-2 (MAB2018, R&D
Systems, Minneapolis, MN) 1:75, Alexa Fluor goat anti-mouse 647 (A21235,
Thermo Fisher Scientific, Waltham, MA) 1:500.

NSC PROLIFERATION
Cell proliferation was assessed by flow cytometry analysis of EdU
immunopositive nuclei using the Click-iT Plus EdU Alexa Fluor Flow Cytometry
Assay Kit. On Day 1, NSCs were plated at a density of 200,000 cells per well in
1ml of N2 media containing growth factors, on a 6-well plate (25-105, Genesee,
San Diego, CA) coated with Matrigel hESC-qualified matrix (354277, Corning,
Corning, NY). Six technical replicates were prepared for each cell line. Cells were
incubated at 37
o
C/5% CO
2
overnight. On Day 2 media was changed and 5µM
EdU was added to each well. On Day, 24 hours after EdU addition, media was
aspirated and Accutase was added. After 5 minutes, disassociated cells were
removed to 15ml tube with 5ml Neurobasal media and centrifuged at 400 x g for
5 minutes. After supernatant was aspirated, cells were resuspended in 300µl of
PBS and gently disassociated to single cells with pipette. While gently vortexing
the tube, 700µl of ice-cold 100% Ethanol (EX0276-1, Merck Milipore, Billerica,
MA) was added dropwise. Fixed cells were held on ice for at least 30 minutes

52
and then stored at -20
o
C until flow cytometry analysis. On day of flow cytometry
cells with washed with 3ml of 1% BSA in PBS and centrifuged at 400 x g for 5
mins. After supernatant was aspirated, cells were permeabilized using Click-iT
saponin-based reagent for 15 minutes. Cells were then incubated in 125µl Click-
iT reaction cocktail (109.4µl PBS, 2.5µl copper protectant, 12.5µl Reaction Buffer
Additive) with 0.63µl fluorescent dye picolyl azide for 30 minutes at room
temperature. Cells were then washed with 3ml of wash buffer, centrifuged at 180
x g for 5 minutes, supernatant aspirated and the cell pellet resuspended in 200µl
of Click-iT wash reagent. Cell suspension was filtered through 35µM nylon filter
into round-bottom tube (352235, Corning Inc, Corning, NY) and the filter washed
with an additional 200µl of wash reagent. Samples were then read on a LSR II
flow cytometer (BD Biosciences, San Jose, CA). NSCs were gated by forward
and side scatter and the % of EdU+ NSCs analyzed. Data was analyzed using
FlowJo V X 10.0.7r2.

NSC MITOCHONDRIAL RESPIRATION
Mitochondrial function was assessed using the Seahorse XFe96 Analyzer
(Agilent Technologies, Santa Clara, CA). On Day 1, NSCs were plated at a
density of 15,000 cells/well in 50µl of N2 media containing growth factors on 96-
well cell culture microplate from the Seahorse XFe96 FluxPak was coated with
Matrigel hESC-qualified matrix (354277, Corning, Corning, NY). The plate is left
at room temperature for 30 minutes to allow cells to settle and then incubated at
37
o
C/5% CO
2
overnight. On Day 2, to arrest cell proliferation, media was
changed to N2 without growth factors. A XFe96 cartridge from the Seahorse

53
XFe96 Flux Pak (102416, Agilent Technology, Santa Clara, CA) was rehydrated
overnight as per manufacturer’s instructions. On Day 3, 28 hours after growth
factors removed, media was changed to unbuffered Seahorse Media, with 25mM
glucose, warmed to 37
o
C. The plate is incubated at 37
o
C without CO
2
for 45-60
minutes. The XFe96 cartridge is loaded with the following inhibitors: Port A:
Sodium Pyruvate (50mM); Port B: Oligomycin (32µM; to inhibit ATP synthase);
Port C: FCCP (carbonylcyanide p-trifluoromethoxyphenylhydrazone, 9µM; to
uncouple mitochondria); Port D: Rotenone (10µM; to inhibit Complex I of the
electron transport chain) and Antimycin (50µM; to inhibit Complex III of the
electron transport chain). Injection volumes for all ports were 25µl. The inhibitor
concentrations at injection are: Sodium Pyruvate 7mM (S8636, Sigma-Aldrich,
St. Louis, MO), Oligomycin 4µM (151786, MP Biomedicals, Santa Ana, CA),
FCCP 1µM (0453, Tocris Cookson, Bristol, UK), Rotenone 1µM (0215015410,
MP Biomedicals, Santa Ana, CA) / Antimycin 5µM (100046, MP Biomedicals,
Santa Ana, CA). Three baseline measurements of oxygen consumption rate
(OCR) and extracellular acidification rate (ECAR) were taken before sequential
injection of mitochondrial inhibitors. Three readings were taken following inhibitor
addition and prior to automated addition of subsequent inhibitors. OCR and
ECAR were automatically calculated and recorded by Seahorse XFe96 Wave
software. To correct for any plating differences, protein readings of all wells
(5000006, Bio-Rad Laboratories, Hercules, CA) were used to normalize wells
within the Seahorse XFe96 Wave software. Parameters of mitochondrial function
are calculated as follows: Basal Respiration (Average first six OCR readings);
Non-Mito Respiration (Average last three readings after Rotenone/Antimycin

54
addition); Basal Mitochondria Function (Basal Respiration – Non-Mito
Respiration); ATP Production (Basal Respiration – first reading after Oligomycin
addition); Proton Leak (First reading after Oligomycin addition – Non-Mito
Respiration); Maximal Capacity (First reading after FCCP addition – Non-Mito
Respiration); Reserve Capacity (First reading after FCCP addition – Basal
Respiration). Wells with negative values, as well as OCR or ECAR readings and
calculated parameters of mitochondrial function that are +/- 2 standard deviations
of the mean are removed as outliers.

CELL MEDIA RECIPES
MEF Media
§ Fetal Bovine Serum (302020, ATCC, Manassasa, VA): 15%
§ MEM Non-Essential Amino Acids (11140-050, Thermo Fisher
Scientific, Waltham, MA): 1%
§ Glutamax (35050-061, Thermo Fisher Scientific, Waltham, MA):
1%
§ 2-mercaptoethanol (21985-023, Thermo Fisher Scientific,
Waltham, MA): 0.1%
§ KnockOut DMEM (10829-018, Thermo Fisher Scientific, Waltham,
MA): to 100%

Add all to KnockOut DMEM and sterile filter (431096, Corning, Corning,
NY).

55
hESC Media
§ KnockOut Serum Replacement (10828-028, Thermo Fisher
Scientific, Waltham, MA): 20%
§ MEM Non-Essential Amino Acids (11140-050, Thermo Fisher
Scientific, Waltham, MA): 1%
§ Glutamax (35050-061, Thermo Fisher Scientific, Waltham, MA):
1%
§ 2-mercaptoethanol (21985-023, Thermo Fisher Scientific,
Waltham, MA): 0.1%
§ bFGF (02634, STEMCELL Technologies, Vancouver, BC,
Canada): 88µl of 10µg/ml stock (Final concentration: 4ng/ml).
§ DMEM/F12 1:1 (11320-033, Thermo Fisher Scientific, Waltham,
MA): to 100%

Add all to DMEM/F12 and sterile filter (431096, Corning, Corning, NY).
Protect from light and store at 4
o
C for up to two weeks.

SRM Media
§ KnockOut Serum Replacement (10828-028, Thermo Fisher
Scientific, Waltham, MA)” 15%
§ Glutamax (35050-061, Thermo Fisher Scientific, Waltham, MA):
1%
§ MEM Non-Essential Amino Acids (11140-050, Thermo Fisher
Scientific, Waltham, MA): 1%

56
§ 2-mercaptoethanol (21985-023, Thermo Fisher Scientific,
Waltham, MA): 0.1%
§ KnockOut DMEM (10829-018, Thermo Fisher Scientific, Waltham,
MA): to 100%

Add all to KnockOut DMEM and sterile filter (431096, Corning, Corning, NY).
Protect from light and store at 4
o
C for up to 1 month.

N2 Media
Base:
• B27 without Vitamin A (12587-010, Thermo Fisher
Scientific, Waltham, MA): 2%
• N2 (17502-048, Thermo Fisher Scientific, Waltham, MA):
1%
• Glutamax (35050-061, Thermo Fisher Scientific, Waltham,
MA): 1%
• Neurobasal media without phenol red (12348-017, Thermo
Fisher Scientific, Waltham, MA): to 100%

Growth Factors:
• bFGF (02634, STEMCELL Technologies, Vancouver, BC,
Canada): 50µl of 10µg/ml stock (Final concentration
10ng/ml)

57
• EGF (2633, STEMCELL Technologies, Vancouver, BC,
Canada): 50µl of 10µg/ml stock (Final concentration
10ng/ml)
• Heparin: (07980, STEMCELL Technologies, Vancouver,
BC, Canada): 125µl of 0.2% stock solution (Final
concentration 0.0005%).

Add all to neurobasal media, store at 4
o
C and use within 1 month (no growth
factors). Once growth factors added, use within 1 week.

Seahorse Media
§ Dulbecco′s Modified Eagle′s Medium Powder (without glucose, L-
glutamine, phenol red, sodium pyruvate and sodium bicarbonate)
(D5030-1L, Sigma-Aldrich, St. Louis, MO): 1 vial
§ Glutamax (35050-061, Thermo Fisher Scientific, Waltham, MA):
10ml
§ NaCl (0241, Amresco LLC, Solon, OH): 1.85g
§ Sterile diH
2
O: to 1L

Measure out ~900ml Sterile diH
2
O into large beaker, add DMEM powder and
dissolve with stirring. Add glutamax and NaCl. pH to 7.4 at 37
o
C with NaOH/HCl.
Make up to 1000ml with sterile diH
2
O and stir briefly. Sterile filter (431096,
Corning, Corning, NY), and store at 4
o
C. On day of use, warm to 37
o
C and add
25mM glucose (BDH9230, VWR International, Radnor, PA).

58

Plasmid Acknowledgements:
pCXLE-hOCT3/4-shp53-F was a gift from Shinya Yamanaka (Addgene plasmid #
27077)
127

pCXLE-hSK was a gift from Shinya Yamanaka (Addgene plasmid # 27078)
127

pCXLE-hUL was a gift from Shinya Yamanaka (Addgene plasmid # 27080)
127

pCXWB-EBNA1 was a gift from Shinya Yamanaka (Addgene plasmid # 37624)
124

pCE-GFP was a gift from Shinya Yamanaka (Addgene plasmid # 41858)
124

Statistics
Statistical analysis was performed in Prism 6.0. For subgroup analysis, pooled groups
were determined to be non-parametric, and so Mann-Whitney test was used. The
difference between groups was considered significant if the p-value was <0.05.

Results
IPSC GENERATION FROM PERIPHERAL BLOOD
As described in Okita et al, reprogramming was successfully completed using the
Yamanaka factors and PBMCs isolated from peripheral whole blood
124
. The
effect of whole blood storage and processing on reprogramming efficiency was
compared. Whole blood was either stored at 4
o
C and PBMCs separated by
density gradient centrifugation on the same day as collection, or blood samples
were stored overnight (up to ~28hrs) on ice or at room temperature, and PBMCs
separated using Sepmate tubes. All PBMC samples were frozen at -80
o
C prior to

59
transfection. In all storage and processing comparisons, reprogramming was
done identically. In all cases, following plasmid transfection, colonies attached to
culture dish were observed from approximately Day 10 – Day 14. Cells show
typical iPS round morphology that then become tightly packed with well-defined
margins (Figure 3.3).

A B C
D E F
Figure 3.3. iPSC induction from lymphocytes using nonviral, integration-free
plasmids. A) Day 5 post-transduction. Scattered GFP+ cells indicate cells took up
plasmids. B) Day 10, proliferation and colony formation. C) Day 12, colony expansion. D)
Day 19, loss of GFP in rapidly proliferating cells is indicative of dilution of episomal
plasmids. E) Day 19, brightfield. iPSC morphology is emerging. MEF’s visible in
background. F) Mature iPSCs after several passages, plated on matrigel with mTeSR-1.

60

At around Day 21 - Day 28, the number of colonies were counted. For 13 PBMC
samples processed on the same day as blood collection, the average number of
colonies is 56.7±12.3, with a reprogramming efficiency of 0.001-0.013% (Figure 3.4).
While some of the difference can be attributed to operator experience, but this does not
account for all reprogramming efficiency variation, especially when some PBMC
samples were reprogrammed side-by-side. When the association between PBMC deep
freeze storage length and reprogramming efficiency was evaluated, the r
2
value was
0.45 (p<0.05) (Figure 3.5). This indicates that the longer a PBMC sample is stored, the
fewer iPSC colonies it will produce. Blood samples from five of the clinical trial
participant donors were also reprogrammed after overnight storage at 4
o
C and room-
temperature. On average, compared to same-day density centrifugation, PBMCs stored
0
50
100
150
105
109
110
111
112
207
219
305
307
310
311
312
313
Colony Count (with % Efficiency)
Participant ID
0.013%
0.01%
0.012%
0.005%
0.002%
0.012%
0.004%
0.006%
0.005%
0.002%
0.002%
0.001%
0.001%
0 100 200 300
0
50
100
150
Associated Between
Time to Reprograming and Colony Count
Time to Reprograming (Days)
Colony Count
r
2
=0.45
p<0.05
Figure 3.4. Reprogramming Colony Count.
Colony counts from reprogramming of 13
clinical trial donors indicate that overall
efficiency of PBMCs ranges from 0.001 –
0.013%.
Figure 3.5. Effect of Storage Time on
Reprogramming Efficiency. Data from
reprogramming of 13 PBMC samples indicates
that a longer time period of liquid nitrogen
storage prior to episomal reprogramming is
associated with reduced reprogramming
efficiency. All samples were reprogrammed
successfully, despite relatively low colony
counts for some donors.

61
at 4
o
C and room temperature overnight result in 58±10% and 32±26% fewer colonies
respectively (Figure 3.6). While on average, neither result in a statistically significant
difference compared to reprogramming of same day isolated PBMCs, the obviously
increased number of colonies in the latter indicates it is the ideal sample handling
process. Contrary to expectations, the data indicates that whole blood samples kept at
room temperature overnight have great reprogramming efficiencies compared to
samples that are stored at 4
o
C. It is important to note however that no matter the storage
and processing conditions, all samples were successfully reprogrammed.
100
100 100 100 100 100
42
26
135
41
31
53
68
37
40
39
93
91
0
20
40
60
80
100
120
140
160
Overall 305 307 310 312 313
Patient ID
Colony Count per Well
(% Normalized to Density Centrifugation)
Peripheral Blood Storage and PBMC Isolation Comparison
Same Day Density
Centrifugation
4C Overnight
Room Temperature
Overnight
Figure 3.6. Effect of Blood Sample Storage on Reprogramming Efficiency.
Comparing the % increase/decrease in colony counts for reprogramming of PBMCs
isolated from whole blood samples handled three ways. PBMCs isolated from whole
blood samples via density centrifugation on the same day as collection consistently
resulted in the highest number of colonies, with room temperature storage overnight
provides the next best sample storage conditions.

62

Established iPSC clones had typical human embryonic stem cell morphology, growing
as flat, compact colonies, with a high nucleus-to-cytoplasm ratio. They expressed the
stem cell markers Tra-1-60, Tra-1-81, and Nanog (Figure 3.7). After one iPSC clone was
expanded for each donor, g-banding for chromosomal abnormalities induced by
reprogramming was evaluated. In establishing the iPSC lines for the Allo Phase I clinical
trial, there were no chromosomal abnormalities in 91.7% of the clones evaluated. Initial
clones from two lines demonstrated defects which were not present in second clones
evaluated for each line. It is inferred that the initial abnormal karyotype was due to
chromosomal aberrations induced during the reprogramming process. Due to lack of
success with neural differentiation despite initial normal karyotype and successful
pluripotency characterization, a second clone was karyotyped for two iPSC lines, 109
and 204. A summary of all karyotyping results is shown in Table 3.2. and representative
g-banding results can be seen in Figure 3.8.

Tra-1-81 Nanog
Tra-1-60 Brightfield
Merged
Merged
Figure 3.7. iPSC Immunocytochemistry. iPSC lines express Tra-1-81, Nanog, and
Tra-1-60

63

Figure 3.8. Representative iPSC Karyotyping. Male and
female iPSC cultures demonstrating normal kayotypes.

Karyotype Number of clones
46, XX [20] 12
46, XY [20] 11
45,X,-
X,add(1)(p22),der(6)t(1;6)(p22;q21)[20]
1
46,XY,t(1;18)(p32;q23)[20] 1
Table 3.2. Summary of iPSC Karyotyping. Cell lines with
abnormal karyotypes were not differentiated or used for assays and
a second clone was karyotyped.

64

IPSC-NSC: PROLIFERATION
After iPSCs were differentiated, they were characterized for expression of the
NSC markers nestin and sox-2 (Figure 3.9). After expansion over several
passages, proliferation of the NSCs was evaluated by 5-ethynyl-2´-deoxyuridine
(EdU) incorporation using flow cytometry. NSCs from 22 clinical trial participants,
grown in media with all growth factors for 24 hours, demonstrated a range of
proliferation rates ranging from 15.4% to 92.2% (Figure 3.10.A). Although
proliferation from NSCs lines derived for male participants trended lower at a
mean of 47% Edu+ NSC’s versus 62% for female participant NSCs, this does not
reach statistical significance (p=0.17) (Figure 3.10.B). There is no effect of ApoE
genotype on proliferation rates (p >0.99) (Figure 3.10.C).

Figure 3.9. NSC Immunocytochemistry. iPSC-NSC lines express nestin and sox-2.

65

102
103
104
105
109
110
111
112
201
203
204
207
215
217
219
221
301
303
305
307
310
312
0
20
40
60
80
100
Participant Cell Line
% EdU+ NSCs
Baseline NSC Proliferation
ApoE 3/3
ApoE 3/4
ApoE 4/4
0
20
40
60
80
% EdU+ NSCs
NSC Proliferation:
ApoE Genotype
N=4 N=11 N=1
Female
Male
0
20
40
60
80
Baseline NSC Proliferation:
Sex
% EdU+ NSCs
N=12 N=10
A. B.
C.
Figure 3.10. NSC baseline proliferation for 22 clinical trial participant cell lines. There is no
effect of sex or ApoE genotype on baseline NSC proliferation. Bars represent the mean ± SEM
(N=4-6).

102
103
104
105
109
110
111
112
201
203
204
207
215
217
219
221
301
303
305
307
310
312
0
20
40
60
80
100
Participant Cell Line
% EdU+ NSCs
Baseline NSC Proliferation
ApoE 3/3
ApoE 3/4
ApoE 4/4
0
20
40
60
80
% EdU+ NSCs
NSC Proliferation:
ApoE Genotype
N=4 N=11 N=1
Female
Male
0
20
40
60
80
Baseline NSC Proliferation:
Sex
% EdU+ NSCs
N=12 N=10
A. B.
C.
102
103
104
105
109
110
111
112
201
203
204
207
215
217
219
221
301
303
305
307
310
312
0
20
40
60
80
100
Participant Cell Line
% EdU+ NSCs
Baseline NSC Proliferation
ApoE 3/3
ApoE 3/4
ApoE 4/4
0
20
40
60
80
% EdU+ NSCs
NSC Proliferation:
ApoE Genotype
N=4 N=11 N=1
Female
Male
0
20
40
60
80
Baseline NSC Proliferation:
Sex
% EdU+ NSCs
N=12 N=10
A. B.
C.

66
iPSC-NSC: Mitochondrial Respiration
102
103
104
105
109
110
111
112
201
203
204
207
215
217
219
221
301
303
305
307
310
312
0
10
20
30
40
50
Participant Cell Line
OCR (pmol/min)
Mitochondrial Respiration Profile
ATP production Proton Leak Maximal Respiration Spare Respirational
Capacity
Basal Mitochondrial
Respiration
Figure 3.41. NSC mitochondrial respiration profile for 22 participant cell lines. Mitochondrial parameters: basal
mitochondrial respiration, ATP production, proton leak, maximal respiration, and spare respirational capacity. Bars represent the
mean±SEM (N=3-11 wells).

67
Concurrent with proliferation analyses, NSC mitochondrial respiration was also
evaluated in live cells using the Seahorse Extracellular Flux platform. NSCs from
22 clinical trial participants demonstrated variable metabolic profiles (Figure
3.11). There was no effect of donor sex on the 5 mitochondrial parameters (basal
mitochondrial respiration, ATP production, proton leak, maximal respiration, and
reserve respirational capacity (Figure 3.12). When subgrouped by ApoE
genotype all of the parameters except for spare respirational capacity appear to
trend lower in ApoE 3/4. However, this reduction reached statistical significance
only in proton leak. Due to only having one participant with an ApoE 4/4
genotype, it is not possible to determine if this phenotype is exacerbated in
homozygous NSCs.

68

Female
Male
0
5
10
15
OCR (pmol/min)
Basal Mitochondrial Respiration:
Sex
ApoE 3/3
ApoE 3/4
ApoE 4/4
0
5
10
15
20
25
Basal Mitochondrial Respiration:
ApoE
OCR (pmol/min)
Female
Male
0
5
10
15
OCR (pmol/min)
ATP Production:
Sex
ApoE 3/3
ApoE 3/4
ApoE 4/4
0
5
10
15
20
ATP Production:
ApoE
OCR (pmol/min)
Female
Male
0
1
2
3
4
Proton Leak:
Sex
OCR (pmol/min)
ApoE 3/3
ApoE 3/4
ApoE 4/4
0
1
2
3
4
5
Proton Leak:
ApoE
OCR (pmol/min)
*
Female
Male
0
5
10
15
20
25
Maximal Respiration:
Sex
OCR (pmol/min)
ApoE 3/3
ApoE 3/4
ApoE 4/4
0
10
20
30
40
Maximal Respiration:
ApoE
OCR (pmol/min)
Female
Male
0
5
10
15
Spare Respirational Capacity:
Sex
OCR (pmol/min)
ApoE 3/3
ApoE 3/4
ApoE 4/4
0
5
10
15
Spare Respirational Capacity:
ApoE
OCR (pmol/min)
Figure 3.12. Parameters of Mitochondrial Respirational Profile subgrouped by sex and ApoE genotype. There was no
difference between females and males on any of the mitochondrial parameters (Female N= 12; Male N=10). ApoE genotype
had a significant effect on proton leak which is increased in ApoE 3/3 compared to ApoE 3/4 (p<0.05) (ApoE 3/3 N=4; ApoE
3/4 N=11). All bars represent the mean ± SEM.

69
Discussion
Experiments evaluating the effect of whole blood handling prior to reprogramming
indicate that for maximum efficacy, PBMC’s should be isolated and frozen the same day
as blood collection. However, if distance between site of sample collection and
processing precludes this possibility, samples can be shipped overnight at room
temperature before processing. The data also demonstrate that the SepMate isolation
system is not a barrier to reprogramming, an important observation in improving
efficiency of sample processing. Compared the Vaccutainer CPT tube which is subject to
short expiration dates and high costs for the tube alone, the Sepmate system only
requires that blood be collected in a standard K2 EDTA vacutainer, which is readily
available in the clinic setting. Sepmate processing tubes are not subject to expiration
dates and stock would be managed at the processing lab only, not individual clinical trial
sites, both of which should increase efficient use of resources. The success of
reprogramming PBMCs that had been handled in a variety of ways prior to processing
indicates that episomal reprogramming technology, while having a very low efficiency,
has some plasticity within the system which allows for iPSC generation even if samples
are handled in less than ideal circumstances.

Here we’ve demonstrated successful iPSC generation in a significant AD population.
Previous published work in this area has used up to 7 donors. Combined with directed
differentiation to NSC’s, we have developed and refined process that can take primary
somatic cells to the relevant NSC in numbers sufficient for therapeutic evaluation in 9-11
weeks. The variability seen in the characterization of NSC proliferation and
mitochondrial function illustrates the challenges associated cell phenotyping. One of the

70
limitations of this work is that experiments were not designed to truly compare data
between cell lines, but to test therapeutic effect within each cell line individually. Even
with standard protocols in place and a single operator for all the cell work, such inter-
cellular comparisons may not fully reflect true baseline variability.

71
Chapter 4: NSC Proliferation & Allopregnanolone

Introduction:
Having established the iPSC and resultant NSC lines for all Allo clinical trial participants
and evaluated the baseline proliferation, the next objective was to evaluate the efficacy
of Allo to promote cellular proliferation of these cell lines. As detailed more fully in
Chapter 1, the Brinton lab previously demonstrated that Allo promotes neurogenesis,
increasing both the number and survival of newly generated cells and reversing
neurogenic deficits in the triple transgenic AD (3xTgAD) mouse brains
15,18
. In this study,
the working hypothesis was that similarly, in vitro iPSC-derived NSCs treated with Allo
for 24 hours would have increased proliferation.

Materials & Methods
NSC PROLIFERATION
As described in Chapter 3. Cells were treated with Allo on Day 2. Media was
aspirated from all wells. Cells grown in growth factors act as the positive control
and so receive fresh media complete with growth factors. Treatment groups are
without growth factors to avoid proliferation ceiling effects. Media with growth
factors was aspirated and for vehicle and for treatment groups, fresh N2 media
without growth factors was added. Allo or vehicle solution was added 4 hours
later, at same time as EdU, to ensure cells were handled the same as they are in
the mitochondrial assay. 6 wells of cells were used for each treatment condition.
The three treatment conditions were: vehicle (0.001% ethyl-alcohol (EtOH)), Allo

72
10nM, Allo 100nM. 24 hours after treatment and EdU addition, cells were fixed
and processed for flow cytometry analysis as described in Chapter 3.

ALLOPREGNANOLONE PREPARATION
Allo powder was provided by Dr. M.A. Rogawski (University of California, Davis).
Allo was dissolved in 200 proof EtOH (EX0276, Merck Milipore, Billerica, MA),
10mM, and was agitated until Allo was visibly dissolved.

STATISTICS
Statistical analysis was performed in Prism 6.0. For intra-line treatment effects a
student’s t-test was conducted. For inter-line comparisons, pooled groups were
determined to be non-parametric, and so Mann-Whitney test was used. The
difference between groups was considered significant if the p-value was <0.05.

Results
INTRA-LINE NSC PROLIFERATION WITH ALLO TREATMENT
NSCs derived from 22 donors, all of whom completed the Phase I clinical trial,
were evaluated for proliferation after treatment with two Allo doses (10nM and
100nM). An assay was deemed to be successful if the positive control (cells
grown with growth factors) had statistically significantly increased proliferation
when compared to the same cell line grown without growth factors and treated
with vehicle (0.001% EtOH). Of the 22 donors, NSCs derived from 4 participants
did not show a difference in proliferation after 24 hours grown with and without

73
growth factors (109, 204, 303, and 305) (data not shown). Subsequent
differentiations of these iPSC lines were also unsuccessful. It was observed that
generally, these NSCs failed to establish and expand immediately following
completion of dual-SMAD directed differentiation, or failed to proliferate
sufficiently in later passages to provide enough cells for analyses. To determine if
this issue was a problem specific to the iPSC clone selected or due to inter-donor
differences, a second iPSC clone was selection for 109 and 204. Each new iPSC
line was characterized, karyotyped, and differentiated. In the case of 109,
proliferation analyses of NSCs derived from the second clone were also
unsuccessful (data not shown). However, the second clone selected for iPSC line
204 was successful with NSC proliferation in cells growth in growth factors
significantly increased over those grown without. Data from all successful
proliferation analyses is presented in Figure 4.1.

74

75

INTER-LINE NSC PROLIFERATION WITH ALLO TREATMENT
Unexpectedly, overall NSCs did not show an increase in proliferation after 24hr
Allo treatment. 3 of the 19 lines, Participants 103, 301, and 310, were positive
responders and showed a statistically significant increase at both Allo 10nM and
Allo 100nM, averaging a 11.4±2.8% (p<0.001) and 6.9±1.7% (p<0.001) increase
in proliferation respectively versus vehicle treated NSCs. Unexpectedly, a
number of lines responded negatively to Allo treatment (participants 102, 105,
110, 207, 215, and 221) and labelled as negative responders with an average
6.81±1.5% (p<0.001) decrease in proliferation with Allo 10nM and a 7.5±1.7%
(p<0.001) decrease with Allo 100nM versus vehicle treatment (Figure 4.2.A). The
magnitude of the Allo effect was not statistically significantly different between
the positive and negative responding lines, which were 9.1±1.7% and 7.2±1.1%
(p=0.33) respectively (Figure 4.2.B). Subanalyses were conducted for each
therapeutic response by sex. It appears that more cell lines from female donors
respond to Allo treatment, either positively or negatively (Figure 4.2.C). As ApoE
genotype is available so far for 15 of the 19 lines, it is impossible to interpret its
effect on Allo response (Figure 4.2.D). While it is not appropriate to assume a
Figure 4.1. Inter-line NSC Proliferation with Allo treatment (10nM and 100nM).
(+) control are cells grown in EGF, bFGF, and heparin. Cell lines are classes are
positive responder or negative responder based on a significant therapeutic
response on the % of EdU+ N Inter-line NSC Proliferation with Allo treatment
(10nM and 100nM). (+) control are cells grown in EGF, bFGF, and heparin. Cell
lines are classes are positive responder or negative responder based on a
significant therapeutic response on the % of EdU+ NSCs normalized to vehicle.
No significant response to treatment is labeled a non-responder. Bar represent
mean ± SEM (N=4-6). *p<0.05, **p<0.01.

76
correlation, it was observed that of the 6 negative responder cell lines, 4 were
from female donors and all had an ApoE3/4 genotype.
Figure 4.2. Inter-line treatment proliferation data from NSC cell lines classed as negative,
non-, or positive responders. (A) The data indicate that when averaged, the inter-line
averages are consistent with the intra-line classifications for therapeutic response. B) Taking the
average of the two Allo doses, there is not significant difference in the magnitude of the positive
and negative treatment response. C) Breaking down therapeutic response by sex, it appears
that more cell lines from female donors respond to Allo treatment, either positively or negatively.
D) Although N per group is low, there does not appear to be a strong effect of ApoE genotype
on therapeutic response. Bars represent mean ± SEM. *p<0.05, **p<0.01, ***p<0.001.

77

NSC PROLIFERATIVE PHENOTYPE AND ALLO THERAPEUTIC RESPONSE
In Chapter 3, it was observed that the baseline level of NSC proliferation was
highly variable (15.4% - 92.2% EdU + NSCs). When pooled by Allo treatment
response, positive responders have a mean % of EdU+ NSCs of 50.9±7.3, which
is not significantly reduced from the mean of the negative responders [67.2±3.5
(p=0.16)]. However, versus negative responders, non-responders do show a
statistically significant decrease in (+) control NSC proliferation with a mean
baseline proliferation rate of 54.5±3.5 % EdU + NSCs (p<0.05) (Figure 4.3.A).
There is no correlation between baseline (+) control NSC proliferation and the
percentage change in proliferation as a result of Allo treatment (data not shown).

As described above, proliferation assays had two control groups: the (+) control,
growth in growth factors, which served to demonstrate the maximal proliferative
capacity of each cell line. The vehicle treated group was cultured without growth
factors and served to demonstrate a lower proliferative phenotype, the
hypothesis being that Allo doesn’t have a therapeutic effect on cells that are
already proliferating at their healthy, maximal rate. As for the baseline NSC
proliferation in the (+) control group, the difference between the (+) control and
the vehicle treatment proliferation rates was highly variable. When normalized to
the vehicle group, the (+) controls are between 2.5% and 62.5% greater
compared to the vehicle group (Figure 4.3.B). When this difference between
control groups is pooled by Allo therapeutic response, the data indicate that the
non-responder NSC lines have higher differences between the (+) and vehicle

78
controls. This is statistically significant for comparisons with both the positive
responders (p<0.001) and negative responders (p<0.01). As with baseline NSC
proliferation, there is no correlation between this difference in the control groups
and the percentage change in proliferation as a result of Allo treatment (data not
shown).

79
Discussion
Overall, the proliferative results from Allo treatment of NSCs were unexpected. While the
assay was successful in a majority of the cell lines tested, it has proven challenging.
Such challenges and variability is almost never reported in the literature and but based
on discussion with others in the field, not uncommon with iPSC derivatives. The
negative effect of Allo on proliferation has not been observed before in in vitro work done
by the Brinton lab at a similar dose
16
. Additionally, it is unclear why about 50% of the
negative responder lines had significantly reduced proliferation at the lower Allo dose of
10nM but not at 100nM. Characterizations of the Allo dose response curve have
indicated that a higher dose often results in worse outcome compared to lower doses,
but the inverse has not been noted previously. Although previous studies of Allo from
the Bäckström research group have reported increased hippocampal shrinkage
128
and
accelerated AD decline
129
in mouse models, they have exclusively utilized chronic
treatment regimens, administering Allo continuously for up to several months. The
Brinton team on the other hand has determined that the optimal in vivo Allo dosing
Figure 4.3. Association between Baseline Proliferation and Allo response. A)
Baseline NSC proliferation within the (+) control is variable. When pooled by Allo
response, the data indicate that positive responders, though trending, are not
significantly lower than negative responder cell lines. However, non-responders have
significantly lower baseline proliferation within (+) control compared to negative
responders. B) The difference in NSC proliferation between (+) control and vehicle
control groups is also variable between lines. When pooled by Allo response, the data
indicates that non-responder cell lines have a significant increase in the proliferative
difference between controls groups compared to positive and negative Allo response.
Bars represent mean ± SEM.*p<0.05, **p<0.01, ***p<0.001.

80
regimen for neurogenesis is once per week and that more frequently is detrimental to
this outcome
17
. Previous in vitro work in the lab has utilized a 24 hour Allo dosing
regimen, and successfully shown increased proliferation of human fetal NSCs
16
. It is
possible that as these cells are derived from AD donors, they display a phenotype
unique to the disease, exacerbated by the 24hr Allo treatment. As this clinical trial did
not necessitate enrolling cognitively normal control subjects, we did not have a normal
control group to compare with the AD proliferation and Allo dosing response. A future
study could evaluate the effect of various Allo incubation times on NSC proliferation,
mimicking the PK profile seen in our trial participants to determine if this improves NSC
proliferation profiles.

81
Chapter 5: NSC Mitochondrial Function & Allopregnanolone

Introduction
Having established the iPSC and resultant NSC lines for all Allo clinical trial participants
and evaluated the baseline mitochondrial function, the next objective was the evaluate
the efficacy of Allo to promote mitochondrial function of these cell lines. As detailed more
fully in Chapter 1, the Brinton Lab has previously demonstrated that Allo increases
mitochondrial respiration after 24-hour treatment, along an inverted-U dose response
curve. In this study, my working hypothesis was that similarly, NSCs derived from AD
clinical trial participants, treated with Allo for 24 hours would have increased
mitochondrial respiration parameters.

Materials & Methods
NSC MITOCHONDRIAL RESPIRATION
As described in Chapter 3. Cells were treated on Day 2, 4 hours after media was
changed to N2 media without growth factors (to arrest cell proliferation). The
three treatment conditions were vehicle (0.001% EtOH), Allo 10nM, Allo 100nM.
A fourth, no treatment group, was also included. On Day 3, 24 hours after cell
treatment, the assay was run as described in Chapter 3.

AMYLOID-Β MEASURMENTS
Extracellular and intracellular Aβ concentrations were assessed by MesoScale
Discovery V-PLEX Aβ Peptide Panel 1 (4G8) Kit (K15199E-1, Rockville, MD). On

82
Day 1, NSCs were plated at a density of 300,000 cells per well in 1ml of N2
media containing growth factors, on a 6-well plate (25-105, Genesee, San Diego,
CA) coated with Matrigel hESC-qualified matrix (354277, Corning, Corning, NY).
One well per treatment condition. Cells were incubated at 37
o
C/5% CO
2

overnight. On Day 2 media was changed to media without growth factors and
treated. The three treatment conditions were vehicle (0.001% EtOH), Allo 10nM,
Allo 100nM. On Day 3, 24 hours after cell treatment, the supernatant was
removed and centrifuged at 4,000 x g for 10 minutes to removed dead cells.
1250ul of supernatant was then removed to clean 1.7ml microfuge tube and
104ul of 6M guanidine solution [6M guanidine (G4630, Sigma-Aldrich, St. Louis,
MO), 50mM Tris-HCl (T3253, Sigma-Aldrich, St. Louis, MO), Protease and
Phosphatase Inhibitor Mini, EDTA-free (88669, Thermo Fisher Scientific,
Waltham, MA), 0.7μg/ml Pepstatin A (P5318, Sigma-Aldrich, St. Louis, MO) in
diH20, pH to 7.6] was added, to give 0.5M final guanidine concentration. After
supernatant was removed, cells were collected using a cell scraper (353085,
Corning, Manassas, VA), 83µl of 6M guanidine solution was added and cell
solution was transferred to 1.7ml microfuge tube. Tubes with supernatant and
cell suspension were sonicated in a water bath for 15 minutes. 996µl of diH
2
O
was added to the cell suspension tubes to give 0.5M final guanidine
concentration, inverted to mix, and all samples were frozen at -20
o
C. On day of
MSD assay, cells were thawed on ice and Aβ evaluated as per the
manufacturer’s directions.

83
Allopregnanolone Preparation
Allo powder was provided by Dr. M.A. Rogawski (University of California, Davis).
Allo was dissolved in 200 proof EtOH (EX0276, Merck Milipore, Billerica, MA),
10mM, and was agitated until Allo was visibly dissolved.

Statistics
Statistical analysis was performed in Prism 6.0. For intra-line treatment effects a
student’s t-test was conducted. For inter-line comparisons, pooled groups were
determined to be non-parametric, and so Mann-Whitney test was used. The
difference between groups was considered significant if the p-value was <0.05.

Results
NSC derived from the 19 donors, all of whom completed the Phase I clinical trial, and
were successfully evaluated for proliferation, were evaluated for mitochondrial function
after treatment with two Allo doses (10nM and 100nM). Through analysis of extracellular
flux using the Seahorse Platform, I was able to examine functional treatment effects on
basal mitochondrial respiration, cellular ATP production, proton leak, maximal
respiration, and reserve respiratory capacity. As with proliferation, wide variability was
observed in the mitochondrial respiration of untreated NSCs (Chapter 3). For those cell
lines where proliferation evaluation was considered unsuccessful due to lack of a
proliferative difference in cells grown with and without growth factors, mitochondrial data
from these cells lines was excluded from overall analyses. This is because it is
presumed that the lack of proliferative change is indicative of some other underlying

84
difference in the NSCs, which would increase variability between the cell lines evaluated.
Of the five functional outcomes, maximal respiration and reserve respiratory capacity
were the two that showed the most consistent increase in response to Allo 10nm and
100nM treatment. These data from all the successful NSCs lines are presented in Figure
5.1.

85

Figure 5.1. Inter-line NSC Mitochondrial Respiration: Maximal Respiration and
Reserve Respirational Capacity with Allo treatment (10nM and 100nM). Treatment
groups are compared to the vehicle treated group and cell lines are classed are
positive responder or negative responder if one or other of the metabolic parameters
has a statistically significant therapeutic response on the % oxygen consumption rate
as normalized to vehicle. Those with an increase with a p value >0.05 and <0.1 are
Trending Responders. No significant change (p>0.1) with treatment is labeled a non-
responder. Bar represent mean ± SEM (N=4-6). *p<0.05, **p<0.01, ***p<0.001.
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
103
*
Maximal Respiration Reserve Capacity
0
50
100
150
200
220
OCR (pmol/min)
(Normalized to Vehicle)
110
*
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
201
*** **
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
204
*
Maximal Respiration Reserve Capacity
0
50
100
150
200
400
OCR (pmol/min)
(Normalized to Vehicle)
221
*
*
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
307
***
***
***
**
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
310
*** ***
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
312
*
*
Positive Responders
Negative Responders
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
102
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
104
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
105
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
111
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
207
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
215
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
217
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
219
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
301
Non-Responders
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
112
**
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
201
*** **
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
203
** *
*
Trending Responders
Maximal Respiration Reserve Capacity
0
50
100
150
200
OCR (pmol/min)
(Normalized to Vehicle)
312
*
*
Vehicle No Treatment Allo 10nM Allo 100nM

86
Individual participant NSC lines were considered positive or negative responders if there
was a statistically significant increase or decrease in treatment versus vehicle, with a
p<0.05. Results were considered to be trending if the p was <0.1. Once a cell line was
classed as a positive responder, trending positive responder, non-responder, or negative
responder, the data normalized to vehicle, was pooled to determine average treatment
effects. The pooled data were then analyzed again for overall evaluation of treatment
effect.

4 out of 17 lines showed a statistically positive increase in maximal capacity after 24-
hour treatment versus vehicle treatment, with a 16.9±4.6% (p<0.001) and 31.9±3.2%
(p<0.001) average increase in OCR after 10nM and 100nM Allo treatment respectively.
An additional 4 lines trended towards positive effects, with Allo 100nM inducing an
average 19.9±4.6% increase in maximal respiration (p<0.001). 11 NSC lines had no
significant changes in maximal respiration in response to Allo treatment (Allo 100nM
100.1±2.9). As was observed during proliferation analyses, some negative effects of Allo
treatment were observed. Only one cell line (203) demonstrated a statistically significant
negative response to treatment on maximal respiration, reducing maximal respiration by
31.7±3.8% after 10nM treatment and by 28.1±3.4% (p<0.001) after 100nM treatment
(Figure 5.2.A.). As with the effect of proliferation after Allo treatment, comparing the
magnitude of the positive response versus the one negative response to Allo, there is no
statistically significant difference (Figure 5.2.C). Evaluating the number of participant cell
lines in each therapeutic response by sex, there does not appear to be an effect of sex
on Allo response and maximal respiration (Figure 5.2.E). As ApoE genotype is available

87
so far for only 15 of the 19 lines, it is impossible to interpret its effect on Allo response as
it relates to maximal respiration as yet (Figure 5.2.G).

When evaluating the reserve respiratory capacity, 6 out of the 17 participant NSCs lines
had a statistically significant increase after 24-hour Allo treatment versus vehicle
treatment, with an average increase of 43.3±8.8% increase with 10nM Allo and
47.4±11.5% increase with 100nM Allo (Figure 5.2.B). The two lines that trended towards
an increase (201, 215) did so only at the higher dose with a mean increase of 47.6±18%
(p<0.1), nearly identical to the positive responder lines, except with larger SEMs. When
included with the non-responder group, there was still no significant effect of treatment
and as a result, these two cell lines were no graphed separately. Interestingly, these two
lines also showed decreases in reserve capacity at the lower dose with a mean
decrease of 29±7.1%. 8 NSC lines had no significant changes in reserve capacity as a
result of Allo treatment, with a mean increase of 3.6±6% after Allo 100nM versus vehicle
(p=0.74). Only three lines (112, 201, 203) demonstrated a statistically significant
decrease in reserve capacity after Allo treatment, averaging 37.9±5.3% and
10.49±10.2% with Allo 10nM and Allo 100nM, respectively. As in all other analyses,
when comparing the magnitude of the effect between the positive and negative Allo
responses, there is no statistically significant difference (Figure 5.2.D). Evaluating the
number of participant cell lines in each therapeutic response by sex, there does not
appear to be an effect of sex on Allo response and spare respirational capacity (Figure
5.2.F). As ApoE genotype is available so far for only 15 of the 19 lines, it is impossible
to interpret its effect on Allo response as it relates to spare respirational capacity as yet
(Figure 5.2.H).

88

89

NSC MITOCHONDRIAL PHENOTYPE AND ALLO THERAPEUTIC RESPONSE
In Chapter 3, it was observed that the baseline level of NSC maximal respiration
and spare respirational capacity (i.e. mitochondrial assessment of cells with no
treatment) was highly variable. When pooled by Allo treatment response, the
data indicate that positive responders have a mean baseline OCR of 22.9.3±6.5
pmol/min, which is not significantly elevated over the mean of the non-
responders (16.3.2±2.3 pmol/min) (p=0.37) (Figure 5.3.A, 5.3.C). There is no
correlation between baseline, no treatment NSC maximal respiration and the
percentage change in maximal respiration after Allo treatment (data not shown).

For spare respirational capacity, when pooled by Allo treatment response, the
data indicate that positive responders have a mean baseline OCR of 11.3±3.3
pmol/min, which is not statistically different from the non-responders (7.2±1.3
pmol/min) and negative responders (5.7±1.9 pmol/min) (p=0.36 and p=0.48
Figure 5.2. Inter-line Treatment Mitochondrial Respiration from NSC cell
lines classed as negative, non-, or positive responders. Inter-line treatment
Maximal Respiration (A) and Spare Respirational Capacity (B) data from NSC
cell lines classed as negative, non-, or positive responders were averaged. The
data indicate that when averaged, the inter-line averages are generally consistent
with the intra-line classifications for therapeutic response. An additional
classification of “trending positive responder” was added to the evaluation of
maximal respiration and includes those cell lines that had a p<0.1 (N=4 NSC
lines). When included in the non-responder average, the Allo 100nM dose was
significantly increased compared to vehicle. Taking the average of the two Allo
doses, there is no significant difference in the magnitude of the positive and
negative treatment response for maximal respiration (C) or spare respirational
capacity (D). Breaking down therapeutic response by sex (E, F) and ApoE status
(G, H) there does not appear to be any relation with Allo response. Bars
represent mean ± SEM. *p<0.05, **p<0.01, ***p<0.001.

90
respectively) (Figure 5.3.B, 5.3.D). There is no correlation between no treatment
spare respiration capacity and the percentage change in spare respirational
capacity after Allo treatment (data not shown).

As there were some observed differences between the maximal respiration and
spare respirational capacity of NSCs with no treatment versus those cells treated
with the vehicle only, we also wanted to determine if therapeutic response was
related to how the cells responded to the vehicle alone. As before, the data were
pooled by Allo treatment response and indicate that positive responders have a
mean vehicle treatment OCR of 20.2.3±5.4 pmol/min, which is not significantly
elevated over the mean of the trending responders (13.4±6.3 pmol/min; p=0.2),
non-responders (15.9±2.2 pmol/min; p=0.5), or negative responder (15.04
pmol/min) (Figure 5.3.E, 5.3.G). There is no correlation between vehicle treated
NSC maximal respiration and the percentage change in maximal respiration after
Allo treatment (data not shown).

For spare respirational capacity, positive responders have a mean vehicle
treatment OCR of 9.3±3.3 pmol/min, which is not significantly elevated over the
mean of the non-responders (7.2±1.1 pmol/min, p=0.6) or the negative
responders (6.4±1.8 pmol/min; p=0.6) (Figure 5.3.F, 5.3.H). There is no
correlation between vehicle treated NSC spare respiration capacity and the
percentage change in spare respiration capacity after Allo treatment (data not
shown).

91
NSC proliferation was directly correlated with to these two mitochondrial
outcomes at both Allo 10nM and 100nM to determine if there was any link
between the treatment effects of the disparate outcomes. Correlation analyses
demonstrated that there is no correlation between proliferation and either
maximal respiration or spare respirational capacity at either Allo dose. The r
values ranged from -0.3 to 0.09 (data not shown).

92

93

NSC AΒ CONCENTRATION
Intracellular and extracellular Aβ concentrations were evaluated in NSCs from a
subgroup of 12 clinical trial participants after two doses of Allo, 10nM and 100nM.
There were no measurable levels of intracellular Aβ from 300,000 cells.
However, 24 hour extracellular Aβ40 and Aβ42 species were measured for all
cell lines. Extracellular Aβ38 was not measureable. The Aβ42/40 ratio was not
altered from vehicle by 24 hr Allo treatment at either dose (Figure 5.4.A). No
analysis of genotype was done because of the 12 participants evaluated for Aβ,
only 8 (from cohorts 1 and 2) had been genotyped for ApoE, and all were ApoE 4
carriers. There was no effect of sex on the Aβ40:42 ratio (Figure 5.4.B.).

Figure 5.3. Association between Baseline and Vehicle Mitochondrial
Function and Allo Response. Association between NSC Maximal Respiration
with no treatment (A) and vehicle (E) as well as NSC spare respirational capacity
with no treatment (B) and vehicle (F). Baseline NSC maximal respiration in no
treatment and vehicle treated groups is variable. When pooled by Allo response,
the data indicate that positive responders with no treatment trend towards a higher
maximal respiration (C) and reserve respirational capacity (D), but does not reach
significance. Pooled Allo responses indicate that there is no difference between
Allo response groups treated with vehicle in maximal respiration (G) or reserve
respirational capacity (H). Bars represent mean ± SEM.*p<0.05, **p<0.01,
***p<0.001.

94
Discussion
Compared to the proliferative effects of Allo on the NSCs, the data presented here on
the effects of Allo on the maximal respiration and spare respirational capacity
parameters of mitochondrial respiration, is equally unexpected. We would not have
hypothesized that the NSCs derived from iPSC would demonstrate stronger, more
consistent positive in vitro effects in these mitochondrial outcomes when compared to
Vehicle
Allo 10nM
Allo 100nM
0.0
0.1
0.2
0.3
0.4
Aβ40:42 ratio
Overall Allo Effect
Vehicle
Allo 10nM
Allo 100nM
Aβ40
Aβ42
0
5
10
15
20
25
Conc. (pg/ml)
Aβ42/40
0.0
0.1
0.2
0.3
0.4
Aβ42/40 Ratio
Males
Females
0.0
0.1
0.2
0.3
0.4
By Sex
Aβ42/40 Ratio
Veh Allo 10nM Allo 100nM
A
B
Figure 5.4. Effect of Allo on extracellular amyloid-b levels. A) There is no significant
effect of 24 hour Allo treatment on levels of extracellular amyloid-beta species (N=12 NSC
lines). B) Broken down by sex, there is no significant differences in the Ab 42:40 ratios
between males and females (N=6 each sex).

95
NSC proliferation. We hypothesize that one possible explanation for the improved
mitochondrial response is that these cells are recapitulating a very early stage of AD,
where NSC proliferation is not yet impaired (and so cannot be improved upon), but the
mitochondria are already impaired and thus there is a window for Allo treatment
response.

Based on the experiments and analyses presented here, it is as yet unclear, why
mitochondrial parameters are improved in some cell lines with Allo treatment and others
are not. One hypothesis is that the cell lines are demonstrating varying levels of an AD
phenotype that are not detected by the assays in this study and Allo is only efficacious in
those with the “correct” phenotype, but whether that responder phenotype is of mild vs
moderate disease is the focus of future studies. Such studies should utilize cognitively
normal PBMC donors, whose iPSC and NSCs are derived using the protocols described
here.

In this cellular model of NSCs from AD donors, it is possible to reliably evaluate
extracellular Aβ concentrations of the Aβ40 and Aβ42 species and to calculate the ratio
between the two. The lack of an Allo effect to change the concentration of any of these
measurements is not totally unexpected. During in vivo studies, Allo treatment was only
associated with a decrease in Aβ pathology after chronic treatment of 6 months
17
. As
Allo is not known to target Aβ directly, the mechanism behind such a decline is as yet
unexplained. Here, NSCs, not neurons, were treated once with Allo for 24 hours. Future
studies on neurons could utilize treatment regimens similar to the clinical paradigm, i.e.

96
once per week for several weeks, and then evaluate Aβ levels to determine effects of
Allo.

Other questions regarding the effects of Allo on mitochondrial function are the focus of
research within the Brinton lab and the neurosteroid field more broadly; specifically
through what mechanisms does Allo have an effect on these specific outcomes? The
known mechanism through the GABA
A
receptor complex has been established for the
neurogenic effects of Allo, but is less plausible for mitochondrial specific effects. This,
along with evidence for effects on Ab and cholesterol homeostasis, lends support to the
hypothesis that Allo acts as a systems biology regulator
12,13
.

97
Chapter 6: In vitro and Clinical Correlations

Introduction
Having completed in vitro analyses for 22 out of the 24 participants in the Allo Phase I
clinical trial, we wanted to determine if this data had any strong correlations with the
clinical outcomes evaluated. At this time, the first two dosing cohorts have been fully
completed, facilitating analysis of pre- and post- dosing analyses of MRI scans and
cognitive tests. The third and final cohort, testing the highest doses of Allo, is almost
complete with anticipated study closeout in approximately February 2018. At the outset
of in vitro work, we hypothesized that iPSC-NSC proliferation with Allo treatment would
correlate with MRI volumetric measures, which themselves would serve as a clinical
surrogate measure of neuroregeneration. It is important to note that as blood samples
for iPSC generation were drawn at baseline, the in vitro cell culture work was agnostic to
the donor’s study arm randomization. This allowed for an unbiased in vitro evaluation of
a participant’s treatment response status. Correlational analyses were conducted on all
participants with complete data to evaluate overall effects of treatment, and then on
population subsets to determine if correlations within a dosing cohort varied due to the
effects of the higher dose on clinical outcomes or if specific ApoE genotypes have an
effect on the ability to use in vitro data to predict the clinical outcome.

98
Materials and Methods
As the entire study was not yet complete, all data remained blinded and the end-of-
cohort correlational analyses were completed by the clinical biostatistics team headed by
Wendy Mack, PhD, in USC’s Department of Preventative Medicine. Proliferative and
mitochondrial (maximum respiration and reserve respirational capacity) data normalized
to vehicle, was sent to the biostatistics team, for each individual participant. The data
was a percentage change due to treatment and did not indicate if it was a statistically
significant change. The in vitro data was NSC Proliferation after 10nM Allo and 100nM
Allo, NSC Mitochondrial Maximum Respiration after 10nM Allo and 100nM Allo, and
NSC Mitochondrial Reserve Respirational Capacity after 10nM Allo and 100nM Allo. The
clinical outcomes used in these correlational analyses were MRI volumetric analyses
(left and right: total hippocampal volume, hippocampal tail, subiculum, hippocampal
fissure, presubiculum, parasubiculum, molecular layer HP, granule cell and molecular
layers of the dentate gyrus (GCMLDG), Cornu Amonis (CA) 3, CA4, fimbria,
hippocampal amygdala transition area (HATA)) and cognitive analyses (ADAS-Cog 13-
minute battery, MOCA, five CogState domains). Initial Spearman Rank Correlation
analyses were conducted with 1) all participant data regardless of treatment are, and 2)
only the participants randomized to the treatment arm in both cohorts 1 and 2. The latter
necessitated pooling data from participants who had been clinically treated with two
different Allo doses (2mg and 4mg). Further correlational analyses were done on
subgroups of participants in order to understand refined treatment effects on discrete
populations, i.e. ApoE genotypes, clinical dosage groups, and sex.

99
Statistical Analysis
Correlational analysis was performed to explore relationships between mitochondrial
parameters (maximal respiration, spare respirational capacity) and changes in
volumetric MRI hippocampal measurements and cognitive tests. Spearman's rank
correlation coeffcients are reported. Tests of significance were not conducted due to
small sample size. Analysis was run using SAS software v9.4.

Results
CLINICAL DATA OVERVIEW
Blinded clinical data from dosing cohorts 1 and 2 was summarized by overall
positive (+) and negative (-) effect (Table 6.1.A; 6.1.B.). The most apparent
improvement appeared to be in the left total hippocampal volume where 8 out of
the 11 participants to receive Allo showed an increase in volume from baseline,
compared to only 1 out of 4 placebo arm participants. Other hippocampal
volumes that appeared to be increasing included right GCMLDG and the right
and left parasubiculum. There was not a strong positive effect in these first two
cohorts on cognitive test score improvement, although the Cog State detection
and one card learning domains may prove promising in Cohort 3. If, within an
individual participant, the number of improved outcomes is totaled (35 data points
per participant), there is no statistically significant difference between either of
the cohorts versus the placebo arm (p= 0.85) (Table 6.1.C).

100
CORRELATIONAL ANALYSES
Correlational analyses were conducted to determine if in vitro data was predictive
of a clinical change after Allo treatment. Only data from Allo treated participants
were included in the analyses, because, if the hypothesis holds true, placebo
treated participants will have negative clinical outcomes and positive in vitro
outcomes, which will reduce the overall strength of any positive correlations
among the treated participants. All the Spearman correlation coefficients are
presented in Table 6.2 and Table 6.3. Due to errors in the MRI scans for two
participants, volumetric MRI results are based on data from 8-9 participants,
while there is complete cognitive data for all 10 Allo treated participants. With 35
clinical and 6 in vitro outcomes, there are 210 possible correlations. Of these, 10
have a positive correlation with r > 0.5. Several have a negative correlation with r
< -0.5.

101

102

Table 6.1. Clinical Volumetric MRI and Cognitive Test Changes after 12 Weeks of Allo. Blinded clinical data
indicating a participant's changes on volumetric MRI parameters (A) and cognitive testing (B) from baseline after 12
weeks of treatment. For hippocampal volumes, positive (+) indicates an increase in volume from baseline and
negative (-) indicates a decrease in volume from baseline.L= left, R = right. For cognitive tests (+) indicates
improvement in score over study visits and (-) indicates worsening of score over study visits. CogState tests: DET =
detection; ID = identificaiton; OCL = one card leaning; OBS = one back speed; OBA = one back accuracy. The
number of participants showing improvement for each parameter are totaled at the bottom. (C) The number of positive
outcomes for each participant. *Subject 1 is missing HC subfield data; Subject 2 is missing all MRI data.

103
IN VITRO PROLIFERATION & CLINICAL MRI CORRELATIONS
Surprisingly and in exact opposite direction to what we hypothesized, the
correlations between the clinical outcomes and NSC proliferation at both Allo
doses have strong negative correlations, especially at the higher Allo dose,
where 10 of the MRI outcomes have a r < -0.5. There is one positive correlation
between 10nM Allo NSC Proliferation and molecular layer HP (r = 0.88) (Table
6.2. At Allo 100nM however, the r drops to 0.17. When evaluating the MRI data in
isolation, the molecular layer after treatment does not have strong positive
changes versus placebo. It remains to be seen if this holds true in the highest
dosing cohort. Correlation with left total hippocampal volume, the most promising
measurement clinically, is negative at both Allo doses, with r = -0.57 for Allo
100nM and r =-0.72 for Allo 10nM (Figure 6.1).
r = -0.57 r = -0.73
A. B.
Figure 6.1. . Correlations between NSC Proliferation and Left Hippocampal
Volume. Spearman Correlation between Clinical MRI Total Left Hippocampal
Volume Change from Baseline to post-treatment and NSC Proliferation % change
with Allo 10nM (A) and Allo 100nm (B). Correlations were conducted using data
from participants in Allo Cohort 1 and Allo Cohort 2 only (N=9).

104
Table 6.2. Spearman Correlations for Allo Participants Only. Spearman
Correlation Coefficients between MRI volumetric parameters and in vitro NSC
outcomes in participants randomized to the Allopregnanolone active arm only.
Coefficients >0.5 are indicated in pink and coefficients <-0.5 are indicated in
purple. * ADAS-Cog: Negative slopes indicate improvement in scores. MOCA and
CogState: positive slops indicate improvement in scores.

105

N
NSC
Proliferation
Allo 100nM
NSC
Proliferation
Allo 10nM
NSC
Maximal
Respiration
Allo 100nM
NSC
Maximal
Respiration
Allo 10nM
NSC Spare
Capacity
Allo 100nM
NSC Spare
Capacity
Allo 10nM
L 14 -0.27 -0.46 0.31 -0.07 0.31 -0.20
R 13 -0.12 0.14 -0.19 0.28 -0.02 -0.20
L 12 0.06 0.41 -0.22 -0.08 -0.25 -0.24
R 12 -0.59 0.13 -0.36 0.11 -0.30 -0.16
L 12 0.22 0.35 0.01 -0.21 -0.10 0.01
R 12 -0.49 0.22 -0.22 0.20 -0.22 0.19
L 12 -0.22 -0.32 -0.10 0.03 -0.02 -0.11
R 12 -0.38 0.27 -0.65 -0.08 -0.56 -0.38
L 12 -0.24 0.07 -0.21 0.27 -0.08 0.34
R 12 -0.43 -0.16 -0.17 0.27 0.00 0.09
L 12 -0.23 0.23 -0.78 -0.24 -0.64 -0.46
R 12 -0.21 0.03 -0.17 -0.16 -0.23 -0.57
L 12 0.40 0.13 -0.04 -0.45 -0.15 -0.65
R 12 -0.48 0.40 -0.43 -0.15 -0.49 -0.55
L 12 0.06 -0.38 0.15 -0.28 0.08 -0.31
R 12 -0.14 -0.25 -0.28 0.31 -0.06 -0.09
L 12 0.08 0.47 -0.26 -0.38 -0.36 -0.57
R 12 -0.36 0.05 -0.45 -0.16 -0.41 -0.32
L 12 -0.16 0.20 -0.22 0.21 -0.13 -0.08
R 12 -0.36 0.05 -0.41 0.09 -0.29 -0.19
L 12 0.06 0.22 -0.01 0.01 -0.02 -0.26
R 12 -0.17 0.11 0.24 0.41 0.20 0.03
L 12 -0.10 0.28 -0.16 0.13 -0.14 -0.20
R 12 -0.36 0.08 -0.43 0.02 -0.33 -0.31
L 12 -0.04 0.20 -0.64 -0.22 -0.50 -0.27
R 12 -0.20 0.22 -0.32 -0.55 -0.50 -0.50
L 12 0.03 0.23 -0.06 -0.64 -0.26 -0.54
R 12 -0.67 0.10 -0.15 0.08 -0.15 -0.29
14 0.20 -0.02 0.26 0.22 0.35 0.56
14 -0.31 -0.17 -0.13 0.05 -0.13 -0.35
DET 14 -0.27 -0.31 0.18 -0.09 0.18 0.20
ID 14 -0.07 -0.22 0.61 0.38 0.65 0.71
OCL 14 0.02 -0.46 0.36 0.07 0.34 0.00
OBS 14 -0.08 0.03 0.39 0.15 0.35 0.31
OBA 14 -0.38 -0.22 0.16 -0.12 0.05 -0.06
All Participants
Molec HP
GCMLDG
CA3
CA4
Fimbria
MRI Parameters:
TOTAL HC VOL
HC
SUBFIELD
VOLUME
Whole HC
HC Tail
Subiculum
CA1
Fissure
Presub
Parasub
HATA
COGNITIVE
SLOPES
ADAS-Cog*
MOCA
Cog State
Table 6.3. Spearman Correlations for All Clinical Trial Participants. Spearman
Correlation Coefficients between MRI volumetric parameter and in vitro NSC outcomes
in all participants, regardless of randomization arm. Coefficients >0.5 are indicated in
pink and coefficients <-0.5 are indicated in purple. *ADAS-Cog: Negative slopes indicate
improvement in scores. MOCA and CogState: positive slops indicate improvement in
scores.

106
In Vitro Mitochondrial & Clinical MRI Correlations:
When the correlations with the mitochondrial maximal respiration and spare
respirational capacity were evaluated, there are strong positive correlations with
the left total hippocampus volume at the Allo 100nM dose (maximal respiration r
= 0.64; spare respirational capacity r = 0.61) (Figures 6.2.B, 6.2.D). There was
no correlation at the lower Allo 10nM dose where the r was -0.02 and -0.2 for
maximal respiration and spare capacity respectively (Figure 6.2.A., 6.2.B.). This
is exciting because left total hippocampal volume is the MRI volumetric outcome
that to date shows the most promise for a clinical biomarker of efficacy (Table
6.1.). There are a number of negative correlations, especially between the MRI
outcomes and spare capacity. However, these also have similar correlations
when all placebo participants are included in the analyses (Table 6.3.), indicating
that such correlations are unlikely due to any treatment effect.

In Vitro Mitochondrial & Cognitive Correlations:
When focusing on the cognitive correlations, while the spare capacity and ADAS-
Cog slope is positively correlated (Table 6.2.), due to the scoring system of this
particular cognitive test where a lower score indicates better cognitive function,
this is interpreted to mean that as the spare capacity improves with treatment in
vitro, ADAS-Cog performance decreases. However, comparing the clinical and
placebo arms it is apparent that there is no effect of treatment on ADAS-Cog,
implying that such a correlation may be erroneous. Overall, it has become clear
that all the data needs to be considered holistically in order to explain the context
of correlational analyses. Having done so, it appears that mitochondrial maximal

107
respiration and spare respirational capacity of participant-derived NSCs treated
with 100nM Allo are somewhat predictive of their clinical response.

Allo 10nM Allo 100nM
Maximal Respiration Spare Respirational Capacity
r = -0.02
r = 0.61
r = 0.64
r = -0.2
A.
B.
C.
D.
Figure 6.2. Correlations between NSC Mitochondrial Function and Left Hippocampal
Volume Spearman Correlation between Clinical MRI Total Left Hippocampal Volume
Change from Baseline to post-treatment and Maximal Respiration % change with Allo 10nM
(A), Allo 100nm (B), and Spare Respirational Capacity % change with Allo 10nM (C) and
Allo 100nM (D). Correlations were conducted using data from participants in Allo Cohort 1
and Allo Cohort 2 only (N=10).

108

Left Total Hippocampal Volume & Mitochondrial Outcomes: Subgroup
Correlations
In order to more fully understand and characterize this therapeutic response
further, we asked a couple of questions that are becoming increasingly relevant
in Alzheimer’s disease drug development:

1) In this initial analysis, we have pooled all the Allo participants. However,
as each cohort of participants receives a higher dose, we hypothesize
that any clinical efficacy will be of a greater magnitude compared to the
lower dose(s). Although the N is small (6) for each cohort, does the
correlation between left total hippocampal volume and mitochondrial
outcomes differ get stronger in the second cohort (4mg) versus the first
(2mg)?

2) What is the effect of ApoE genotype? With ApoE genotypes of the trial
participants as detailed in Chapter 2, we repeated the correlational
analysis between left total hippocampal volume and mitochondrial
outcomes in the ApoE 3/4 Allo treated participants only (N=8). This
allowed us to test our hypothesis that the correlation would increase in
the ApoE 3/4 sub-group. Due to the low N in ApoE 3/3 (N = 3) and
ApoE 4/4 (N=1), analyses on these subgroups was not conducted.

109
3) What is the effect of sex? Again, the small N (4-5) is a significant
limitation. However, such initial analyses will contribute to the overall
picture of therapeutic response and development of an appropriate
biomarker.

Correlational Subgroup: Dosing cohort
A correlation analysis of the left total hippocampal volume and mitochondrial
maximal respiration and reserve respirational capacity with Allo 100nM, was
run within each of the two dosing cohorts. The data indicate that for maximal
respiration there is an increased correlation at the higher clinical dose in
cohort 2 (r =0.6) compared to lower clinical dose in cohort 1 (r = 0.4) (Figure
6.3.A., 6.3.C). For spare respirational capacity the same effect is seen with
Cohort 1 having a weaker correlation (r = 0.4) compared to cohort 2 (r = 0.54)
(Figure 6.3.B, 6.3.D). Neither mitochondrial outcome has relevant difference
in correlations between cohorts with Allo 10nM treatment. This proves our
hypothesis that a higher dose resulting in a greater clinical effect, has a
stronger correlation with in vitro data due to the reduced clinical variability
across dosages. See below for a note on clinical vs in vitro dosing. As this
strengthens the credence of the correlation between these in vitro and clinical
data, we hypothesize further that the correlation will increase again in the
highest dosing cohort (Cohort 3, 6mg-14mg).

110

Maximal Respiration
Spare Respirational
Capacity
Cohort 1
Cohort 2
r = 0.4 r = 0.6
r = 0.4 r = 0.54
A.
B.
C.
D.
Figure 6.3. Spearman Correlations Between NSC Mitochondrial Function and Left
Hippocampal Volume by Dosing Cohort. Spearman Correlation between Clinical MRI
Total Left Hippocampal Volume Change from Baseline to post-treatment and Maximal
Respiration % change with Allo 100nm (A) and Spare Respirational Capacity % change
(B) in Cohort 1. Coefficients increased slightly in Left Hippocampal Volume change from
baseline and Maximal Respiration % change with Allo 100nM (C) and Spare
Respirational Capacity (D) in Cohort 2. Cohort 1 N=4, Cohort 2 N=6.

111
Correlational Subgroup: ApoE genotype
A correlational analysis of the left total hippocampal volume and
mitochondrial maximal respiration and reserve respirational capacity with Allo
100nM, was conducted on data from only participants with an ApoE 3/4
genotype. Due to the small number of 3/3 participants treated with Allo (N=3),
it wasn’t possible to run a separate correlation on this group. The data
indicate that there is a slight increase in correlation for maximal respiration
with Allo 100nM in the ApoE3/4 genotype, where r increased from 0.64 to
0.67. Spare respirational capacity after Allo 100nM in the ApoE 3/4 subgroup
increased more substantially, with the r increasing from 0.61 to 0.71 (Figure
6.4). There is no change in correlation with ApoE genotype in these
mitochondrial parameters after Allo 10nM treatment. Additionally, in
evaluating the placebo arm data is observed that 3 of these participants with
the ApoE 3/4 genotype had a decreased total left hippocampal volume at the
end of the study but a positive increase in mitochondrial outcomes (Figure
6.4.). Interpreted in the light of the above correlational data, this supports our
hypothesis that ApoE 3/4 genotypes are most likely to respond positively to
Allo treatment. This is exciting as it indicates it may be possible to enrich
future clinical studies of Allo in AD with participants who are ApoE4 positive.

112

Correlational Subgroup: Participant Sex
A correlational analysis of the left total hippocampal volume and
mitochondrial maximal respiration and reserve respirational capacity with Allo
100nM, was conducted on data from the female participants randomized to
Allo and compared to that of the male participants randomized to Allo. The
data indicate that for maximal respiration, the correlation increases in females
compared to males (r = 0.8 vs r = 0.6) (Figure 6.5.A, 6.5.C). In spare
respirational capacity, the correlation widens further with females having a r =
0.8 and males with an r = 0.5 (Figure 6.5.B, 6.5.D). This difference in the
correlation between the clinical and in vitro outcomes is due to the sex of the
participant. This is important when considered along with the data presented
in Chapter 4 where NSC proliferation from male participant cell lines trended
r = 0.67 r = 0.71
Figure 6.4. Spearman Correlations Between NSC Mitochondrial Function and Left
Hippocampal Volume by ApoE Genotype. Spearman Correlation between Clinical
MRI Total Left Hippocampal Volume Change from Baseline to post-treatment and
Maximal Respiration % change with Allo 100nm (A) and Spare Respirational Capacity %
change (B) in Participants with the ApoE 3/4genotype only (N=8). Only participants
randomized to Allo were included in correlational analyses.

113
lower compared to female cell lines, but never reach significance. There is no
effect of sex on maximal respiration or spare respirational capacity. At this
stage, it is unclear if the increase in female mitochondrial and clinical
correlations then is truly due to more accurate patient modeling in female
patient cell lines. It’s an important observation to consider in the analysis of
Cohort 3, where sex differences in the levels of sedation with equivalent
doses have been observed. The difference has resulted in the women
receiving a higher average Allo dose compared to the men.

Drug Dosing: In vitro vs Clinical
It is important to note that in vitro Allo doses of 10nM and 100nM were selected due to
previous data from rat, mouse, and human NSCs and before clinical PK data from the
Phase I study was available. Cohort 2 (4mg) PK data revealed a Cmax of
125.48±19.13nM, which is close to our most effective in vitro dose of Allo 100nM. Direct
comparisons are limited by the time spent at this concentration as the 125nM level is
sustained only very briefly during the IV infusion (~15 mins) and Allo is completely
cleared within 4 hours (AUC 31.06±7.69 ng.hr/ml). In vitro, the 100nM concentration is
maintained for 24 hours.

114
Maximal Respiration
Spare Respirational
Capacity
Female
Male
r = 0.8 r = 0.6
r = 0.8 r = 0.5
A.
B.
C.
D.
Figure 6.5. Spearman Correlation Spearman Correlation between NSC
Mitochondrial Function and Left Hippocampal Volume by Sex. Clinical MRI Total
Left Hippocampal Volume Change from Baseline to post-treatment and Maximal
Respiration % change with Allo 100nm (A) and Spare Respirational Capacity % change
(B) in female participants randomized to Allo (N=5). Coefficients decreased slightly in
Left Hippocampal Volume change from baseline and Maximal Respiration % change
with Allo 100nM (C) and Spare Respirational Capacity (D) in males only (N=5).

115
Chapter 7: Discussion

The work presented here describes a high-risk research project, which aimed to bring a
new technology into the lab and to apply it translationally. The data allow us to reject our
null hypothesis by demonstrating that the effect of Allo on iPSC-derived NSCs correlates
with clinical changes in neuroimaging after treatment. In order to more fully understand
how this work contributes to the bigger picture, it is best to consider the two key
components individually: 1) the iPSC-based model and 3) the in vitro/clinical
correlations.

iPSC Model of SAD and Allopregnanolone Response
The first aim of the project was to establish a process for efficiently and reproducibly
generating iPSCs in a manner that was scalable for this and future clinical trial settings.
We were encouraged by the wide-patient acceptability for stem cell research. It was
included as part of the overall clinical trial informed consent process, but required a
separate signature, which was given by 100% of the study participants. Our evaluation
of the effects of sample storage and processing demonstrates the feasibility of
expanding somatic cell collection for iPSC studies to multi-site, clinical trials, requiring
standard collection tubes and no special shipping conditions. Reprogramming and iPSC
evaluation has been refined to minimize the time and resources required to ensure
quality iPSC lines have been established. In order to establish a relatively high
throughput protocol, we included a limited number of cell (iPSC and NSC)
characterization methods. This meant that following clonal isolation and expansion, cells

116
were determined to be karyotypically normal and pluripotency was based on morphology
followed by live and fixed cell imaging (Tra-1-60, Nanog, and Tra-1-81). After
differentiation, cells were characterized as NSCs based on fixed cell staining for nuclear
sox-2 and cytoplasmic nestin. While our NSC characterization is comparable to a
number of other studies
55,87,107
, our approach to pluripotency characterization is more
minimalist, not incorporating teratoma formation, germline staining, or gene expression
profiling. While this may be seen by some to be a major limitation, it allowed for a more
efficient, streamlined process from somatic to NSC. This approach has proven
successful and more than 85% of participant cell lines were evaluated for an Allo
response using the first iPSC clone characterized. This is not to minimize the importance
of cell characterization, but to illustrate how time and resources can be managed to
facilitate large-scale applications.

Addressing the ~15% of cell lines that were characterized for proliferation at baseline but
subsequently failed to show a difference between the (+) and vehicle controls therefore
impeding a successful characterization of the Allo response, we hypothesize that a more
complete iPSC characterization may have prevented such issues. This is supported by
observations for line 204, where the first iPSC clone failed to result in NSCs that could
be characterized. However, the second iPSC clone, which was seemingly identical to the
first based on our protocols, differentiated to successful NSCs. On the other hand, in the
assessment of line 109, a second iPSC clone was also unsuccessful. To more fully
investigate this, a second clone will be assessed for any unsuccessful cell lines from the
third dosing cohort. Comparing full characterizations (including teratoma formation,
germline staining etc), of iPSC lines from the first and second clones in these

117
troublesome lines would determine if inadequacies in the iPSC lines explained the
differences between differentiated cells. Then, if the 15% error rate is deemed too high,
additional iPSC characterization methods could be added to ensure more appropriate
iPSC lines are selected prior to differentiation.

As outlined in Chapter 3, the baseline phenotypes were variable. However, because this
study was designed to evaluate drug effects within cell lines, it is difficult to determine if
baseline differences are due to true participant variation or inter-experimental differences
as most cell lines were processed at different times. Encouragingly, such variation does
not appear to impact the overall effect of Allo on maximal respiration and spare
respirational capacity. This is an important observation, demonstrating that even with
variability between assays, it is still possible to detect a relatively accurate in vitro
therapeutic response. This is the first time a research group has reported on the
mitochondrial function of AD iPSC-derived neural cells. Even in the absence of
demonstrating an AD-associated phenotype, this is an important contribution to the
current paucity of literature reporting on iPSC-models of sporadic AD.

One of the most obvious strengths of this study is the number of somatic cell AD donors.
Reporting data on 22 derived NSCs lines, this study is the largest on AD-derived iPSCs,
with previous publications reporting on cell lines derived from 1 – 7 donors (average: 3;
Table 1.2). It is widely acknowledged that developing patient lines is highly valuable in
so much as such cell line resources are available to other researchers. To this end, all
the clinical trial participant derived iPSC lines will be available though the University of
Arizona Center for Innovation in Brain Science biorepository. Another strength of this

118
experimental work is that it was all conducted by one person, thereby limiting any inter-
operator variability. In future studies, it will be important to demonstrate reproducibility
with other personnel, to ensure generalizability of the data and protocols in a broader
context.

We experienced the usual limitations of all iPSC-based studies: immature phenotype of
the NSCs, indeterminate sub-types of NSCs, the lack of other neural cell types, and the
reset of participant epigenetics that may be contributing to their disease status.
Depending on the applications iPSCs are being developed and differentiated for, one of
the major limitations with any iPSC model is the length of time from somatic cell
collection to a readout from the terminally differentiated cell type of interest. In our
hands, assuming the best-case scenario, an in vitro Allo response is available 9 to 10
weeks after somatic cell collection. iPSC technology is still relatively new and this time
lag is widely acknowledged
130
. There are a number of potential workarounds, including
testing Allo on a primary somatic cells from patients (fibroblasts or PBMCs) or direct
reprogramming of somatic cells into NSCs
131-134
, thus bypassing the time required to go
through an iPSC intermediary. The reports of both disease phenotype and therapeutic
response in blood, fibroblasts and undifferentiated iPSCs have been mixed
105,106
. Each
approach has it’s own pros and cons which would need to be balanced against the final
aims of any project, but none obviously preclude these from consideration for modeling
the Allo response in vitro.

One of the other limitations of this study was the lack of cognitively healthy control
donors. Without iPSC lines reprogrammed, differentiated, and evaluated using the

119
protocols detailed above, it is not possible to demonstrate a distinct AD phenotype.
While we report on the variability of the proliferative, mitochondrial, and Ab phenotypes
observed, this was to be expected based on the work of others describing phenotypic
variability in sporadic AD iPSC derivatives
60,61,68,72,89
. This study was designed not to
demonstrate an AD phenotype in NSCs and to compare across donors, but to look
within the cell line to determine a therapeutic response. However, if even a small control
lines could be added, it would be possible to go back and conduct such a study. The
number of control lines would depend on the number of controlling variables (age, ApoE
status, sex), but based on the use on controls in the literature could be as few as 2-3
cognitively normal donors.

One of the most important questions raised by this data is what is the mechanism by
which Allo is increasing mitochondrial function in the responder NSCs? Allo’s diverse
treatment effects in AD models, which besides neurogenesis include in vivo reduction of
Ab and improved cholesterol homeostasis, indicate that it acts beyond the GABA
A

complex, possibly as system’s biology regulator
17,38,135
. However this iPSC-based
method is not a systems biology model, indicating that the mitochondrial improvements
induced by Allo must be due to a direct effect within the cell. The next stage of this
project is to investigate such potential mechanisms, which we hypothesize include
increased mitochondrial biogenesis, by mitochondrial copy number and RNA-Seq.
Advancing our understanding of the mechanism will facilitate understanding of potential
biomarkers of response or to identify responders as well as novel analogues of Allo with
optimized physicochemical properties.

120
To summarize, the proposed next steps in this project have a number of aims. 1) to
determine if fibroblasts, PBMC’s, or iPSCs from sporadic AD donors have a similar
mitochondrial response with Allo treatment, 2) to investigate the mechanisms underlying
Allo’s effects on mitochondrial respiration in iPSC-NSCs, and 3) in future iterations of
this study, to incorporate a panel of cognitively healthy controls, in order to establish an
AD phenotype in sporadic AD derived neural cells.

In vitro - clinical correlations
The first portion of the study, focusing on the in vitro effects of Allo and the sporadic AD
NSC established that the model can be used to identify a treatment response. However,
the ability of the model to translate clinical efficacy was the focus of the second portion.
Published work, linking the effects of treatment on iPSC-derived cells to a clinical
response has been limited, with less than half a dozen published reports
57,105-108
. Three
out of the five are in neurological indications (SMA, IEM, and Bipolar) but did not include
AD. Our study is the first to demonstrate that an in vitro therapeutic response can be
correlated with improvements in MRI imaging in an AD population. As before, our study
characterized a greater number of patients than the other therapeutic in vitro-clinical
studies (which ranged from 1-8 donors). This is advantageous not only because of the
variability reported previously in sporadic AD iPSC work, but also due to the statistical
requirements for responder/non-responder studies
136
. Here, we were limited to 24
donors, the required number of participants for the clinical trial, which was powered for
safety as the primary endpoint rather than efficacy. Future clinical studies of efficacy will
require a much greater number of patients, and if powered primarily for a responder

121
analysis, will require more again
137
. Described simply, this is due to the loss of statistical
power that occurs when a continuous variable (% increase in an outcome versus no
treatment) is dichotomized into a binary variable (responder versus non-responder).
Obviously, many more complicated factors are taken into account during such power
calculations including the mean difference between the two groups, the selected cut-
point above or below which a study participant is deemed a responder or not, and the
response rate in control and active study arms
136
. Regardless, the N is always higher
than what is required for a more traditional statistical analysis of clinical response. It was
rewarding to determine that, comparing in vitro mitochondrial data from the first two
dosing cohorts with their changes on volumetric MRI that an outcome, namely the
change in left total hippocampal volume, popped out with a Pearson correlation
coefficients greater than 0.5, for both maximal respiration and spare respirational
capacity when treated with Allo 100nm. So, what does this positive correlation actually
represent? It means that the mitochondrial response induced by Allo, whether positive,
neutral or negative, is similar to the change in hippocampal volume after 12 week of Allo
infusions. In those participants for whom clinical and in vitro data obviously fail to
correlate (Figure 7.1), this doesn’t necessarily mean that Allo will be unsuccessful, but
rather raises the question, why the discrepancy? One should consider the breakdown of
the types of error the cell model, i.e. is it falsely identifying participants as responders
when they show a clinical decline (Type I error, false positive) or is it indicating a decline
in mitochondrial function when they have an increase in left hippocampal volume (Type
II error, false negative). It appears that the iPSC model have a mix of the two, with very
few obvious errors, but rather more variation in magnitude of treatment effect. This may
be an indication of the level of accuracy, sensitivity, and specificity of the Seahorse

122
Extracellular flux assay, which is an important consideration when determining the most
appropriate biomarker development strategy. The correlational data can be evaluated
either to identify responders versus non-responders, or to determine the functionality of
the model and these are represented in Figure 7.2. In the overall study population, those
who did not respond either clinically or in vitro were two male participants who were both
ApoE 3/3. It might be expected that such participants will skew correlations negatively. In
cell culture, treatment would have little to no negative effect, while over 12 weeks, one
would expect some degree of disease progression (measured in this instance as a
decrease in left total hippocampal volume). Even if treatment proved negative in vitro,
any treatment-precipitated decline would be compounded by natural disease
Maximal Respiration
Spare Respirational Capacity
r = 0.61 r = 0.64
Figure 7.1. Potential Errors in NSC Model. Spearman Correlation between Clinical
MRI Total Left Hippocampal Volume Change from Baseline to post-treatment and
Maximal Respiration % change with Allo 100nM, indicating participants who have
Type I and Type II errors for the in vitro marker.

123
progression. The disconnect between clinical decline and little in vitro response, would
result in decreased correlation coefficient.

Initial evaluations of clinical data indicate that in ApoE 4 carriers are the most likely to
have increased hippocampal volume after 12 weeks of 4mg Allo treatment. In the in
vitro-clinical data subgroup analyses, the correlation between left hippocampal
volume and maximal respiration as well as spare respirational capacity is increased
when the data is limited to those participants with ApoE 3/4 genotypes only. This
indicates that the in vitro iPSC-derived model more accurately predicts clinical
r = 1
Clinical (+)
Marker (+)
Clinical (-)
Marker (+)
Clinical (+)
Marker (-)
Clinical (-)
Marker (-)
Figure 7.2. Correlational Analysis Interpretation. Interpretation of correlational
analyses, with theoretical, perfect positive correlation trend line. The green box
indicates responders correctly identified by the in vitro model. The red box indicates
those negative responders correctly identified by the in vitro model. The blue boxes
indicate discrepancies between the clinical and the in vitro Allo response.

124
patient response in those with the ApoE 3/4 genotype, although why this is, is as yet
unclear. This isn’t to say that Allo is unsuccessful in other ApoE genotypes. Here, it
is important to consider the diagnosis of each study participant/donor. Previous AD
clinical trials have noted that poor efficacy rates and a failure to meet primary
endpoints is due to the fact that a proportion of participants do not actually have AD.
The Phase III bapineuzumab study quantified this proportion is a post-hoc analysis to
determine that in non-ApoE4 carriers, 36% of the randomized participants did not
have Aβ plaques at baseline
10
. In this Phase I study, it was not clinically necessary or
economically feasible, to quantify Aβ levels; therefore this data is unavailable to us,
although it is built into future Phase II clinical protocols. Within the 3/4 genotype, a
strong positive correlation doesn’t necessarily mean that all those ApoE4 carriers will
respond to treatment, but rather that in this subpopulation iPSC-derivatives are a
stronger disease model. Therefore, the next step is to examine those ApoE 3/4
participants who do not respond positively to Allo treatment. In this instance, there
were two participants who did not show improvements in hippocampal volume or
mitochondrial function. Both were female participants, one in each dosing cohort.
While the correlation between the two outcomes was similar, in order to advance our
understanding of responders versus non-responders, it is these participants that
need to be more fully understood.

In understanding the interaction between ApoE genotype and therapeutic response,
the placebo group proves to be an important tool. In the 4 participants randomized to
the placebo arm, two demonstrate a strong improvement in mitochondrial response
to Allo. Of the other two placebo participants, one demonstrates a strong negative

125
response with Allo on both platforms and the other is essentially neutral. The two
who demonstrated a strong in vitro response and negative clinical response (the
latter to be expected in the placebo arm of the study) were both female participants
with an ApoE 3/4 genotype. The two who had negative or neutral responses were
male with an ApoE 3/3 genotype. Such an observation lends credence to the
proposal that ApoE carriers may be a more appropriate subgroup for Allo treatment.

Based on everything discussed above, we are now asking the question: how does
this integrate with and advance the larger Allopregnanolone drug development
program? Both feasibility and regulatory requirements must be considered. There
are two main possibilities: further development to aid clinical trial enrichment and true
biomarker development. Considering first the FDA’s comprehensive guidance
138
on
clinical trial enrichment, it is important to define enrichment as “the prospective use
of any patient characteristic to select a study population in which detection of a drug
effect (if one is in fact present) is more likely than it would be in an unselected
population”, that is, laying out the original purpose of this research project:
developing a predictive enrichment strategy. As noted in the guidance, such an
approach facilitates a larger effect size with a smaller study population. Applied to
our data, if we assigned a cut point of Mitochondrial Maximal Respiration/Spare
Respirational Capacity % change greater than 0, 60%-70% of the treated
participants would be considered marker positive (Green Box - Figure 7.2). However,
10% of the participants are marker negative but show a clinical response (Blue box,
upper left quadrant Figure 7.2). The greatest sample size reduction occurs when
there is a low prevalence of marker (+) patients and a low effect size in the marker (-)

126
patients. It is critical however, to ensure that in enriching a study population, one
does not inadvertently limit generalizability and future FDA approved labeling.

If we account for the above estimates for marker (+) and (-) participants, as well as
acknowledging that data on marker (-) patients is important to determine if they
respond less well or if there is no effect at all, and based on the length of time to get
data from an iPSC-based marker, there are a couple of possible clinical trial designs
recommended by FDA. The first, and most advisable at this stage, would evaluate
include marker (+) and (-) participants, as this would allow for further characterization
of the predictiveness of mitochondrial function as well as provide more data for
selection of a future cut point and to understand responsiveness in marker (-)
participants. A sample for stem cell development and evaluation of mitochondrial
respiration responsivess would be collected at baseline and patients randomized
without regard to future marker status based on in vitro outcomes. The primary
Figure 7.3. Recommended Clinical Trial Design. NSC marker status is determined,
but does not affect participant randomization. However, it can be powered for a primary
endpoint of efficacy within the marker (+) population only. This allows for further
refinement and a deeper understanding of the in vitro model. From FDA Guidance:
Enrichment Strategies for Clinical Trials to Support Approval of Human Drugs and
Biological Products

127
endpoint could be based on either the whole population, or the study alpha divided
between two endpoints (overall population and marker (+) only) (Figure 7.3.).
Alternatively, the study could be powered based on primary outcome in the (+) only,
with a risk-benefit analysis in all study participants. The second design, which would
not be advisable at this stage, would require an in vitro readout for each participant
prior to randomization. Marker (+) and (-) participants would be enrolled, but stratified
based on their status (Figure 7.4.). For long-term development plan, it is vital to
consider how patient management decisions will be made on the basis of the trial
results. Because, if it is determined that marker classification is necessary prior to
treatment initiation, the labeling of an approved drug would be limited and the marker
would have to be developed as a companion diagnostic, which has significant
regulatory thresholds. The decision to utilize an enrichment strategy lays with the
sponsor, not FDA, however the agency should be engaged in order to ensure they
have confidence in the adequacy of the study and negate future concerns about drug
labeling
138
. On the other hand, approval of a companion diagnostic requires
significant dialogue with the FDA and because it becomes part of the drug labeling,
is included in the IND and NDA submissions
8,139
.

128

Other regulatory approaches to biomarker development include pursuing biomarker
qualification or letter of support through FDA or EMA
8,139
. These are two distinct
pathways depending on the overall goals of the research program. The former is
meant to establish a biomarker for use in multiple development programs, requiring
an analytically validated assay and significant regulatory hurdles. Due to this, such
qualifications are usually done through consortium and not an individual drug
sponsor. As the marker in this study is indicative of a change in hippocampal volume,
this is an outcome specific to potential neuroregneretive medicines, which is a rare
mechanism in the current AD development portfolio. A letter of support on the other
hand allows the FDA to highlight potential biomarkers as being in need of additional
evidence to support specific contexts of use.

Figure 7.4. Alternative Clinical Trial Design. Clinical Trial Design where marker
status is determined, and used to stratify participants prior to randomization. Both (+)
and (-) participants are randomized, which allows for further understanding of the in
vitro model, but may result in a limited drug label and require companion diagnostic
development. From FDA Guidance: Enrichment Strategies for Clinical Trials to
Support Approval of Human Drugs and Biological Products

129
The model and marker of mitochondrial function as described here isn’t quite ready
for the big leagues yet, and would necessitate further studies randomizing both (+)
and (-) participants. However, the data we describe is critical for future power
calculations for both the size of in vitro effect, number of marker (+) participants, and
error estimates. The model currently can predict to an extent a neuroimaging
change. To effectively translate this discovery into a criterion for enrichment, it is
necessary to advance this still further and demonstrate a positive correlation
between an increase in left hippocampal volume and improved cognitive function
with Allo treatment. This could not be accomplished in the current Phase I trial due to
the limited sample size and short treatment duration, but will be a key component of
the Phase II study design. In summary, this work has nicely established a possible in
vitro marker, worthy of further refinement and development in upcoming clinical
studies.

Others have indicated that such research is most efficient and has the greatest
translational impact when conducted as part of larger pre-existing clinical research
team, a recommendation we echo
109
. Here, existing collaborations with neurology
and preventative medicine faculty were strengthened and relationships initiated with
the expanding USC neuroimaging department. Such collaborations engage clinical
and pre-clinical researchers, galvanizing and enabling translation both to and from
the bedside. At the study completing, dissemination of resulting data from study
members to their respective fields ensures that the message reaches a broader
audience, and in doing so, highlights the scientific advantages of strong,
collaborative teams.

130

iPSCs derived from patients and clinical trial participants provide a valuable
opportunity to shift the drug development paradigm. This study has highlighted
mitochondrial respiration as possible novel method of patient stratification based on
Allo response. While an improvement in mitochondrial function was observed in
previous animal studies, using iPSC technology translated this from the bench into a
relevant, clinical trial setting. Future work will further this approach and advance our
understanding of Alzheimer’s disease and an individual patient’s response to Allo.

131
References

1. 2017 Alzheimer's disease facts and figures. Alzheimer's & Dementia: The
Journal of the Alzheimer's Association. 2017;13(4):325-373.
2. 2015 Alzheimer's disease facts and figures. Alzheimer's & dementia : the journal
of the Alzheimer's Association. 2015;11(3):332-384.
3. Wainaina MN, Chen Z, Zhong C. Environmental factors in the development and
progression of late-onset Alzheimer's disease. Neurosci Bull. 2014;30(2):253-
270.
4. FDA. Approval Package for: Application Number 21-487. 2003;
http://www.accessdata.fda.gov/drugsatfda_docs/nda/2003/21-
487_Namenda_Approv.pdf. Accessed 11/12/2015.
5. Alzheimer's_Association. Medications for Memory Loss.
http://www.alz.org/alzheimers_disease_standard_prescriptions.asp. Accessed
11/12/2015.
6. Cummings JL, Morstorf T, Zhong K. Alzheimer's disease drug-development
pipeline: few candidates, frequent failures. Alzheimer's research & therapy.
2014;6(4):37.
7. Axovant Announces Negative Topline Results of Intepirdine Phase 3 MINDSET
Trial in Alzheimer's Disease [press release]. Basel, Switzerland: Axovant
Sciences2017.
8. Hampel H, Frank R, Broich K, et al. Biomarkers for Alzheimer's disease:
academic, industry and regulatory perspectives. Nat Rev Drug Discov.
2010;9(7):560-574.
9. Berkrot BP, R. Ahead of Alzheimer’s meeting, researchers seize on signs of
progress. Reuters2015.
10. Salloway S, Sperling R, Brashear HR. Phase 3 trials of solanezumab and
bapineuzumab for Alzheimer's disease. The New England journal of medicine.
2014;370(15):1460.
11. Brinton RD. Neurosteroids as regenerative agents in the brain: therapeutic
implications. Nature reviews Endocrinology. 2013;9(4):241-250.
12. Irwin RW, Brinton RD. Allopregnanolone as regenerative therapeutic for
Alzheimer's disease: translational development and clinical promise. Progress in
neurobiology. 2014;113:40-55.
13. Brinton RD. Neurosteroids as regenerative agents in the brain: therapeutic
implications. Nature reviews Endocrinology. 2013;9(4):241-250.
14. Reddy DS. Neurosteroids: endogenous role in the human brain and therapeutic
potentials. Prog Brain Res. 2010;186:113-137.
15. Irwin RW, Solinsky CM, Loya CM, et al. Allopregnanolone preclinical acute
pharmacokinetic and pharmacodynamic studies to predict tolerability and efficacy
for Alzheimer's disease. PloS one. 2015;10(6):e0128313.
16. Wang JM, Johnston PB, Ball BG, Brinton RD. The neurosteroid allopregnanolone
promotes proliferation of rodent and human neural progenitor cells and regulates
cell-cycle gene and protein expression. J Neurosci. 2005;25(19):4706-4718.

132
17. Chen S, Wang JM, Irwin RW, Yao J, Liu L, Brinton RD. Allopregnanolone
Promotes Regeneration and Reduces beta-Amyloid Burden in a Preclinical
Model of Alzheimer's Disease. PloS one. 2011;6(8):e24293.
18. Singh C, Liu L, Wang JM, et al. Allopregnanolone restores hippocampal-
dependent learning and memory and neural progenitor survival in aging 3xTgAD
and nonTg mice. Neurobiology of aging. 2012;33(8):1493-1506.
19. Brinton RD GA, Schmidr PJ, Morrison JH. Reproductive Aging of Female: Neural
Systems. In: Doanld W. Pfaff APA, Anne M. Etgen, Fahrbach SE, Rubin RT, ed.
Hormones, Brain and Behavior. 2nd ed. San Diego: Academic Press; 2009:2199-
2216.
20. Brinton RD. The healthy cell bias of estrogen action: mitochondrial bioenergetics
and neurological implications. Trends in neurosciences. 2008;31(10):529-537.
21. Yao J, Hamilton RT, Cadenas E, Brinton RD. Decline in mitochondrial
bioenergetics and shift to ketogenic profile in brain during reproductive
senescence. Biochimica et biophysica acta. 2010;1800(10):1121-1126.
22. Brinton RD. Estrogen-induced plasticity from cells to circuits: predictions for
cognitive function. Trends in pharmacological sciences. 2009;30(4):212-222.
23. Blalock EM, Grondin R, Chen KC, et al. Aging-related gene expression in
hippocampus proper compared with dentate gyrus is selectively associated with
metabolic syndrome variables in rhesus monkeys. The Journal of neuroscience :
the official journal of the Society for Neuroscience. 2010;30(17):6058-6071.
24. Kadish I, Thibault O, Blalock EM, et al. Hippocampal and cognitive aging across
the lifespan: a bioenergetic shift precedes and increased cholesterol trafficking
parallels memory impairment. The Journal of neuroscience : the official journal of
the Society for Neuroscience. 2009;29(6):1805-1816.
25. Landfield PW, Blalock EM, Chen KC, Porter NM. A new glucocorticoid
hypothesis of brain aging: implications for Alzheimer's disease. Current
Alzheimer research. 2007;4(2):205-212.
26. Blalock EM, Geddes JW, Chen KC, Porter NM, Markesbery WR, Landfield PW.
Incipient Alzheimer's disease: microarray correlation analyses reveal major
transcriptional and tumor suppressor responses. Proceedings of the National
Academy of Sciences of the United States of America. 2004;101(7):2173-2178.
27. Moreira PI, Zhu X, Wang X, et al. Mitochondria: a therapeutic target in
neurodegeneration. Biochimica et biophysica acta. 2010;1802(1):212-220.
28. Yao J, Irwin RW, Zhao L, Nilsen J, Hamilton RT, Brinton RD. Mitochondrial
bioenergetic deficit precedes Alzheimer's pathology in female mouse model of
Alzheimer's disease. Proceedings of the National Academy of Sciences of the
United States of America. 2009;106(34):14670-14675.
29. Mosconi L, Berti V, Swerdlow RH, Pupi A, Duara R, de Leon M. Maternal
transmission of Alzheimer's disease: prodromal metabolic phenotype and the
search for genes. Human genomics. 2010;4(3):170-193.
30. Spuch C, Ortolano S, Navarro C. New insights in the amyloid-Beta interaction
with mitochondria. J Aging Res. 2012;2012:324968.
31. Yao J, Rettberg JR, Klosinski LP, Cadenas E, Brinton RD. Shift in brain
metabolism in late onset Alzheimer's disease: implications for biomarkers and
therapeutic interventions. Molecular aspects of medicine. 2011;32(4-6):247-257.

133
32. Area-Gomez E, de Groof AJ, Boldogh I, et al. Presenilins are enriched in
endoplasmic reticulum membranes associated with mitochondria. The American
journal of pathology. 2009;175(5):1810-1816.
33. Hansson CA, Frykman S, Farmery MR, et al. Nicastrin, presenilin, APH-1, and
PEN-2 form active gamma-secretase complexes in mitochondria. The Journal of
biological chemistry. 2004;279(49):51654-51660.
34. Pavlov PF, Wiehager B, Sakai J, et al. Mitochondrial gamma-secretase
participates in the metabolism of mitochondria-associated amyloid precursor
protein. FASEB journal : official publication of the Federation of American
Societies for Experimental Biology. 2011;25(1):78-88.
35. Young KJ, Bennett JP. The mitochondrial secret(ase) of Alzheimer's disease.
Journal of Alzheimer's disease : JAD. 2010;20 Suppl 2:S381-400.
36. Swerdlow RH, Burns JM, Khan SM. The Alzheimer's disease mitochondrial
cascade hypothesis: progress and perspectives. Biochim Biophys Acta.
2014;1842(8):1219-1231.
37. Reddy PH, Beal MF. Amyloid beta, mitochondrial dysfunction and synaptic
damage: implications for cognitive decline in aging and Alzheimer's disease.
Trends in molecular medicine. 2008;14(2):45-53.
38. Irwin RW, Solinsky CM, Brinton RD. Frontiers in therapeutic development of
allopregnanolone for Alzheimer's disease and other neurological disorders. Front
Cell Neurosci. 2014;8:203.
39. Grant KA, Helms CM, Rogers LS, Purdy RH. Neuroactive steroid
stereospecificity of ethanol-like discriminative stimulus effects in monkeys. The
Journal of pharmacology and experimental therapeutics. 2008;326(1):354-361.
40. Kask K, Backstrom T, Lundgren P, Sundstrom Poromaa I. Allopregnanolone has
no effect on startle response and prepulse inhibition of startle response in
patients with premenstrual dysphoric disorder or healthy controls. Pharmacol
Biochem Behav. 2009;92(4):608-613.
41. Kask K, Backstrom T, Nilsson LG, Sundstrom-Poromaa I. Allopregnanolone
impairs episodic memory in healthy women. Psychopharmacology (Berl).
2008;199(2):161-168.
42. Timby E, Balgard M, Nyberg S, et al. Pharmacokinetic and behavioral effects of
allopregnanolone in healthy women. Psychopharmacology (Berl).
2006;186(3):414-424.
43. van Broekhoven F, Backstrom T, van Luijtelaar G, Buitelaar JK, Smits P, Verkes
RJ. Effects of allopregnanolone on sedation in men, and in women on oral
contraceptives. Psychoneuroendocrinology. 2007;32(5):555-564.
44. clinicaltrials.gov. Identifier NCT01673828, Allopregnanolone for the
Treatment of Traumatic Brain Injury. June 28, 2017;
https://clinicaltrials.gov/ct2/show/study/NCT01673828?term=Allopregnanolone
&cond=Traumatic+Brain+Injury&rank=1. Accessed Sep 28 2017, 2017.
45. Broomall E, Natale JE, Grimason M, et al. Pediatric super-refractory status
epilepticus treated with allopregnanolone. Ann Neurol. 2014;76(6):911-915.
46. Vaitkevicius H, Husain AM, Rosenthal ES, et al. First-in-man allopregnanolone
use in super-refractory status epilepticus. Ann Clin Transl Neurol. 2017;4(6):411-
414.
47. ClinicalTrials.gov. Identifier NCT02052739, Study to Evaluate SAGE-
547 Injection as Adjunctive Therapy for the Treatment of Super-Refractory

134
Status Epilepticus (SRSE). April 26, 2017;
https://clinicaltrials.gov/ct2/show/NCT02052739?term=Sage-547&rank=5.
Accessed Sep 28, 2017, 2017.
48. ClinicalTrials.gov. Identifier NCT02285504, Evaluate SAGE-547 in Female
Patients With Severe Postpartum Depression. April 26, 2017;
https://clinicaltrials.gov/ct2/show/NCT02285504?term=Sage-547&rank=1.
Accessed Sep 28 2017, 2017.
49. ClinicalTrials.gov. Identifier NCT02614547, A Study to Evaluate SAGE-
547 in Patients With Severe Postpartum Depression. Aug 8, 2017;
https://clinicaltrials.gov/ct2/show/NCT02614547?term=Sage-547&rank=3.
Accessed Sep 28, 2017, 2017.
50. ClinicalTrials.gov. Identifier NCT02277106, Evaluate SAGE-547 in Patients
With Essential Tremor. April 26, 2017;
https://clinicaltrials.gov/ct2/show/NCT02277106?term=Sage-547&rank=2.
Accessed Sep 28 2017, 2017.
51. Wang JY, Trivedi AM, Carrillo NR, et al. Open-Label Allopregnanolone Treatment
of Men with Fragile X-Associated Tremor/Ataxia Syndrome. Neurotherapeutics.
2017.
52. Ligsay A, Van Dijck A, Nguyen DV, et al. A randomized double-blind, placebo-
controlled trial of ganaxolone in children and adolescents with fragile X
syndrome. J Neurodev Disord. 2017;9(1):26.
53. Brinton RD, Schneider LS, Law M. Allopregnanolone, Regenerative Therapeutic
for Alzheimer’s Disease: Phase 1b/2a Update. Alzheimer's Association
International Conference; 2017; London, UK.
54. Takahashi K, Yamanaka S. Induction of pluripotent stem cells from mouse
embryonic and adult fibroblast cultures by defined factors. Cell. 2006;126(4):663-
676.
55. Israel MA, Yuan SH, Bardy C, et al. Probing sporadic and familial Alzheimer's
disease using induced pluripotent stem cells. In: Nature. Vol 482.
England2012:216-220.
56. Chung SY, Kishinevsky S, Mazzulli JR, et al. Parkin and PINK1 Patient iPSC-
Derived Midbrain Dopamine Neurons Exhibit Mitochondrial Dysfunction and
alpha-Synuclein Accumulation. Stem Cell Reports. 2016;7(4):664-677.
57. Terrenoire C, Wang K, Tung KW, et al. Induced pluripotent stem cells used to
reveal drug actions in a long QT syndrome family with complex genetics. J Gen
Physiol. 2013;141(1):61-72.
58. Broccoli V, Giannelli SG, Mazzara PG. Modeling physiological and pathological
human neurogenesis in the dish. Front Neurosci. 2014;8:183.
59. Liu Q, Waltz S, Woodruff G, et al. Effect of potent gamma-secretase modulator in
human neurons derived from multiple presenilin 1-induced pluripotent stem cell
mutant carriers. JAMA neurology. 2014;71(12):1481-1489.
60. Hossini AM, Megges M, Prigione A, et al. Induced pluripotent stem cell-derived
neuronal cells from a sporadic Alzheimer's disease donor as a model for
investigating AD-associated gene regulatory networks. BMC genomics.
2015;16:84.

135
61. Kondo T, Asai M, Tsukita K, et al. Modeling Alzheimer's disease with iPSCs
reveals stress phenotypes associated with intracellular Abeta and differential
drug responsiveness. Cell stem cell. 2013;12(4):487-496.
62. Shi Y, Kirwan P, Smith J, MacLean G, Orkin SH, Livesey FJ. A human stem cell
model of early Alzheimer's disease pathology in Down syndrome. Science
translational medicine. 2012;4(124):124ra129.
63. Duan L, Bhattacharyya BJ, Belmadani A, Pan L, Miller RJ, Kessler JA. Stem cell
derived basal forebrain cholinergic neurons from Alzheimer's disease patients
are more susceptible to cell death. In: Mol Neurodegener. Vol 9. England2014:3.
64. Muratore CR, Rice HC, Srikanth P, et al. The familial Alzheimer's disease
APPV717I mutation alters APP processing and Tau expression in iPSC-derived
neurons. In: Hum Mol Genet. Vol 23. England: The Author 2014. Published by
Oxford University Press For Permissions, please email:
journals.permissions@oup.com.; 2014:3523-3536.
65. Koch P, Tamboli IY, Mertens J, et al. Presenilin-1 L166P mutant human
pluripotent stem cell-derived neurons exhibit partial loss of gamma-secretase
activity in endogenous amyloid-beta generation. The American journal of
pathology. 2012;180(6):2404-2416.
66. Mertens J, Stuber K, Wunderlich P, et al. APP processing in human pluripotent
stem cell-derived neurons is resistant to NSAID-based gamma-secretase
modulation. Stem cell reports. 2013;1(6):491-498.
67. Oka S, Leon J, Sakumi K, et al. Human mitochondrial transcriptional factor A
breaks the mitochondria-mediated vicious cycle in Alzheimer's disease. Sci Rep.
2016;6:37889.
68. Lu R, Wang J, Tao R, et al. Reduced TRPC6 mRNA levels in the blood cells of
patients with Alzheimer's disease and mild cognitive impairment. Mol Psychiatry.
2017.
69. Raja WK, Mungenast AE, Lin YT, et al. Self-Organizing 3D Human Neural Tissue
Derived from Induced Pluripotent Stem Cells Recapitulate Alzheimer's Disease
Phenotypes. PLoS One. 2016;11(9):e0161969.
70. Mutations. http://www.alzforum.org/mutations. Accessed Sep 28 2017, 2017.
71. Cooper O, Seo H, Andrabi S, et al. Pharmacological rescue of mitochondrial
deficits in iPSC-derived neural cells from patients with familial Parkinson's
disease. In: Sci Transl Med. Vol 4. United States2012:141ra190.
72. Armijo E, Gonzalez C, Shahnawaz M, Flores A, Davis B, Soto C. Increased
susceptibility to Abeta toxicity in neuronal cultures derived from familial
Alzheimer's disease (PSEN1-A246E) induced pluripotent stem cells. Neurosci
Lett. 2017;639:74-81.
73. Brownjohn PW, Smith J, Portelius E, et al. Phenotypic Screening Identifies
Modulators of Amyloid Precursor Protein Processing in Human Stem Cell Models
of Alzheimer's Disease. Stem Cell Reports. 2017;8(4):870-882.
74. Hossini AM, Megges M, Prigione A, et al. Induced pluripotent stem cell-derived
neuronal cells from a sporadic Alzheimer's disease donor as a model for
investigating AD-associated gene regulatory networks. In: BMC Genomics. Vol
16. England2015:84.
75. Jones VC, Atkinson-Dell R, Verkhratsky A, Mohamet L. Aberrant iPSC-derived
human astrocytes in Alzheimer's disease. Cell Death Dis. 2017;8(3):e2696.

136
76. Mahairaki V, Ryu J, Peters A, et al. Induced pluripotent stem cells from familial
Alzheimer's disease patients differentiate into mature neurons with
amyloidogenic properties. Stem Cells Dev. 2014;23(24):2996-3010.
77. Pires C, Schmid B, Petraeus C, et al. Generation of a gene-corrected isogenic
control cell line from an Alzheimer's disease patient iPSC line carrying a A79V
mutation in PSEN1. Stem Cell Res. 2016;17(2):285-288.
78. Poon A, Li T, Pires C, et al. Derivation of induced pluripotent stem cells from a
familial Alzheimer's disease patient carrying the L282F mutation in presenilin 1.
Stem Cell Res. 2016;17(3):470-473.
79. Poon A, Schmid B, Pires C, et al. Generation of a gene-corrected isogenic
control hiPSC line derived from a familial Alzheimer's disease patient carrying a
L150P mutation in presenilin 1. Stem Cell Res. 2016;17(3):466-469.
80. Schroter F, Sleegers K, Cuyvers E, et al. Lymphoblast-derived integration-free
iPS cell line from a 65-year-old Alzheimer's disease patient expressing the
TREM2 p.R47H variant. Stem Cell Res. 2016;16(1):113-115.
81. Schroter F, Sleegers K, Cuyvers E, et al. Lymphoblast-derived integration-free
iPS cell line from a female 67-year-old Alzheimer's disease patient with TREM2
(R47H) missense mutation. Stem Cell Res. 2016;17(3):553-555.
82. Schroter F, Sleegers K, Van Cauwenberghe C, et al. Lymphoblast-derived
integration-free iPSC lines from a female and male Alzheimer's disease patient
expressing different copy numbers of a coding CNV in the Alzheimer risk gene
CR1. Stem Cell Res. 2016;17(3):560-563.
83. Sproul AA, Jacob S, Pre D, et al. Characterization and molecular profiling of
PSEN1 familial Alzheimer's disease iPSC-derived neural progenitors. In: PLoS
One. Vol 9. United States2014:e84547.
84. Woodruff G, Young JE, Martinez FJ, et al. The presenilin-1 DeltaE9 mutation
results in reduced gamma-secretase activity, but not total loss of PS1 function, in
isogenic human stem cells. Cell Rep. 2013;5(4):974-985.
85. Woodruff G, Reyna SM, Dunlap M, et al. Defective Transcytosis of APP and
Lipoproteins in Human iPSC-Derived Neurons with Familial Alzheimer's Disease
Mutations. Cell Rep. 2016;17(3):759-773.
86. Yagi T, Ito D, Okada Y, et al. Modeling familial Alzheimer's disease with induced
pluripotent stem cells. In: Hum Mol Genet. Vol 20. England2011:4530-4539.
87. Yang J, Zhao H, Ma Y, et al. Early pathogenic event of Alzheimer's disease
documented in iPSCs from patients with PSEN1 mutations. Oncotarget.
2017;8(5):7900-7913.
88. Young JE, Boulanger-Weill J, Williams DA, et al. Elucidating Molecular
Phenotypes Caused by the SORL1 Alzheimer's Disease Genetic Risk Factor
Using Human Induced Pluripotent Stem Cells. Cell Stem Cell. 2015;16(4):373-
385.
89. Zhang S, Lv Z, Liu L, et al. Characterization of human induced pluripotent stem
cell (iPSC) line from a 72year old male patient with later onset Alzheimer's
disease. Stem Cell Res. 2017;19:34-36.
90. Zhao J, Davis MD, Martens YA, et al. APOE epsilon4/epsilon4 diminishes
neurotrophic function of human iPSC-derived astrocytes. Hum Mol Genet.
2017;26(14):2690-2700.
91. Neofytou E, O'Brien CG, Couture LA, Wu JC. Hurdles to clinical translation of
human induced pluripotent stem cells. J Clin Invest. 2015;125(7):2551-2557.

137
92. Mandai M, Watanabe A, Kurimoto Y, et al. Autologous Induced Stem-Cell-
Derived Retinal Cells for Macular Degeneration. N Engl J Med.
2017;376(11):1038-1046.
93. Kimbrel EA, Lanza R. Current status of pluripotent stem cells: moving the first
therapies to the clinic. Nat Rev Drug Discov. 2015;14(10):681-692.
94. Knoepfler P. Landmark IPSC clinical study on hold due to genomic issue. In. The
Niche. Vol 20172015.
95. Knoepfler P. Good stem cell news as Takahashi IPS Cell Trial to Resume. In.
The Niche. Vol 20172016.
96. Knoepfler P. Japan IPS cell trial for vision to start in new incarnation. In. The
Niche. Vol 20172017.
97. Takamatsu K, Ikeda T, Haruta M, et al. Degradation of amyloid beta by human
induced pluripotent stem cell-derived macrophages expressing Neprilysin-2.
Stem Cell Res. 2014;13(3 Pt A):442-453.
98. Nizzardo M, Simone C, Rizzo F, et al. Minimally invasive transplantation of iPSC-
derived ALDHhiSSCloVLA4+ neural stem cells effectively improves the
phenotype of an amyotrophic lateral sclerosis model. Hum Mol Genet.
2014;23(2):342-354.
99. Inoue H, Nagata N, Kurokawa H, Yamanaka S. iPS cells: a game changer for
future medicine. The EMBO journal. 2014;33(5):409-417.
100. Singec I, Simeonov A. Translating Stem Cell Biology Into Drug Discovery. Drug
Target Rev. 2016;3(2):34-38.
101. Papapetrou EP. Patient-derived induced pluripotent stem cells in cancer
research and precision oncology. Nat Med. 2016;22(12):1392-1401.
102. Kremer LC, van Dalen EC, Offringa M, Voute PA. Frequency and risk factors of
anthracycline-induced clinical heart failure in children: a systematic review. Ann
Oncol. 2002;13(4):503-512.
103. Kremer LC, van der Pal HJ, Offringa M, van Dalen EC, Voute PA. Frequency and
risk factors of subclinical cardiotoxicity after anthracycline therapy in children: a
systematic review. Ann Oncol. 2002;13(6):819-829.
104. Swain SM, Whaley FS, Ewer MS. Congestive heart failure in patients treated with
doxorubicin: a retrospective analysis of three trials. Cancer. 2003;97(11):2869-
2879.
105. Burridge PW, Li YF, Matsa E, et al. Human induced pluripotent stem cell-derived
cardiomyocytes recapitulate the predilection of breast cancer patients to
doxorubicin-induced cardiotoxicity. Nat Med. 2016;22(5):547-556.
106. Garbes L, Heesen L, Holker I, et al. VPA response in SMA is suppressed by the
fatty acid translocase CD36. Hum Mol Genet. 2013;22(2):398-407.
107. Mertens J, Wang QW, Kim Y, et al. Differential responses to lithium in
hyperexcitable neurons from patients with bipolar disorder. Nature.
2015;527(7576):95-99.
108. Cao L, McDonnell A, Nitzsche A, et al. Pharmacological reversal of a pain
phenotype in iPSC-derived sensory neurons and patients with inherited
erythromelalgia. Sci Transl Med. 2016;8(335):335ra356.
109. McNeish J, Gardner JP, Wainger BJ, Woolf CJ, Eggan K. From Dish to Bedside:
Lessons Learned While Translating Findings from a Stem Cell Model of Disease
to a Clinical Trial. Cell Stem Cell. 2015;17(1):8-10.

138
110. .gov C. Identifier NCT02221622, Allopregnanolone for Mild Cognitive
Impairment Due to Alzheimer's Disease or Mild AD (Allo). 2017;
https://clinicaltrials.gov/ct2/show/NCT02221622. Accessed Sep 28, 2017.
111. DiFrancesco JC, Longoni M, Piazza F. Anti-Abeta Autoantibodies in Amyloid
Related Imaging Abnormalities (ARIA): Candidate Biomarker for Immunotherapy
in Alzheimer's Disease and Cerebral Amyloid Angiopathy. Front Neurol.
2015;6:207.
112. Sperling RA, Jack CR, Jr., Black SE, et al. Amyloid-related imaging abnormalities
in amyloid-modifying therapeutic trials: recommendations from the Alzheimer's
Association Research Roundtable Workgroup. Alzheimers Dement.
2011;7(4):367-385.
113. Reiman EM, Quiroz YT, Fleisher AS, et al. Brain imaging and fluid biomarker
analysis in young adults at genetic risk for autosomal dominant Alzheimer's
disease in the presenilin 1 E280A kindred: a case-control study. Lancet Neurol.
2012;11(12):1048-1056.
114. Reddy DS, Rogawski MA. Neurosteroid replacement therapy for catamenial
epilepsy. Neurotherapeutics. 2009;6(2):392-401.
115. Simon R, Freidlin B, Rubinstein L, Arbuck SG, Collins J, Christian MC.
Accelerated titration designs for phase I clinical trials in oncology. J Natl Cancer
Inst. 1997;89(15):1138-1147.
116. Albert MS, DeKosky ST, Dickson D, et al. The diagnosis of mild cognitive
impairment due to Alzheimer's disease: recommendations from the National
Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines
for Alzheimer's disease. Alzheimers Dement. 2011;7(3):270-279.
117. Petersen RC. Mild cognitive impairment clinical trials. Nat Rev Drug Discov.
2003;2(8):646-653.
118. McKhann GM, Knopman DS, Chertkow H, et al. The diagnosis of dementia due
to Alzheimer's disease: recommendations from the National Institute on Aging-
Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's
disease. Alzheimers Dement. 2011;7(3):263-269.
119. Irwin RW, Wang JM, Chen S, Brinton RD. Neuroregenerative mechanisms of
allopregnanolone in Alzheimer's disease. Frontiers in endocrinology. 2011;2:117.
120. Mayeux R, Saunders AM, Shea S, et al. Utility of the apolipoprotein E genotype
in the diagnosis of Alzheimer's disease. Alzheimer's Disease Centers Consortium
on Apolipoprotein E and Alzheimer's Disease. N Engl J Med. 1998;338(8):506-
511.
121. Kennedy RE, Cutter GR, Schneider LS. Effect of APOE genotype status on
targeted clinical trials outcomes and efficiency in dementia and mild cognitive
impairment resulting from Alzheimer's disease. Alzheimers Dement.
2014;10(3):349-359.
122. Schneider LS, Kennedy RE, Wang G, Cutter GR. Differences in Alzheimer
disease clinical trial outcomes based on age of the participants. In: Neurology.
Vol 84.2015:1121-1127.
123. Mallone R, Mannering SI, Brooks-Worrell BM, et al. Isolation and preservation of
peripheral blood mononuclear cells for analysis of islet antigen-reactive T cell
responses: position statement of the T-Cell Workshop Committee of the
Immunology of Diabetes Society. Clin Exp Immunol. 2011;163(1):33-49.

139
124. Okita K, Yamakawa T, Matsumura Y, et al. An efficient nonviral method to
generate integration-free human-induced pluripotent stem cells from cord blood
and peripheral blood cells. Stem Cells. 2013;31(3):458-466.
125. Chambers SM, Fasano CA, Papapetrou EP, Tomishima M, Sadelain M, Studer
L. Highly efficient neural conversion of human ES and iPS cells by dual inhibition
of SMAD signaling. In: Nat Biotechnol. Vol 27. United States2009:275-280.
126. Tomishima M. Neural induction - Dual SMAD inhibition. StemBook edJune 10,
2012.
127. Okita K, Matsumura Y, Sato Y, et al. A more efficient method to generate
integration-free human iPS cells. Nat Methods. 2011;8(5):409-412.
128. Bengtsson SK, Johansson M, Backstrom T. Long-term continuous
allopregnanolone elevation causes memory decline and hippocampus shrinkage,
in female wild-type B6 mice. Horm Behav. 2016;78:160-167.
129. Bengtsson SK, Johansson M, Backstrom T, Wang M. Chronic allopregnanolone
treatment accelerates Alzheimer's disease development in
AbetaPP(Swe)PSEN1(DeltaE9) mice. J Alzheimers Dis. 2012;31(1):71-84.
130. Engle SJ, Puppala D. Integrating human pluripotent stem cells into drug
development. Cell Stem Cell. 2013;12(6):669-677.
131. Babos K, Ichida JK. Small Molecules Take a Big Step by Converting Fibroblasts
into Neurons. Cell Stem Cell. 2015;17(2):127-129.
132. Hu W, Qiu B, Guan W, et al. Direct Conversion of Normal and Alzheimer's
Disease Human Fibroblasts into Neuronal Cells by Small Molecules. Cell Stem
Cell. 2015;17(2):204-212.
133. Zhang M, Lin YH, Sun YJ, et al. Pharmacological Reprogramming of Fibroblasts
into Neural Stem Cells by Signaling-Directed Transcriptional Activation. Cell
Stem Cell. 2016;18(5):653-667.
134. Cairns DM, Chwalek K, Moore YE, et al. Expandable and Rapidly Differentiating
Human Induced Neural Stem Cell Lines for Multiple Tissue Engineering
Applications. Stem Cell Reports. 2016;7(3):557-570.
135. Liao G, Cheung S, Galeano J, Ji AX, Qin Q, Bi X. Allopregnanolone treatment
delays cholesterol accumulation and reduces autophagic/lysosomal dysfunction
and inflammation in Npc1-/- mouse brain. Brain Res. 2009;1270:140-151.
136. Snapinn SM, Jiang Q. Responder analyses and the assessment of a clinically
relevant treatment effect. Trials. 2007;8:31.
137. Ard MC, Edland SD. Power calculations for clinical trials in Alzheimer's disease. J
Alzheimers Dis. 2011;26 Suppl 3:369-377.
138. Administration UFaD. Enrichment Strategies for Clinical Trials to Support
Approval of Human Drugs and Biological Products. In: Research CfDEa, ed.
Silver Spring, MDDec 14, 2012.
139. Amur S, LaVange L, Zineh I, Buckman-Garner S, Woodcock J. Biomarker
Qualification: Toward a Multiple Stakeholder Framework for Biomarker
Development, Regulatory Acceptance, and Utilization. Clin Pharmacol Ther.
2015;98(1):34-46.

Abstract (if available)

Abstract Introduction. Alzheimer’s disease (AD) is a national and global epidemic with complex pathoetiology including compromised brain metabolic activity and decreased regenerative capacity. Allopregnanolone (Allo) is an investigational neuroregenerative therapeutic, currently in Phase 1b clinical trial for AD (NCT02221622, https://clinicaltrials.gov/ct2/show/NCT02221622?term=NCT02221622&rank=1). In rodent preclinical models, Allo promotes neural stem cell (NSC) proliferation and neural differentiation, improves mitochondrial function, and reduces amyloid-beta pathology. To develop biomarkers predictive of a clinical regenerative response to Allo, this thesis investigates the impact of Allo on human induced pluripotent stem cells (iPSCs) and iPSC-derived NSCs as well as the feasibility of implementing such an approach in a large clinical trial setting. To evaluate the translatability of in vitro data from clinical trial participant cell lines, this data was then correlated with clinical outcomes in the same study participants. ❧ Methods. Peripheral blood mononuclear cells were isolated from whole blood samples of Allo clinical trial participants and t-lymphocytes reprogrammed to iPSCs via a non-integrating, non-viral, episomal plasmid method. Using dual inhibition of SMAD signaling, iPSCs were differentiated to NSCs. Assays were conducted to assess NSC proliferation and mitochondrial function by flow cytometry and live cell extracellular flux respectively. Resulting data were analyzed to determine the regenerative and bioenergetics effect of Allo on clinical trial participant iPSC-NSCs. In vitro data from the first two dosing cohorts was correlated with volumetric MRI of the hippocampus and cognitive testing clinical outcomes. ❧ Results. In vitro data indicated that Allo treatment was associated with an increase in proliferation in 3 out of the 19 participant NSC lines, while contrary to previous in vivo and in vitro studies, in 6 participants, treatment was associated with a decrease in NSC proliferation. However, the mitochondrial function parameters maximal respiration and spare respirational capacity were significantly increased in 8 NSC lines with Allo treatment (mean increase of 30% and 40% versus vehicle respectively), and maximal respiration was nearly significant in an additional 4 participants (mean increase of 20% versus vehicle). These participant cell lines have been labeled ‘responders’, while those that did not have increased NSC proliferation or metabolic capacity are ‘non-responders’. These data demonstrate that Allo promotes regeneration and mitochondrial function of iPSC-derived NSCs in a subset of clinical trial participants. Analyses with clinical data indicated that mitochondrial parameters have positive correlation with participants’ change in left hippocampal volume after 12 weeks of Allo treatment. Correlations increase when limited to the subgroup of participants with the ApoE3/4 genotype, indicating that an iPSC-approach may be a better model of their therapeutic response to Allo compared to modeling the ApoE 3/3 genotype. ❧ Conclusions. Allo treatment of iPSC-derived NSCs resulted in an increase in mitochondrial function that positively correlates with a clinical increase in left hippocampal volume after 12-weeks of Allo infusions. This study is not only one of the largest investigations linking an in vitro iPSC-derivative response with clinical changes but also serve to translate previous discoveries from mouse models into a human AD population. Future work will advance both our basic understanding of Allo in AD by investigating the mechanisms for improved mitochondrial functioning, and also expanding and refining our understanding of in vitro mitochondrial function as a component of a larger drug discovery program. By proposing mitochondrial respiration as possible novel method of patient stratification, these data form the foundation for developing the first biomarker of regenerative potential in brain to determine and monitor response to Allo as a regenerative therapeutic.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

From risk mitochondrial and metabolic phenotype towards a precision medicine approach for Alzheimer's disease

PDF

Perimenopausal transition increases blood brain permeability: implications for neurodegenerative diseases

PDF

IPS and CNS cell models of peroxisomal disorders

PDF

Averting dementia: renin-angiotensin system and angiogenic cells in cognitive decline

PDF

Development of targeted therapies for peroxisome biogenesis disorders

PDF

ROOT: a novel pharmacotranscritomic pipeline for rescuing age-associated functional decline of hippocampal adult neurogenesis

PDF

The peptide angiotensin-(1-7) as a novel treatment for complications induced by type 2 diabetes mellitus

PDF

Understanding human nephrogenesis and scaling synthesis of organoids facilitate modeling of kidney development and disease

Asset Metadata

Creator Solinsky, Christine Marie (author)

Core Title An iPSC-based biomarker strategy to identify neuroregenerative responders to allopregnanolone

School School of Pharmacy

Degree Doctor of Philosophy

Degree Program Clinical and Experimental Therapeutics

Publication Date 10/19/2019

Defense Date 10/06/2017

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

Tag allopregnanolone,Alzheimer's disease,biomarker,drug development,induced pluripotent stem cells,neuroregeneration,OAI-PMH Harvest,personalized medicine,precision medicine,translational science

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Brinton, Roberta Diaz (committee chair), Chui, Helena Chang (committee member), Ichida, Justin K. (committee member), Rodgers, Kathleen E. (committee member)

Creator Email christinesolinsky@rcsi.ie,solinsky@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-446768

Unique identifier UC11265497

Identifier etd-SolinskyCh-5846.pdf (filename),usctheses-c40-446768 (legacy record id)

Legacy Identifier etd-SolinskyCh-5846.pdf

Dmrecord 446768

Document Type Dissertation

Rights Solinsky, Christine Marie

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