University of Southern California Dissertations and Theses

Using engineered exosomes and gene-editing to target latent HIV

(USC Thesis Other)

Using engineered exosomes and gene-editing to target latent HIV

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Using Engineered Exosomes and Gene-Editing to Target Latent HIV
By Andrea Gadon
Advisor: Dr. Paula Cannon

A Thesis Presented to the
FACULTY of THE UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the Requirements for the Degree
MASTER OF SCIENCE
BIOCHEMISTRY AND MOLECULAR BIOLOGY
December 2016

Copyright 2016 Andrea Gadon

2

Table of Contents

Acknowledgements……………………………………………………………..….…….….…...4
List of Figures……………………………………..………………………….…….………...….5
Abstract…………………………………………………………………….……….……………6
Ch. 1 Introduction and Overview
1.1 Latent HIV…………………………………………………….…………..…….….7
1.2 Genome Editing Nucleases...……………………………………...……………....10
1.3 Engineered Exosomes…………………………………………………….…….…13
1.4 Project Objectives……………………………………………………….……..….17
Ch. 2 Exosome Isolation
2.1 Introduction………………………………………………………………..………18
2.2 Materials and Methods…………………………………………...……….……….20
2.2.1 Cell Culture…………………………………………...….….………….20
2.2.2 Ultracentrifugation…………………………………….….………….....20
2.2.3 Tangential Flow Filtration…………………………………...………….21
2.2.4 Western Blotting……………………………………………...…………21
2.2.5 Exosome Transfection…………………………………………...…...…22
2.2.6 Flow Cytometry…………………………………………….…...………22
2.3 Results………………………………………………………………..….…………22
2.5 Discussion…………………………………………………………….……..……..26
Ch. 3 Cargo Loading
3.1 Introduction………………………………………………………………………...33
3.2 Materials and Methods……………………………………………..…….…..…….35
3.2.1 Plasmids……………………………………………………………...….35
3.2.2 Restriction Site Cloning……………………………………………....…35
3.2.3 Enzymatic Mutation Detection Assay…………………………….……..36
3.3 Results…………………………………………………………………..………….36
3.4 Discussion……………………………………………………….……….…..….…39
3

Ch. 4 Targeting
4.1 Introduction…………………………………………………………………..….42
4.2 Materials and Methods……………………………………………..……...….…45
4.2.1 Plasmids……………………………………………………..……......45
4.2.2 In-Fusion Cloning of Fusion Plasmids………………...….........….…45
4.2.3 Engineering Targeted Exosomes……………………………..….…...45
4.2.4 Western Blotting………………………………………...………...….45
4.2.5 Flow Cytometry……………………………………….………..…….46
4.2.6 Statistical Analysis……………………………………..….….......…..46
4.3 Results………………………………………………………………………...…47
4.4 Discussion…………………………………………………………….……....…51
Summary............…………………………………………..….........................…..……….…54
References…………………………………………………………………………….……....57

4

Acknowledgements

I would like to express my deep gratitude to my advisor Dr. Paula Cannon for her mentorship and
expert guidance throughout this project, and I am especially thankful for the opportunity to train
in her lab.
I would also like to thank the members of the Cannon lab, particularly Drs. Cathy Wang, Danielle
Krasner, Eduardo Seclen, Geoffrey Rogers, and Nicholas Llewellyn. Their support and feedback
during all stages of this study, from experimental design to the editing of this manuscript, has been
invaluable.
Finally, I would like to extend my thanks to Dr. Joseph Hacia and Dr. William DePaolo for
graciously accepting to be members of my committee and dedicating their time and expertise to
reviewing this manuscript.

5

List of Figures
Figure 1. PD-1 as a surrogate marker of HIV latency 09
Figure 2. Classes of genome-editing nucleases 10
Figure 3. Disruption of integrated HIV with TALENs 12
Figure 4. Exosome Biogenesis 14
Figure 5. Exosome Isolation 23
Figure 6. Characterization of isolated exosomes 24
Figure 7. Exosome delivery of GFP 26
Figure 8. Comparison of cargo loading techniques 38
Figure 9. Exosome delivery of nucleases 39
Figure 10. Schematic of targeting experimental design 44
Figure 11. Characterization of targeted exosomes 47
Figure 12. Delivery of Engineered exosomes to PD-1+ cells (I) 48
Figure 13. Delivery of Engineered exosomes to PD-1+ cells (II) 50

6

Abstract
Antiretroviral therapy is extremely effective against human immunodeficiency virus (HIV), but
strategies to suppress the reactivation of the latent HIV reservoir are needed to obtain a drug-free
control of HIV infection. Genome-editing nucleases are a promising approach, as they are effective
at disrupting the latent proviral genome in vitro. The main hurdle, however, is ensuring delivery
of the nucleases in a cell- or tissue-specific manner. Exosomes are biogenic nanovesicles capable
of intercellular transport of nucleic acids and protein and have been engineered to target specific
cells in vivo. As such, in this thesis, we examine the hypothesis that engineered exosomes are
capable of delivering genome editing nucleases to latently infected cells in a targeted manner,
using programmed death 1 (PD-1) receptor as a surrogate marker of HIV latency. First, we
optimized a method to produce functional exosomes from the human embryonic kidney (HEK)
293T cell line using ultracentrifugation. Second, we compared methods of packaging our nuclease
cargo into exosomes and determined that direct transfection into exosomes achieved the most
efficient packaging of cargo into exosomes. For ease of testing, we use plasmid DNA encoding
GFP as cargo. Finally, we engineered exosomes with targeting capability using PD-1 ligand 1 (PD-
L1) that binds PD-1, fused to an exosomal membrane protein, either Lactadherin or Lamp2b. Our
evaluation of the delivery of these engineered exosomes carrying GFP to PD-1 expressing cells
led to the conclusion that although our untargeted exosomes achieved highly efficient gene
delivery, further work is required to enhance their targeted delivery. Ultimately, our study
evaluates a possible strategy for eradicating the latent HIV genome using an exosome-based
delivery system. However, additional studies on targeting capability need to be performed and
optimized in order to validate whether exosomes carrying genome-editing nucleases are capable
of targeted delivery in vivo.
7

Chapter 1
1.1 Latent HIV
HIV infection causes progressive failure of immune function via depletion of CD4
+
T cells.
Currently, there remains no available cure, but antiretroviral therapy (ART) can adequately
control viral load and significantly restrict infection and transmission (Fauci and Folkers,
2012). However, ART selectively targets viral proteins that are expressed at various stages
of the viral life cycle, thus can only target actively replicating virus. In addition, ART
exhibits poor drug penetrance into lymphoid tissues, where there is high frequency of
infected cells (Fletcher, 2014). Because of these complications, HIV is allowed to persist
as an integrated provirus in a latent state even on ART. Latent HIV is transcriptionally
inactive and does not replicate, therefore no viral proteins are produced that can be targeted
by ART. If ART is discontinued because of adverse metabolic, immunologic, and
neurologic side effects or improper drug management, the latent virus will reactivate and
restart infection of the host immune system within 2-3 weeks of stopping ART (Marin,
2009). Given this, it is important to develop a strategy to permanently eliminate HIV by
targeting HIV reservoirs, the pool of latent HIV present in the body. This will not only
obviously aid the health of individuals infected with HIV, but also reduce the economic
burden of treatment costs and eliminate the need for lifelong daily drug therapy, which
costs approximately $379,668 per person as of 2010 (Shackman, 2006).
Strategies to combat latent HIV encompass several areas of focus. Research focusing on
reactivation and subsequent elimination of latent virus, using latency reversing agents
(LRAs) such as PKC agonists (prostratin and bryostatin) or HDAC inhibitors (valproic
8

acid, SAHA) to render the virus detectable by the host immune system is together referred
to as the “shock and kill” method (Deeks, 2012). However, current attempts have achieved
only moderate success in vitro (Ho, 2013 and Shan, 2011). In addition, this method has not
shown to significantly reduce viral load in clinical trials (Rasmussen, 2016). This suggests
that the latently infected cells are somewhat resistant to immune clearance and viral-
activated cell death or that some cells are resistant to reactivation (Kimata, 2016).
Another approach to disrupting latent HIV is gene therapy, which involves using genome-
editing nucleases to disrupt the integrated HIV provirus in latently infected cells (Cary,
2016). Most studies using this approach target the HIV long terminal repeat (LTR), which
plays a crucial role in reverse transcription and integration of the viral genome (Karpinski,
2016). Importantly, LTRs are located on both ends of the integrated virus, providing
genome-editing nucleases with two target sequences. This approach is also being pursued
by other laboratories and by ours using genome-editing nucleases which may prove more
effective as they do not require viral activation as with the shock and kill approach.
To use genome editing to eradicate persistent HIV reservoirs, target cells first need to be
identified. A major challenge in latent HIV is that there are 10
6
-10
8
total latently infected
cells in the body (Massanella, 2016). Nevertheless, this represents a small fraction of the
total infected cells (estimated to be one latently infected cell for every 10
6
infected cells).
Latently infected cells are primarily found in resting memory CD4+ T cells, a subset of T
cells, which are the largest and most widely studied HIV reservoir (Chun, 1997). Their
ability to escape immune surveillance due to their longevity and quiescence poses a barrier
towards eliminating the HIV reservoir (Fletcher, 2014). Because latently infected cells do
not express viral proteins that can be targeted, potential markers are needed in order to
9

identify these rare latently infected cells (Ruelas, 2013). One of these surrogate markers is
programmed cell death-1 (PD-1). PD-1 is a major negative regulator of self-renewal and
expansion of memory T-cells (Kinter, 2008) and is enriched in the memory T-cells of HIV-
infected patients receiving long-term ART (Chomont, 2009).
We chose to target PD-1 because it has been confirmed in a mouse model of latency
developed in our lab by Dr. Llewellyn. To test the contribution of PD-1+ cells to the HIV
latent reservoir, 11 humanized mice were infected with reporter HIV that expresses an HA
tagged surface protein and ART treated (Figure 1). Spleens were then isolated and FACS
sorted for uninfected and latently infected (HA
-
) CD4
+
PD-1
+
and CD4
+
PD-1
-
cells. Those
cells were cultures in unstimulating or stimulating conditions to reactivate any latent HIV
and the latent virus was quantified. In these experiments, it was observed that the vast
majority of latent HIV was found in the PD-1
+
population, which shows that PD-1 is a
good surrogate marker of latently infected cells in our animal model.
Figure 1: PD-1 as a surrogate marker of HIV latency
11 humanized mice were infected with HIV and treated with ART. Spleens were harvested
and productively infected cells were removed. The remaining uninfected and latently infected
cells were sorted for human CD4
+
/PD-1
+
or CD4
+
/PD-1
-
cells. These cells were cultured in
unstimulating of stimulating conditions to reactivate HIV. Reactivated HIV from PD-1+ and
PD-1- cultures was quantified and compared. In general, over 80% of latent HIV was found
from PD-1+ cells. (Unpublished data performed by Dr. Llewellyn)
Ex-vivo latency-PD1 cells
1/2
3
4
5
6
7
8
9
10/11
0
20
40
60
80
100
%Latency PD1-
%Latency PD1+
%of Latent Virus
10

1. 2 Genome Editing
Genome editing nucleases can be used to achieve inactivation of target genes. Specifically,
these enzymes function by creating site-specific DNA double-strand breaks, which is
primarily repaired by the cellular pathway of non-homologous end-joining (NHEJ) (Gaj,
2013). NHEJ is highly error-prone and often produces insertions or deletions in the target
DNA sequence that can result in gene knockout (Wyman and Kanaar, 2006). There are
several classes of genome editing nuclease, which include zinc-finger nucleases (ZFNs),
transcription activator-like effector nucleases (TALENs), and the CRISPR/Cas9 system
(Figure 2). The primary difference between the various platforms is that ZFNs and
TALENs are based on a modular design, linking sequence-specific DNA binding domains
to non-specific DNA cleavage domain of a nuclease, typically the Fok1 enzyme. On the
other hand, the CRISPR/Cas9 system involves an RNA sequence guiding a Cas9 nuclease
to the desired target sequence (Ran, 2015).

Figure 2: Classes of genome-editing nucleases
Diagram showing the three genome-editing platforms. All colored portions of each
system indicates regions with sequence-specificity, while the grey portions are the
nucleases responsible for creating the double-stranded break in the target DNA
sequence.
11

Due to the ability of these nucleases to efficiently disrupt target genes with high specificity,
they can be used as therapeutic tools in HIV. Our lab has extensive expertise using these
genome-editing nucleases to disrupt integrated HIV genomes, which can be treated as
cellular genes. All three genome-editing systems have been used to disrupt integrated HIV
DNA by targeting the long terminal repeat (LTR) sequences present on both ends of the
integrated viral genome. Initial studies used a modified cre-recombinase altered to
recognize a specific sequence in the LTR. It was used to disrupt HIV integrated into HeLa
cells (Sarkar, 2007) and further validated in primary T cells and hematopoietic stem cells
(HSCs) (Hauber, 2013). However, it is restricted to target sequences similar to its natural
recognition sequence and is very challenging to change, making modified cre a difficult
platform to use. On the other hand, due to their modular nature, ZFNs and TALENs are
relatively simple to construct and target. ZFNs, targeting a highly conserved site found in
viral isolates from patient samples, have been reported to disrupt up to 60% of GFP-
expressing HIV genomes in T cell lines and 31% in primary T cells (Qu, 2013). Our lab
has also developed several TALENs that target the LTR and have been able to disrupt
integrated HIV in a cell line model of latency (JLAT cells), which express a GFP tagged
HIV after stimulation with TNF , using nucleofection of TALEN RNA (Figure 3). The
CRISPR/Cas9 system has been used in a similar manner to attain a 25-32% reduction in
GFP-expressing HIV genomes in primary T cells (Ebina, 2013).
In order to test the ability of genome-editing nucleases for gene disruption in cell lines that
lack HIV LTRs, we will use another established nuclease system against CCR5, a
coreceptor required for HIV entry into cells (Bjorndal, 1997). The naturally occurring
CCR5 ∆32 mutation that results in a truncated protein confers protection against HIV
12

infection without detrimental health consequences (Samson, 1996). We have proven highly
efficient gene disruption technology using ZFNs to disrupt the CCR5 gene in human
hematopoietic stem cells (HSCs), progenitors of the target cells of HIV (Holt, 2010). Our
laboratory has shown that gene knockout of CCR5 in those HSCs in a humanized mouse
model results in a decrease in HIV replication and leads to protection against HIV infection.
As such, our validated ZFNs against CCR5 are important tools for genome editing in cell
lines.

Even though there is reasonable success in using genome-editing nucleases for gene
disruption in vitro, delivering the nucleases in vivo in a cell-specific manner is a significant
challenge (Wang, 2015). Control of the specific delivery to target cells remains an active
area of research in gene therapy. Previously, our lab has used viral vectors based on adeno-
associated virus (AAV) and lentivirus. Lentiviral vectors are commonly used delivery
systems that achieve moderate ability of targeting in vitro, but carry with it potential
disadvantages such as low transduction efficiency in vivo, genome integration that may
cause insertional mutagenesis, and immunogenicity (Sakuma, 2012). Even AAV vectors,
Control
+TNF 
RU5
GFP
TAR
TATA Mock
21.6% 72.0% 28.4% 23.4% 46.1%
Figure 3 Disruption of integrated HIV with TALENs
Candidate TALENs targeted to different regions of the HIV LTR, TAR, TATA, and RU5, were
nucleofected into a cell line model of latency (JLAT cells). After activation with TNF , which
promotes expression of GFP-tagged HIV in JLAT cells, GFP expression was measured by flow
cytometry.
13

which are stable and have low immunogenicity, has a small packaging limit and broad
tissue tropism (Mingozzi, 2011). An alternative delivery platform are liposomes, synthetic
nanoparticles that can be loaded with nucleic acid or drugs. Unfortunately, the artificial
membrane-like polyethylene glycol (PEG) corona intended to increase liposomal
circulation time upon systemic administration is also linked to an induction of immune
response (Dugan and Keating, 2011). However, their endogenous counterparts, exosomes,
are biogenic and naturally transport functional nucleic acids between cells (Tan, 2013).
Exosomes are a novel approach for use in gene therapy the next section elucidates their
advantages as a delivery system. Broadly, there are a variety of available delivery systems
each with its own set of benefits and restrictions, and it is important to find the optimal
approach of delivering genome-editing nucleases to PD-1 expressing HIV reservoirs.
1.3 Engineered Exosomes
Exosomes are a class of extracellular vesicles actively released by virtually all cells and
found in various body fluids, such as urine, breast milk, amniotic fluid, and blood (Wistwer,
2013). Extracellular vesicles generally fall into two classes separated by size:
microvesicles, which are 100-1000nm in diameter, and exosomes, roughly 40-120nm
(Simpson, 2009). The two classes differ in their biogenesis, as exosomes are derived from
the late endosome while microvesicles bud from the plasma membrane (Andaloussi, 2013).
Inside their parent cell, exosomes are intraluminal vesicles (ILVs) housed inside
multivesicular bodies (MVBs), which are released into the extracellular environment upon
MVB fusion with the plasma membrane (Figure 4). One hallmark property of exosomes
is they mediate intercellular communication, being naturally capable of transporting a
variety of functional cargo between cells, including nucleic acids and proteins (Thery,
14

2009). Exosomes are stable in the blood, able to shield their cargo from degrading enzymes
present in the bloodstream. These characteristics make them attractive candidates as
delivery vehicles (Batrakova, 2015).

Excitingly, exosomes have many advantages over current methods of delivering
engineered nucleases. The targeted nucleases used as exosome cargo in this study do not
require permanent expression to achieve their desired effect, giving exosomes an advantage
over integrating lentiviral vectors as transient expression of nucleases will lead to fewer
off-target effects in vivo. Exosomes are also generally larger in size than the 20nm AAV
vectors, allowing for a larger payload capacity. Also, syngeneic exosomes do not induce
neutralizing antibody responses or activate the complement system, and thus are stable
inside the body unlike liposomes. As a result, encapsulated cargo are protected from
Figure 4 Exosome Biogenesis
Exosomes are formed when the membrane of the early endosome buds inward,
forming an internal vesicle called intraluminal vesicles (ILV). Multivesicular
endosomes (MVE) are endosomes containing ILVs and can traffic along the
endosomal pathway to the lysosome or fuse with the plasma membrane,
releasing their contents including ILVs into the extracellular space where they
are considered exosomes. (Raposo and Stoorvogel, 2013)
15

degrading enzymes present in the body, making exosomes a promising delivery tool.
Exosomes have been implicated in a number of cancer immunotherapies (Johnsen, 2014)
and drug delivery (Haney, 2015). Their low immunogenicity and low toxicity have proven
to be safe for patients, as a number of therapeutic autologous exosomes have passed Phase
I in clinical trials (Bell, 2016).
In order to use exosomes as a delivery system to target latent HIV, we need to load the
genome-editing nucleases as cargo, as well as confer PD-1 targeting capability onto the
exosomes. To do this, exosomes are isolated from cell culture supernatant and can be
modified to encapsulate specific cargo, as well as incorporate targeting ligands on its
membrane surface (Kooijmans, 2016). Both the cargo loading and targeting aspects of
exosomes necessitate thorough evaluation in order to determine the applicability of
exosome delivery to gene therapy against HIV.
Importantly, modified exosomes have been shown to successfully deliver functional
nucleic acid and drugs to specific cells in vivo. This is done by transfection of the producer
cell, or post-isolation directly transfecting into exosomes. Inclusion of the nucleic acid
‘payload’ can be achieved by bulk over-expression of DNA or mRNA in the producer cell
(Delcayre, 2005). Nucleic acids are also easily incorporated by post-isolation transfection
or electroporation, including DNA or mRNA encoding GFP (Mulcahy, 2014). Given this,
we use exosomes to DNA and mRNA encoding transport reporter GFP and genome-editing
nucleases to specific relevant cell types. Because of this ability to create exosomes carrying
nucleic acid and protein cargo, exosomes have been exploited as vaccine tools for systemic
delivery and has major implications as delivery vehicles in gene therapy.
16

Targeted exosomes are a novel approach to in vivo delivery and is typically achieved by
engineering ligands into membrane proteins that are naturally enriched in exosomes, such
as LAMP2 or lactadherin (Rountree, 2011). Exosomes need to express targeting peptides
in order to achieve specific delivery in vivo, as systemic injection into mice drives
untargeted exosomes into the liver, kidney, and spleen. The first reported targeted
exosomes demonstrated successful delivery in vivo using exosomes carrying siRNA
against a target gene in the brain, suggesting that the targeted exosomes can cross the blood-
brain barrier (Alvarez-Erviti, 2011). Using labeled siRNA cargo, they reported that IV
injected exosomes were found to deliver siRNA to neurons, microglia and
oligodendrocytes to achieve a 60-70% knock down of their target gene in vivo. Targeted
exosomes were also developed to deliver miRNA to EGFR-expressing breast cancer cells
(Ohno, 2013). The exosomes effectively delivered anti-tumor miRNA to tumor-bearing
RAG2
-/-
mice and inhibited tumor formation, further validating the targeting ability of
exosomes. Targeting to PD-1 expressing cells has not been previously reported, and is
important to evaluate in order to create engineered exosomes capable of inactivating the
latent HIV reservoir.
In summary, using exosomes as a candidate targetable delivery system to deliver nucleases
to latent HIV reservoirs holds promise, but is still in its infancy. Currently, there is no
effective delivery system capable of transporting genome editing nucleases to latently
infected cells in vivo. This may be due to the absence of validated cell surface markers that
characterize latency, as well as the difficulty of delivering to latently infected cells that are
situated in tissues largely impenetrable by ART. As biogenic delivery vehicles, exosomes
may be uniquely suitable to deliver nucleases to latently infected T cells.
17

1.4 Project Objectives
The goal of this study is to develop a strategy to deliver genome-editing nucleases to HIV
reservoirs using an exosome-based delivery system. To target specific cells, exosomes can
be modified by incorporating targeting ligands, including peptides or antibodies capable of
binding to these target cells, fused to an exosomal membrane protein. To disrupt latent
HIV, we have chosen to the PD-1 receptor as a putative marker as it has been found to be
enriched in latently infected cells. The exosome cargo will then be the DNA or mRNA
transcripts that expresses reporter GFP as well as the specific anti-HIV ZFNs, TALENs, or
Cas9/gRNA. This study investigates the capability of exosomes to target cells latently
infected with HIV by (i) optimizing exosome production, (ii) improving methods of
packaging cargo, and (iii) examining efficiency of targeting in vitro. We hypothesize that
PD-L1 targeted exosomes will be able to deliver nucleic acid cargo specifically to PD-
1 expressing cells. This research will mark one of the first cases of exosomes targeted to
T cells, and its success will pioneer a novel delivery system for genome-editing nucleases.
Overall, I hope to determine the applicability of targeted exosome delivery to gene therapy
against HIV.

18

Chapter 2 Exosome Isolation
2.1 Introduction
Exosomes are a promising delivery tool because of their ability for intercellular nucleic
acid transport; however, their isolation poses a challenging factor as they are difficult to
distinguish from other extracellular vesicles and proteins of similar size (Tauro, 2012). As
such, various isolation techniques were evaluated to determine the optimal method for
obtaining pure exosome isolates. Strategies to effectively isolate exosomes from cell
culture supernatant vary with regards to time, throughput, and efficiency (Zeringer, 2015).
Each method effectively falls into one of two categories: centrifuge-based or pressure-
driven. Centrifuge-based systems rely on centripetal force to sediment molecules of
different densities. Of these, ultracentrifugation is the gold standard for exosome isolation
(Livshts, 2015). It involves a series of centrifugations at increasing speeds to remove dead
cells, cell debris, and contaminating vesicles before ultracentrifugation to pellet exosomes
(Thery, 2006). Pressure-driven systems such as ultrafiltration uses pressure to concentrate
fluids through membranes of various sizes (Heinemann, 2014). A novel technique,
tangential flow filtration (TFF), uses a peristaltic pump in a pressure-driven system (Rao,
2012). TFF has not been previously reported to isolate exosomes but has been successful
in concentrating lentiviral vectors, which are similar to exosomes in size and makeup
(Cooper, 2011). As exosomes are difficult to distinguish from other molecules of similar
size making co-isolation is a common issue, it is important to carefully select the method
of exosome isolation (Raposo, 2013).
19

The presence of exosomes can be confirmed by identifying proteins ubiquitous in
exosomes regardless of the cell of origin. In general, exosomes are abundant in heat shock,
cytoskeletal, cytosolic, and plasma membrane proteins, as well as proteins involved in the
endosomal sorting complexes required for transport (ESCRT). There is no exclusive
marker for exosomes, but a combination of proteins enriched in exosomes are commonly
used, such as CD63 and TSG101 (Lee, 2012). CD63 is a protein in the tetraspanin family
of membrane proteins. TSG101 is part of the endosomal sorting complex required for
transport (ESCRT) complex regulating vesicular trafficking, located inside the exosome.
(Sokolova, 2011). Using these proteins as markers of exosomes is vital because cell culture
supernatant and other biological fluids contain a number of similarly-sized extracellular
vesicles such as microvesicles or apoptotic vesicles, that are not enriched in these proteins.
Here, we compared the purity of exosomes isolated through a centrifuge-based method of
exosome isolation and a pressure-based system using the human embryonic kidney (HEK)
293T cell line. We tested an established ultracentrifugation protocol which involves
filtration of the cell culture supernatant followed by two ultracentrifugation steps (Thery,
2006). However, this process is lengthy and repeated ultracentrifugation steps may damage
and reduce the quality of isolated exosomes (Lobb, 2015). In addition, ultracentrifugation
cannot process large volumes. Therefore, we also tested the TFF system, a promising new
ultrafiltration technology able to handle volumes of several liters, allowing for high
throughput isolation. The comparison of these two methods for exosome isolation allowed
for the conclusion that ultracentrifugation results in higher purity in exosome isolation.
Upon demonstrating success in exosome production, we then confirmed functionality by
20

proving their ability to delivery DNA and mRNA cargo to target cells is at levels
comparable to commercially available purified exosomes.

2.2 Materials and Methods
2.2.1 Cell Culture
The human embryonic kidney (HEK) 293T cell line (ATCC) was used in this study
as they are established as efficient producers of exosomes (Sokolova, 2011). The
cells were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented
with 10% fetal bovine serum (FBS) as well as penicillin-streptomycin and
incubated at 37°C and 5% CO2 atmosphere. Prior to exosome isolation, the cell
culture medium was changed to DMEM supplemented with 10% exosome-free
FBS and cultured for 48 hours. To produce exosome-free FBS, FBS was
ultracentrifuged for 70 minutes at 100,000g to remove presence of contaminating
bovine exosomes.
2.2.2 Ultracentrifugation
Exosomes were isolated as previously described (Thery, 2006). 10 x 10
6
HEK 293T
cells were seeded in T150 flasks in DMEM supplemented with exosome-free FBS.
After 48 hour incubation, 30ml cell culture supernatant was collected and
centrifuged at 2,000g for 10 min and filtered once with a 0.22um PVDF membrane
filter (Millipore). The supernatant was ultracentrifuged for 70 minutes at 100,000g
at 4C using an XL-90 ultracentrifuge (Beckman) on a SW-28 Ti swinging bucket
21

rotor to pellet exosomes then washed with 30ml PBS and a final ultracentrifugation
for 70 minutes at 100,000g at 4C to pellet exosomes. Exosome pellets were
resuspended in 50-100ul PBS or RIPA buffer for future use.
2.2.3 Tangential Flow Filtration
10 x 10
6
HEK 293T cells were seeded in T150 flasks cultured in DMEM
supplemented with exosome-free FBS. After 48 hour incubation, cell culture
supernatant was collected and centrifuged at 2,000g for 10 min to remove dead cells
and filtered once with a 0.22um PVDF membrane filter (Millipore) to remove
contaminating vesicles. 30ml of starting supernatant was filtered through a
tangential flow filtration system (Spectrum) through a mPES 100kDa hollow fiber
filter and buffer exchanged with an equal amount of 1x PBS before concentrating
into 1ml PBS.
2.2.4 Western Blotting
Exosomes were lysed with RIPA buffer (Thermo Fisher) with PMSF protease
inhibitors (Sigma Aldrich). Protein concentration was determined using DC Protein
Assay reagents (Biorad) using a Berthold Mithras LB940 plate reader. Absorbance
was measured at 620nm in triplicate and analyzed using Mikrowin 2000 software.
30 g of exosome lysate and 50 g of cell lysate was then subjected to
electrophoresis using Criterion TGX gels (BioRad). Membranes were probed with
rabbit anti-CD63 (ab134045), rabbit anti-TSG101 (ab30871), and rabbit anti-
GM130 (ab52649) antibodies (Abcam) and incubated with goat anti-rabbit IgG
HRP antibody (sc-2004) secondary antibody (Santa Cruz Biotechnology). All
22

membranes with blocked with 5% milk and washed in PBST after each incubation
step. Western blots were visualized with Konica Minolta SRX-101A Imager
(Freedom Imaging).
2.2.5 Exosome Transfection
Exosomes were transfected using ExoFect solution as per manufacturer’s protocols
(SBI). Isolated exosomes were mixed with 1ug mRNA or 5ug plasmid DNA,
incubated at 37C for 10 minutes then chilled at 4C for 30 minutes and exosomes
are precipitated and resupended to yield 150ul of transfected exosomes. 50ul
transfected exosomes were added to approximately 2x10
5
cells and incubated for
24 hours.
2.2.6 Flow Cytometry
Cells were collected and analyzed for GFP expression by flow cytometry using
Guava EasyCyte 6-21 (EMD Millipore). FlowJo software version X (Treestar) was
used for data analysis. In each independent analysis, the criteria for gating of GFP
positive cell populations was determined for 0.1% or fewer of untreated or mock
treated cell populations cultured in parallel.

2.3 Results
With the idea of becoming proficient on producing our own exosomes, we compared a
centrifugation-based method to a pressure-driven system of exosome isolation. Total
protein yield obtained from each isolation method was quantified using a Bradford assay.
23

Previous reports of exosome isolation from HEK293 cells using ultracentrifugation
parallels our yields (Ohno, 2013). The TFF-purified exosome samples showed a 3-fold
higher total protein yield from exosomes isolated through ultracentrifugation using the
same starting amount of cells and cell culture supernatant (Figure 5).

Western blot analysis validated the presence of exosomes in our purified exosomes using
commonly used exosome markers CD63 and TSG101 (Figure 6). It is important to note
that although there is no known exclusive marker for exosomes, a combination of proteins
enriched in exosomes are used in their characterization and identification, specifically
CD63 and TSG101 (Lobb, 2015). We confirmed the presence of exosomes in the purified
Figure 5 Exosome Isolation
Exosomes from 50 x 10
6
HEK 293T cells were isolated from cell culture supernatant and subject to
centrifugation and filtration prior to ultracentrifugation (UC) or tangential flow filtration (TFF) using the
parameters shown (left figure). The starting and final amounts as well as total protein recovered are listed
(right table). TFF resulted in 15x greater recovery of total protein than UC given the same amount of
starting amount of cell culture supernatant.
24

samples using antibodies against these markers (Zoller, 2008). We used GM130, a marker
for a cisGolgi protein, as a negative control as Golgi proteins are generally not present in
exosomes. As expected, the western blots for both TFF-purified and ultracentrifuge-
purified exosomes show the exosomal proteins CD63 (26kD) and TSG101 (44kD) and do
not show band the Golgi protein GM130 (120kD). However, for the same amount of total
protein, the amount of exosomal proteins in the TFF-purified exosomes resulted in a much
fainter band than the ultracentrifuge-purified exosomes. Western blot analysis of the
unpurified HEK 293T cell lysate demonstrated the presence of all three proteins, as
expected. Our conclusion is that while TFF recovered a much larger amount of overall
protein from 30ml of cell culture supernatant derived from 50 million 293T cells, it was
enriched in exosomal proteins to a lesser degree than ultracentrifuge-derived exosomes
from the same amount of supernatant and cells. Thus, ultracentrifugation was determined
to be more successful at producing exosome preparations of higher purity and is therefore
used for the duration of this study.

Upon identifying an efficient method to successfully purify exosomes though
ultracentrifugation, the next step is to examine their functionality in cells. To do so, we
tested their capability to deliver functional DNA and mRNA encoding reporter GFP to
Figure 6 Characterization of isolated exosomes
HEK293T purified exosomes were isolated by
ultracentrifugation (UC) and tangential flow filtration
(TFF). 30 g of exosome lysate and 50 g of cell lysate were
examined for the presence of exosome markers CD63,
TSG101, and cis-Golgi marker GM130. Presence of
exosome proteins in exosome and cell lysates, as well as cis-
Golgi protein in cell lysate were as expected. Reduced
exosome proteins enriched in TFF exosome lysate than UC
exosome lysate given same total protein suggests decreased
purity.
25

recipient 293T cells. We used a post-isolation transfection reagent ExoFect (System
Biosciences) to load GFP, in either DNA or mRNA form, into exosomes. The presence of
GFP in cells treated with GFP-loaded exosomes compared to control exosomes lacking
GFP cargo indicates that the GFP-loaded exosomes are able to enter the cells and deliver
DNA and mRNA cargo. In order to establish a baseline level of delivery efficiency, we
performed parallel experiments using commercially available exosomes isolated through
polymer-based precipitation (Tauro, 2012). Our results demonstrate that cells treated with
GFP-loaded exosomes showed 70-80% GFP fluorescence, and these numbers correspond
to that of exosomes developed for commercial use (Figure 7). Surprisingly, when we
delivered the GFP-loaded commercial exosomes to K562, a leukemia suspension cell line,
GFP expression was nominal. Overall, the ability of our exosomes to deliver GFP cargo
to recipient cells validates their functionality and establishes our capability to effectively
produce functional exosomes.

26

2.4 Discussion
Because exosome isolation from cell culture supernatant lacks a standardized purification
method, exosome purification can be performed with a diverse number of methods. To
maintain quality control of our laboratory-produced exosome isolates, it is important to
Figure 7 Exosome delivery of GFP
Commercially available (50-150ug) and
laboratory produced exosomes (50ug) were
transfected with GFP plasmid (1ug) or GFP
mRNA (5ug) using ExoFect reagent (SBI)
and added to 2 x 10
5
recipient 293T cells
cultured in exosome-free culture media
(top). Flow cytometry on recipient cells was
performed 24 hours after exosome addition.
One independent study was performed. The
experiment was repeated with commercial
exosomes on recipient 293T and K562 cells
in one independent study (middle, bottom).
Overall, the results show that our
laboratory-produced exosome isolated
through ultracentrifugation were able to
deliver GFP DNA and mRNA at
comparable levels to commercially
available exosomes. However,
commercially available exosomes were
unsuccessful at delivering GFP DNA and
mRNA to K562 cells, a human leukemia
suspension cell line.
293T Mock 293T DNA 293T RNA
SSC
9% of events were out of range
K562 Mock K562 DNA K562 RNA
66 0.1 72
0.87 0.88
GFP
0.1
27

fully evaluate our production system and comprehensively characterize exosome recovery
and purity. In this study, our exosomes were isolated from HEK 293T cells, an adherent
epithelial embryonic kidney cell line established in the literature and used commercially as
an exosome producer cell. Each isolation involved harvesting cell culture supernatant from
cells cultured in media free of bovine exosomes from serum to avoid contamination of the
final exosome isolates. After an initial spin to discard cells and cell debris, removal of
particulates larger than 220nm was performed using filtration with a low-protein binding
polyvinylidene fluoride (PVDF) membrane to retain the maximum amount of protein prior
to ultracentrifugation or tangential flow filtration.
Comparison of the commonly used ultracentrifugation compared with the newer tangential
flow filtration technology showed that although TFF produced 3-fold higher protein yield
using a Bradford assay, it also had a much lower ratio of exosome particles to protein using
Western blot analysis. This is potentially a result of several factors. Ultracentrifugation
sediments particles based on density, while TFF concentrates on the basis of size. The
100kDa membrane filter in TFF theoretically discards small contaminating proteins
commonly present in serum, such as the 66kDa bovine serum albumin (BSA), but it is
possible that protein aggregation occurred and resulted in co-isolation with exosomes. It
would be useful to determine the identity of the contaminants and whether they are
potentially toxic. Pressure-induced membrane shearing in tangential flow filtration has
been proposed to occur as well, but is not fully assessed (Lobb, 2015). In addition, a loss
of exosome isolates is likely attributable to the exosomes being trapped in the membrane
filter. For this reason, it is possible that TFF would be more effective at isolating exosomes
using a larger volume of starting cell culture supernatant up to 5 liters, as the first 50-100ml
28

may adhere to the membrane and result in less recovery. Overall, ultracentrifugation
resulted in less recovery but higher purity with less co-isolation of large molecules or
particulates.
Ultracentrifugation presents its own set of caveats with regards to exosome isolation. One
challenge is that ultracentrifugation will co-purify any non-exosomal vesicles and protein
present in the cell culture media. Modified ultracentrifugation protocols may include
sucrose cushions or sucrose or iodoxinal gradients to improve on the purity of the exosomes
isolated from centrifugation (Jeppesen, 2014). However, this presents a double-edged
sword, as increasing the purity of the exosome isolates with additional steps caused loss of
exosome sample material in an attempt to separate exosomes from other vesicles and
particles, as reported by Jeppesen et al. Therefore, all of our ultracentrifugation steps were
preceded by an established initial spin to centrifuge exosomes and using a 0.22um pore
PVDF filter to discard as much contaminants as possible, as there is little evidence in the
literature to assume that co-isolated particles interfere with exosome function.
Another important factor to evaluate is the effect of the high centrifugal forces present in
ultracentrifugation on the integrity of the isolated exosomes. Changes in exosome
morphology have been suggested as a direct result from the high g-force exerted during
ultracentrifugation (Lobb, 2015). Repeated 2-hour ultracentrifugation, up to four times,
were performed on HEK 293T cell culture supernatant to evaluate exosome stability and
were found not to cause significant changes in size according to NTA analysis (Sokolova,
2011). However, general knowledge on the effect of ultracentrifugation on exosome
proteome or morphology is limited and requires further investigation. Herein, all
29

ultracentrifugation steps were carried on for 70 minutes as previous reports suggest that
longer ultracentrifugation does not lead to an increase in exosome yield (Jeppesen, 2014).
On the other hand, Sokolova et al. determined that the storage of exosome isolates plays a
role in their stability. Exosomes were found to be more stable at 4C than exosomes stored
at 37C, as measured by exosomes size through NTA analysis. The effect is consistent up
to 4 days, until which exosomes exhibit a reduction in size from 120nm to 80nm, down to
50nm after 25 days. For this reason, upon the final ultracentrifugation step we immediately
resuspend our exosomes in PBS and load them with cargo for treatment to recipient cells
during the same day as the exosome isolation to minimize a reduction in exosome stability
(Sokolova, 2011).
In this project, our exosome isolates are evaluated by two criteria: total protein
quantification and immunoblotting for exosome positive and negative markers. After our
second ultracentrifugation with PBS, the PBS is carefully decanted and the exosome pellet
immediately resuspended in PBS or RIPA buffer. Each exosome preparation was subject
to a Bradford assay with comparison against bovine serum albumin (BSA) standard curve
to determine quantify total protein. The typical amount of total protein concentration from
50 x10
6
cells and 30ml supernatant ranged from 50 to 70 g in a final volume of 50ul and
normalized to 50 g using PBS to dilute. Exosomes, being of endocytic origin, express
CD63 and TSG101, which we use as exosome markers to demonstrate the success of our
exosome isolation strategies. CD63, a membrane protein, and TSG101, a cytosolic protein,
together confirm two regions of exosomes. Still, evaluation of two criteria are the minimum
experimental requirements set forth by the Journal of Extracellular Vesicles (Lotvall,
2014). To further improve our methods, it would be useful to characterize our purified
30

exosomes by examining other characteristics to improve exosome quality control. One
direction is to inspect exosome heterogeneity by examining vesicle size distribution.
Several techniques have been used previously to characterize exosome isolates from a
variety of cell sources, as the nano-scale size of exosomes poses a problem regarding their
isolation and quantification. These include scanning electron microscopy (SEM), atomic
force microscopy (AFM), dynamic light scattering (DLS), and nanoparticle tracking
analysis (NTA) (Lotvall, 2014). Of these, NTA has proven to be a valuable technique to
characterize exosomes and other nano-sized vesicles. NTA essentially measures the
hydrodynamic diameter of exosomes in their respective solution on the basis of Brownian
motion and can provide a robust estimation of the particle size. Ultracentrifugation of 30ml
cell culture supernatant results in an exosome pellet not visible to the naked eye, so the
added measurement of size distribution of our exosome isolates would be helpful to
characterize our exosomes.
Additionally, it would be beneficial to profile the protein content of the purified exosomes
to compare the proteome between exosomes isolated by ultracentrifugation compared to
tangential flow filtration. Exosomes are expected to retain cargo proteins and nucleic acids
representative of their cell of origin, and as such, exosomes from different producer cells
may have a diverse array of proteomic profiles (Xu, 2015). In summary, it is important to
realize that there are a myriad of additional ways to characterize and test for quality control
of exosomes, but for our purposes, validating our exosomes though total protein
quantification, presence of exosome markers, and functionality sufficiently establishes the
proper groundwork for our study.
31

Our results showing successful exosome delivery of DNA and mRNA encoding GFP in
HEK 293T cells is encouraging, particularly as the levels correspond with delivery
efficiencies of commercially available pre-isolated exosomes. In both cases, a chemical-
based transfection reagent was used to introduce nucleic acids into the exosomes by using
charged cationic polymers that bind to the negatively charged nucleic acids, resulting in a
complex that gets taken up into the exosomes (Peterson, 2015). Afterwards, the transfected
exosomes were added to recipient cells and measured for GFP fluorescence 24 hours later.
It would be useful to measure the amount of DNA or mRNA encoding GFP inside the
transfected exosomes at various time points to expand our knowledge of the transfection
efficiency into exosomes.
Changing the exosome cell of origin or having mismatched producer and recipient cells
may modify the properties of the exosomes as seen when we delivered GFP-loaded
exosomes to K562 cells. Exosomes derived from K562 cells have been used to treat human
umbilical vein endothelial cells to measure tube formation in angiogenesis (Mineo, 2011
and Tadokoro, 2013), suggesting that K562 cells can function as producer cells. However,
while HEK 293T and its parent cell line HEK 293 cells have been commonly used as
exosome producer cells, K562s have been less studied. This, in turn, may lead to different
delivery efficiencies dependent on the exosome cell of origin and further tests would be
needed to examine these changes. (Tan, 2013). Throughout the rest of this project, our
exosomes were produced and delivered using the same cell line throughout this study, and
as a result our findings are applicable specifically to 293T cells. Therefore, additional
experiments testing the optimal packaging and targeting strategies need to be reevaluated
32

when switching producer cells, for instance, when moving onto primary cells or in vivo
studies. as the properties of exosomes reflects largely on cell of origin.
Finally, it is important to note that as these exosomes are designed for future gene therapy,
a careful analysis of toxicity should be done in order to realize exosomes as a viable
therapeutic. This could be performed with viability staining of recipient cells with 7-AAD
or propidium iodide (PI) followed by flow cytometric analysis. If results show toxicity, a
potential solution is to use exosomes derived from immature murine dendritic cells, as they
produce exosomes lacking T-cell activators such as MHC-I and CD86, which provokes an
immune response in the host (Quah and O’Neill, 2005). Our preliminary studies on
exosomes did not yield major cell death, but if our studies progress to primary cells or in
vivo, further experiments quantifying apoptosis due to exosome treatment are important to
verify the therapeutic use of exosomes.

33

Chapter 3 Cargo Loading
3.1 Introduction
Exosomes were first discovered in 1983 and were initially thought to be secreted by cells
as part of a cellular waste system. It was not until 2007 when they were revealed to also
play a role in cell-cell communication, by exchange of functional miRNA and mRNA in
mouse and human mast cells (Valadi, 2007). They have since been found to mediate
intercellular delivery of a variety of cargo, including siRNA, lipids, and proteins (Taylor
and Gercel-Taylor, 2013) in various cell types (Ekstrom, 2012) to influence the expression
and function of recipient cells. As an extension of this ability, exosomes play a role in the
regulation of immune recognition and response, promotion of wound healing and normal
homeostasis, and mediation of tumorigenesis, inflammation and other chronic conditions
(Tian, 2014). Therefore, harnessing the fundamental ability of exosomes to modulate
intercellular communication, signaling, and regulation has potential therapeutic
applications for gene therapy.
As discussed in the previous chapter, one way to load cargo onto exosomes is directly
transfecting exosomes using a chemical-based cationic polymer ExoFect (SBI), which we
have determined works for both DNA and mRNA GFP cargo (Peterson, 2015). Other
potential methods include electroporation, incubation, and bulk-overexpression in
producer cells (Mulcahy, 2014). The latter involves transfecting producer cells with the
plasmid encoding the payload and subsequently isolating exosomes. This effectively
exploits the natural biogenesis of exosomes to encapsulate cytoplasmic contents of the cell
by using a mass action driving force to achieve nonspecific incorporation of cargo into
exosomes. Recently, it was found that exosomes secreted by human cells transport mRNA
34

fragments that are largely enriched in the 3’-untranslated region, or UTR (Batagov, 2013).
The 3′ UTR of mRNA transcripts typically contain regulatory elements called zipcodes,
regulatory sequences governing the mRNA localization within a cell (Ekstrom, 2012).
While it still remains unclear how mRNA is directed to exosomes, a zipcode-like sequence,
which when inserted into the 3’UTR of mRNA, has been shown to direct the mRNA
transcript to extracellular vesicles, a group of 40-250nm membrane vesicles which includes
exosomes (Bolukbasi, 2012). This 25-nucleotide sequence was first discovered during a
microarray analysis showing enrichment in extracellular vesicles isolated from primary
glioblastoma and melamona cells. The two components crucial to enrich mRNA transcripts
to vesicles was determined to be the presence of a “CTGCC” core region that forms a stem-
loop structure, as well as a binding site for microRNA-1289. Studies on intracellular
mRNA localization find that miRNA can interact with mRNA to suppress mRNA
translation and localize mRNA to distinct cytoplasmic domains (Martin and Ephrussi,
2009). Essentially, this zipcode-like sequence serves to direct mRNA transcripts to
extracellular vesicles, but the exact mechanism is not known. Dr. Bolukbasi and his team
reported that HEK 293T cells transfected with a GFP reporter plasmid modified to contain
the inserted sequence in the 3’UTR resulted in an approximately a two-fold increase of
GFP mRNA in extracellular vesicles when compared to exosomes derived from producer
cells transfected with a control plasmid (Bolukbasi, 2012). We plan to include this zipcode
sequence to DNA plasmids encoding GFP and nucleases as a method to produce mRNA-
loaded exosomes.
Essentially, our goal is to elucidate the optimal method of packaging our nucleic acid cargo
into exosomes and characterize its packaging efficiency. We engineer cargo plasmids
35

containing a zipcode-like sequence to test whether the specialized motif works as a
consensus sequence that improves packaging of mRNA into exosomes. The mRNA is
incorporated into exosomes through transfection of plasmids into producer cells. We
hypothesize that the addition of the zipcode sequence will result in increased mRNA
localization to exosomes based on the finding that the zipcode sequence was enriched in
mRNA transcripts inside extracellular vesicles produced by primary glioblastoma and
melanoma cells. We then compare the cargo loading efficacy of our nucleic cargo packaged
through producer cells to cargo directly transfected into pre-isolated exosomes and found
that direct transfection into exosomes resulted in higher delivery efficiency.

3.2 Materials and Methods
3.2.1 Plasmids
ZFNs targeting CCR5 have been described previously (Perez, 2008). Expression
cassettes for Cas9 from Staphylococcus aureus, zsgreen1 and ZFNs targeting CCR5
were cloned into a modified version of plasmid pVAX (ThermoFisher) driven by a
CMV promoter/enhancer. For the plasmid expressing gRNA against CCR5, the
Cas9 expression cassette was removed from a commercially available plasmid
encoding a single vector AAV-Cas9 system with Cas9 and its single guide RNA
(Addgene plasmid #61591), where its expression was driven by a U6 promoter.
3.2.2 Restriction Site Cloning
To generate modified plasmids containing the
ACCCTGCCGCCTGGACTCCGCCTGT zipcode sequence, restriction site
cloning was performed to insert the sequence into XhoI sites located in the 3’UTR
36

in the plasmids. The location was determined to be between the coding sequence
for GFP and the polyadenylation tail sequence described previously (Bolukbasi,
2012). Synthesized oligos with the sequence flanked by XhoI restriction sites as the
insert, and expression plasmids for GFP as the vector. The plasmid was digested at
the XhoI sites at the 3’UTR, isolated and purified using a Gel Extraction kit
(Qiagen) and the insert ligated into the vector using reagents from Fast-Link DNA
Ligation kit (Epicentre). A diagnostic restriction digest and Sanger sequencing was
performed on the finished plasmid for size and verification of insert orientation.
3.2.3 Enzymatic Mutation Detection Assay
Cells were harvested 3 days after exosome treatment. Genomic DNA was extracted
from cell pellets using a DNeasy Blood and Tissue kit (Qiagen) and PCR amplified
using human CCR5-specific primers. The PCR product was then denatured and
reannealed to allow for small DNA mismatch annealing, which is then cleaved
using reagents from GeneArt Genomic Cleavage Detection Kit (Thermo Fisher
Scientific) per manufacturer’s instructions. The digested PCR product was then
subjected to electrophoresis in a 10% polyacrylamide gel (BioRad) and imaged
with Gel Doc XR (BioRad). Gene disruption analysis was performed based on
densitometry of cleaved to uncleaved products.

3.3 Results
Here, our goal was to evaluate the best method to load nucleic acid cargo into exosomes,
either by overexpression of a zipcode-containing plasmid in producer cells, or by direct
transfection of pre-isolated exosomes. In both cases, exosomes were isolated using
37

ultracentrifugation and produced and delivered to the HEK 293T cell line. By comparing
various ways to deliver our nucleic acid cargo, we hope to select an efficient method of
loading our exosomes moving forward.
We first test whether the addition of a zipcode-like sequence reported to direct mRNA
transcripts into extracellular vesicles is generally applicable to our mRNA transcripts. Our
cargo plasmid encoding GFP was modified to include the zipcode-like sequence at the
3’UTR using restriction site cloning. Exosomal total protein content was unchanged when
they were produced with a plasmid encoding expression cassettes with or without the
3’UTR zipcode sequence. Flow cytometric analysis performed in cells treated with GFP-
packaged exosomes show no significant difference in GFP expression between the zipcode
and control group (Figure 8), suggesting that the presence of the zipcode-like sequence
did not improve mRNA packaging into exosomes, with the caveat that only one
independent study was performed. Additional replicates are needed to fully confirm the
functionality of the zipcode motif in enriching mRNA transcripts inside exosomes. In any
case, cells treated with exosomes directly transfected with plasmid DNA encoding GFP
showed a two-fold higher GFP expression that cells treated with the zipcode-containing
exosomes. These results show that direct transfection of exosomes may be the most suitable
method of cargo loading into exosomes.
38

While GFP cargo facilitates testing of the delivery capabilities of exosomes, it is important
to ensure that they can deliver functional nucleases. To confirm whether nuclease-loaded
exosomes effectively disrupted the endogenous CCR5 gene present in recipient 293T cells,
an enzymatic mutation detection assay was performed on recipient cells 3 days after
addition of ZFN-loaded exosomes packaged by direct transfection (Figure 9). As a control,
the transfected cells were collected and analyzed for nuclease activity. Our results showed
that the ZFN-loaded exosomes are able to achieve 9.3% disruption compared to 13.4% in
cells directly transfected with ZFN plasmids, allowing us to conclude that plasmid DNA
encoding ZFNs can be delivered by exosomes into recipient cells.
Figure 8. Comparison of cargo loading techniques
Two techniques to load plasmid GFP cargo into exosomes were examined. Passive transfection required
packaging pre-isolation, via transfection of producer cells with plasmid encoding GFP using
Lipofectamine (Thermo Fisher) and isolating exosomes through ultracentrifugation. Through this
method, a zipcode motif proposed to direct mRNA transcripts to extracellular vesicles was tested by
cloning the sequence into the same GFP plasmid. Direct transfection into exosomes required already
isolated exosomes which are then directly transfected using ExoFect (SBI). One independent study was
performed. Data shows that direct transfection into exosomes resulted in the highest GFP expression in
recipient 293Ts.
39

3.4 Discussion
Here, our objective was to optimize a method to load nucleic acid cargo into exosomes.
We tested whether the presence of a zipcode-like motif sequence in the cargo plasmids
resulted in increased packaging into exosomes. The zipcode sequence was reported to
result in a two-fold increase of mRNA transcripts into extracellular vesicles when the
plasmids containing the 3’UTR zipcode sequence are transfected into producer cells
(Bolukbasi, 2012). We performed a similar experiment using plasmids encoding GFP and
anti-CCR nucleases and compared the outcomes of the zipcode-containing plasmids to the
control group. In addition, we evaluate exosomes directly transfected with nucleic acid
cargo, as we have previously done in Chapter 2, as a potential method of loading exosomes.
Successful cargo packaging was measured by the resulting GFP fluorescence and nuclease-
induced mutations in the recipient cells, assuming that the exosome cargo is functional
Transfection
ZFN Exosomes
0
5
10
15
% Gene Disruption
Figure 9: Gene disruption of 293Ts at the
CCR5 locus
HEK293T cells were transfected with DNA
plasmids encoding a ZFN pair against CCR5
using Lipofectamine (Transfection bar). 293T
cells were treated with exosomes purified from
untransfected 293Ts by ultracentrifugation and
loaded with CCR5 ZFN plasmids using ExoFect
(ZFN Exosomes bar). Cells were then harvested
after 3 days and extracted for DNA and an
enzymatic mutation detection assay was
performed to obtain levels of gene disruption at
the CCR5 locus. One independent study was
performed. Results show that exosomes loaded
with ZFNs are able to deliver functional ZFNs to
recipient cells and cause gene disruption.
40

once inside the recipient cell. Translation of the mRNA transcripts is necessary to achieve
this outcome, and there are several factors, such as degrading enzymes or transport into
lysosomes, that may have prevented protein expression. It is also possible that the location
of the zipcode may have resulted in unknown consequence that prevented its function,
however the results from the control group did not suggest that this was the case.
We found that the most effective way to load exosomes is directly transfecting exosomes
with cargo nucleic acids using a cationic polymer reagent ExoFect. This method results in
60-70% GFP expression in target cells compared to 20-30% in the overexpressed
exosomes. Importantly, our exosomes are able to deliver functional ZFNs at 70% the
efficiency of cells directly transfected with ZFNs. Furthermore, our examination of the
applicability of the zipcode-like sequence to our plasmid cargo did not correspond to the
findings resulting in double the GFP expression (Bolukbasi, 2012). However, were able to
determine that for our purposes, direct transfection of pre-isolated exosomes is an effective
method of loading exosomes. This observation is in accordance to reports that exosome
transfer of nucleic acids and proteins are likely dependent on the specific cargo and cell
types used (Chevillet, 2014 and Kanada, 2015). Hence, further studies on the mechanism
of exosome uptake are needed to completely understand exosome-mediated delivery.
An alternative method to load Cas9 specifically is to use the commercially available Guide-
it CRISPR/Cas9 system (Clontech) that allows efficient delivery of a Cas9 protein with its
guide RNA inside microvesicles. This system, together with the iDimerize technology
(Clontech), by which a combination protein complex is generated by two dimerization
domains (DmrA and DmrC) and a dimerizing ligand (AP21967), can be used to redirect
Cas9 to the plasma membrane, through having one protein partner be a membrane protein.
41

The redirection of the protein to the plasma membrane allows enrichment of the protein
into extracellular vesicles, including microvesicles and exosomes. Upon exosome isolation
and delivery to target cells, absence of the dimerizing ligand allows for the dissociation of
the complex, thereby releasing the cargo protein. In short, this system can tether the Cas9
protein to exosome membranes.
We have identified an optimal method to effectively load exosomes with our desired
nucleic acids, and have shown this to work with plasmids encoding GFP and ZFNs. This
technique may broaden the use of exosomes as a delivery vector, adding yet another tool
for gene therapy with potential in vivo application.

42

Chapter 4 Targeting
4.1 Introduction
Engineered exosomes are novel tools in the field of in vivo delivery. Recent innovations in
exosome engineering achieve targeting capability through the display of targeting peptides
on exosomal membranes using membrane anchors (Alvarez-Erviti, 2011 and Ohno, 2013).
These membrane proteins, which include Lactadherin and Lamp2b, can be fused to desired
targeting peptides to anchor the targeting moiety on the surface of exosomes. At this stage,
we have demonstrated that our ultracentrifuge-produced exosomes have the capability to
transport functional GFP and genome editing nucleases to recipient cells, however our
main goal is to achieve targeted delivery to HIV reservoirs. In this study, we target latent
reservoirs using PD-1 as a surrogate marker, as PD-1 is enriched in latently infected cells
as previously described (Chomont, 2009). To provide targeting capability to exosomes, we
use PD-1 ligand (PD-L1) as the targeting peptide, which has previously been confirmed to
bind PD-1 (Lin, 2008). The general utility of this approach has been validated to target
EGFR-expressing cells and to the central nervous system, but engineered exosomes
targeting PD-1 have not been accomplished. Given the need to develop a delivery system
against latent reservoirs, we use Lactadherin and Lamp2b to anchor the PD-L1 to the
surface of exosomes in order to target PD-1 expressing cells.
Lactaderin, also known as its murine homolog milk fat globule epidermal growth factor 8
(MFG-E8), is a major glycoprotein of the human milk fat globule membrane. The mature
protein has three domains, an EGF-like domain with an RGD cell adhesion surface, and a
C1 and C2 domain (Oshima, 2002). The protein is co-translationally translocated into the
lumen of the endoplasmic reticulum, where it is recruited to intraluminal vesicles (ILVs)
43

or intercellular exosomes. The mechanism behind this is the protein’s 15-amino acid long
C-terminal C1C2 domain, which binds phosphatidylserine present on ILV membranes. The
C1C2 domain of lactadherin is commercially available as a vector plasmid (SBI) and has
been used as a technique to localize ligands to the exosome surface, including CD40, IL-
2, Her2, as well as to neurons and the gastrointestinal tract (Zeelenberg, 2008).
The alternate exosomal anchor we use in our targeting strategy is lysosomal-associated
membrane protein 2b (Lamp2b), which was used to produce the first reported engineered
exosomes (Alvarez-Erviti, 2011). The targeting peptide used was a 29-amino acid sequence
derived from the rabies viral glycoprotein (RVG) that targets the alpha-7 subunit of the
nicotinic acetylcholine receptor. Using labeled siRNA cargo, Alvarez-Erviti et al reported
that IV injected exosomes were found to deliver siRNA to neurons, microglia, and
oligodendrocytes to achieve a 60-70% knock down of the target gene in vivo. The target
gene BACE, or β-site amyloid precursor protein-cleaving enzyme 1, is a therapeutic target
in Alzheimer’s disease. This technique has since been applied to target integrin avb3 on
breast cancer cells (Tian, 2013).
To evaluate the targeting capabilities of our engineered exosomes, the final fusion plasmids
were transfected into exosome-producer cells, with the expectation that the Lactadherin
and Lamp2b proteins will be recruited to ILVs and will then act as exosome membrane
anchors for the PD-L1 targeting peptide once secreted as exosomes (Figure 10). Once the
presence of the targeting peptide on exosomes is confirmed, plasmid DNA encoding GFP
is packaged into exosomes using Exofect reagent. In the meantime, recipient 293T cells
are transfected with a PD-1 expression plasmid using Lipofectamine 2000 (Thermo Fisher
Scientific) to achieve an approximately 50% PD-1 expression at the time of exosome
44

treatment. We hypothesize that a majority of GFP expression will be found in the PD-1+
population, which will demonstrate the targeting capability of our engineered exosomes.

Figure 10 Schematic of targeting experimental design
To generate targeting plasmids, PDL1 was cloned into Lamp2b and Lactadherin constructs (top). A
general diagram of their expression in cells and exosomes is shown (middle). The targeting experimental
design is also shown (bottom). Producer 293T cells were transfected with PD-L1 fusion plasmids using
Lipofectamine (Step 1). Cell culture supernatant was collected after 48 hours and subjected to
ultracentrifugation to isolate exosomes (Step 2). Exosomes were either lysed for western blotting or
transfected with GFP plasmid using Exofect (Step 3) followed by addition to recipient PD-1 expressing
cells and flow cytometry to analyze GFP expression (Step 4).
45

4.2 Materials and Methods
4.2.1 Plasmids
Lamp2b construct used for engineering targeted exosomes have been described
previously and Dr. Wood, who pioneered this technology, has generously supplied
us with a Lamp2b expression plasmid that we use in this study (Alvarez-Erviti,
2011). The lactadherin construct is commercially available as the C1C2 domain in
a pMSCV-MCS-EF1-Puro lentivector (SBI). PD-L1 was obtained commercially
(Origene plasmid # RC213071) and the PD-1 expression plasmid is also
commercially available in a pCMV6-XL5 vector (Origene plasmid #SC117011).
4.2.2 In-Fusion Cloning of Fusion Plasmids
PDL1 was cloned into the Lactaderin and Lamp2b vectors using In-Fusion cloning
reagents (Clontech). To generate the PDL1-Lamp2b fusion construct, PDL1 was
cloned after the 39-amino acid signal peptide between linker proteins as previous
described (Alvarez-Erviti, 2011). For the PDL1-lactadherin fusion construct, PDL1
was cloned into the 5' secretion signal sequence (SS) within the multiple cloning
site and fused to the C1C2 domain of Lactadherin.
4.2.3 Engineering Targeted Exosomes
HEK 293T cells were transfected with Lipofectamine 2000 (Thermo Fisher)
according to manufacturer’s protocols and incubated for 48-72 hours. Cell culture
supernatant was collected and exosomes were isolated using ultracentrifugation.
Exosomes were then loaded with plasmid DNA encoding GFP using ExoFect
(SBI).
46

4.2.4 Western Blotting
Exosomes were lysed with RIPA buffer (Thermo Fisher) with PMSF protease
inhibitors (Sigma Aldrich). Protein concentration was determined using DC Protein
Assay reagents (Biorad) using a Berthold Mithras LB940 plate reader. Absorbance
was measured at 620nm in triplicate and analyzed using Mikrowin 2000 software.
30 g of exosome lysate and 50 g of cell lysate were then subjected to
electrophoresis using Criterion TGX gels (BioRad). Membranes were probed with
rabbit anti-PDL1 (ab205921) antibodies (Abcam) and incubated with goat anti-
rabbit IgG HRP antibody (sc-2004) secondary antibody (Santa Cruz
Biotechnology) and blocked with 5% milk and washed in PBST after each
incubation step. Western blots were visualized with Konica Minolta SRX-101A
Imager (Freedom Imaging).
4.2.5 Flow cytometry
Cells were collected and stained with Alexa Fluor 647 mouse anti-huCD279 (PD-
1) antibody (BD Pharmingen) 24 hours after exosome addition. Flow cytometry
was performed using Guava EasyCyte 6-21 (EMD Millipore). FlowJo software
version X (Treestar) was used for data analysis. In each independent analysis, the
criteria for gating of GFP positive and PD-1 positive cell populations was
determined for 0.1% or fewer of untreated or mock treated cell populations cultured
in parallel.
4.2.6 Statistical Analysis
Statistics were calculated using GraphPad Prism software. All data with error bars
are presented as mean plus or minus SEM. Statistical differences between two
47

parameters were determined using Student’s t-test, with P value <0.05 considered
significant.

4.3 Results
Our objective is to evaluate targeting strategies to deliver nucleic acid cargo inside
exosomes by exposing the PD-L1 targeting ligand using exosome membrane anchors. We
generated 2 different PD-L1-fusion proteins with the exosome membrane proteins
Lactadherin and Lamp2b and transfected the final fusion plasmids into 293T producer cells
48 hours prior to exosome isolation. Due to the enrichment of Lamp2b and Lactadherin to
exosomal membranes, we expect the fusion proteins to be located in membranes involved
with exosome biogenesis, including endosomes, multivesicular bodies, and intraluminal
vesicles, which are the intracellular form of exosomes. Targeting ligand incorporation was
confirmed by western blot analysis checking for PD-L1 (33kD) (Figure 11), demonstrating
that our targeting strategy effectively took advantage of the localization of Lactadherin and
Lamp2b on exosomes to express PD-L1 on the exosome surface.

26kDa
33kDa
PDL1
CD63
Figure 11 Characterization of
engineered exosomes
Engineered exosomes (30ug) were
examined for the presence of PDL1 and
CD63. Cellular lysates (CL) (50ug) were
also prepared and analyzed for the indicated
antibodies. Presence of PDL1 protein was
confirmed for cell lysates transfected with
targeted constructs, including the exosomes
derived from transfected cells.
Untransfected cell lysate control does not
show presence of PDL1 protein, as
expected.
48

Subsequently, the engineered exosomes were packaged with DNA plasmid encoding GFP
using the ExoFect transfection reagent (SBI). We then compared GFP expression of PD-1
transfected to mock transfected 293T cells treated with exosomes produced from an
equivalent number of starting cells to establish a baseline level of passive uptake in the
absence of the target receptor (Figure 12). With either construct, gene delivery was
observed in both PD-1
+
and PD-1
-
cells, with the former slightly more effective.
Surprisingly, we observed that untargeted exosomes were most effective overall, though
the ratio among PD-1
+
and PD-1
-
was consistent among all exosomes tested. However, it
is possible that transfection of the targeting molecules resulted in cell death or otherwise
reduced exosome yields compared to mock transfected cells. To control for effects of
transfection, we repeated the experiment with exosomes generated from cells transfected
with Lamp2b and Lactadherin alone, as well as their PD-L1 fusion counterparts.

Lamp2b-PDL1 Lactadherin-PDL1 Untargeted
PD-1
GFP
49

For the second experiment, exosomes were quantified by total protein content using a
Bradford assay, and 50 g of exosome particles were added to PD-1 transfected 293T cells.
As before, GFP expression was observed in both PD-1
+
and PD-1
-
target cells (Figure 13).
Lamp2b-PD-L1 exosomes showed no enhancement compared to Lamp2b alone, whereas
Lactadherin-PD-L1 showed elevated GFP expression compared to Lactadherin alone.
However, the ratio of GFP positivity in PD-1
+
and PD-1
-
target cells was comparable for
all constructs tested. Thus, we conclude that the constructs we generated do not facilitate
ligand-specific delivery of exosome cargo in 293T cells; rather, nonspecific mechanisms
of uptake appear to dominate in this scenario.

Figure 12 Delivery of Engineered exosomes to PD-1+ cells (I)
Flow cytometry analysis of recipient cells treated with targeted exosomes gated for PD-1 and GFP
expression (top). GFP expression from Q3 and Q4 was analyzed (bottom left) as well as the ratio of GFP+
cell in PD-1+ to PD-1- cells (bottom right). HEK293T cells were treated with untargeted, Lamp2b-PDL1,
and Lactadherin-PDL1 exosomes and collected 24 hours after exosomes addition. One independent study
of three replicates was performed.
50

Lactadherin-PDL1
No exosomes
Lamp2b Lamp2b-PDL1 Lactadherin
PD-1
GFP
Figure 13 Delivery of Engineered exosomes to PD-1+ cells (II)
Flow cytometry analysis of recipient cells treated with targeted exosomes gated for PD-1 and GFP expression
(top). GFP expression from Q3 and Q4 was analyzed (bottom left) as well as the ratio of GFP+ cell in PD-1+
to PD-1- cells (bottom right) was repeated to include Lamp2b and Lactadherin controls. HEK293T cells were
treated with untargeted, Lamp2b only, Lamp2b-PDL1, Lactadherin only, and Lactadherin-PDL1 exosomes
and collected 24 hours after exosomes addition. One independent study of three replicates was performed.
51

4.4 Discussion
Recently, exosomes have been the focus of targeted delivery because of their clinical
applicability as a vaccine and drug delivery tool (Kooijmans, 2012). Here, we test recent
innovations in exosome engineering that imbues targeting ability to exosomes using
peptide display technology. In our study, we generated chimeric proteins with PD-L1
ligand that binds to PD-1 to an exosomal membrane protein partner. Together, these
experiments examined the targeting capability and specificity of our engineered exosomes
to PD-1 expressing cells.
In our results, we observed that engineered exosomes do not exhibit a significant difference
in efficiency of GFP delivery than untargeted exosomes. In addition, PD-1
+
cells exhibited
a higher uptake of exosomes than PD-1
-
cells regardless of the presence of targeting ligand.
This experiment should be repeated in a stable cells line expressing PD-1, as changes in
cell behavior could be due to transfection and our surprising observation may be an artifact
of the lipofection process. Additionally, testing the targeted exosomes in K562s is another
viable experiment to plan, as a different recipient cell line may introduce a gain of function.
Taken together, our results suggest that the majority of our engineered exosomes did not
deliver to recipient cells in a receptor-mediated manner, but should be further confirmed
in additional testing.
Besides PD-1, other immune checkpoint molecules such as T cell immunoreceptor with Ig
and ITIM domains (TIGIT) and lymphocyte activation gene 3 (LAG-3) have also been
found positively associated with T cells from HIV-infected patients on ART and can serve
as surrogate markers of latent HIV (Fromentin, 2016). Currently, the strong association of
latently infected cells with the expression of TIGIT and LAG-3 has not been validated in
52

an animal model. For this reason, we chose to target PD-1, as our lab has established a PD-
1 mouse model of latency. It is possible that our receptor-ligand of choice were not optimal.
PD-1 may not been more efficiently targeted with PD-L2, a second ligand that has been
found to bind with higher affinity to PD-1 (Butte, 2008). In addition, testing a panel of
alternate surrogate markers for latency such as LAG-3 and TIGIT holds potential in
developing a targeting strategy to direct engineered exosomes to latently infected cells.
To improve our understanding of how engineered exosomes enter cells, further
examination of the mechanism of exosome uptake into recipient cells is needed. Generally,
once the exosome reaches the target cell, receptor-mediated endocytosis can follow as a
response to internalize receptor ligands. Three routes of receptor-mediated endocytosis are
known: clathrin-mediated endocytosis, caveolae-mediated endocytosis, and a clathrin and
caveolae independent pathway. In order to determine the mechanism of uptake, chemical
inhibitor assays to block specific endocytic pathways can be performed to elucidate which
pathway contributes to exosome entry into cells. Dynamin and Fumonisin-1 are chemical
inhibitors known to block clathrin-mediated and caveolae-mediated endocytosis
specifically in 293Ts, respectively (Mulcahy, 2014). The use of known chemical inhibitors
may elucidate the specific endocytic pathway responsible for the intracellular trafficking
of targeted exosomes into recipient cells, or conversely, determine whether the majority of
uptake is not receptor-mediated.
Another approach to generate engineered exosomes using the transmembrane domain of
platelet-derived growth factor (PDGF) (Ohno, 2013). PDGF has been previously used as a
fusion protein with the GE11 peptide to target exosomes to EGFR-expressing breast cancer
cells. While Lamp-2b and Lactadherin are membrane proteins found in exosomes, PDGF
53

is localized to the cell surface. Since exosome formation is initiated by a budding
endosome, the exosome membrane will also include plasma membrane proteins. Thus, this
alternate approach is another arsenal of tools using peptide display.

54

Summary
Although current antiretroviral drug treatment (ART) for HIV is highly effective at suppressing
viral replication, the drugs do not have the ability to target latent HIV reservoirs. Strategies to
suppress the reactivation of latent HIV reservoir is needed to obtain a drug-free control of HIV
infection. Current antiretroviral drug therapy targets only actively replicating virus and does not
completely eradicate all virus from the body. Nonreplicating proviral genomes are able to persist
in HIV-infected cells during even prolonged antiretroviral drug therapy and are capable of
reactivation if treatment is stopped. Thus, the key to a comprehensive cure for HIV is to remove
the proviral HIV genome inside latently-infected cells.
In this study, we use genome-editing nucleases, such as zinc-finger nucleases (ZFNs), transcription
activator-like effector nucleases (TALENs), and the CRISPR/Cas9 system, as tools to disrupt the
integrated HIV genome that persists in latently infected cells to achieve targeted reduction of the
HIV reservoir. Our lab is currently developing genome editing nucleases targeting the HIV
genome. However, the ability to deliver nucleic acids to a specific cell type in vivo is a major
challenge in the field of gene therapy. Systemic delivery of naked nucleic acids in the blood is
ineffective, as they are quickly degraded by nucleases. In addition, the negative charge on DNA
and RNA limits cellular uptake by cells that also have a negatively charged surface. Taken
together, this means that the nucleases require delivery vehicles to be effective in vivo. Our
approach to overcome these challenges is by using exosomes as a novel delivery system.
Exosomes are membraned vesicles naturally capable of transporting nucleic acids and proteins
between cells. To deliver their cargo to target cells, exosomes have a variety of surface adhesion
proteins as well as specific vector ligands that reflect the physiological or pathological state of the
cells they originate from. Conversely, exosome uptake can occur through a variety of mechanism
55

including phagocytosis, micropinocytosis, receptor-mediated endocytosis, depending on the type
of recipient cell (Mulcahy, 2014). Importantly, modified exosomes have been shown to
successfully deliver functional nucleic acid and drugs to specific cells in vivo. Previous research
on targeted exosomes have shown in vivo delivery of functional siRNA across the blood-brain
barrier to the microglia, astrocytes, and other cells of the central nervous system (Alvarez-Erviti,
2011). Targeted exosomes have also been developed to deliver miRNA to EGFR-expressing tumor
mouse model (Ohno, 2013). Given this, exosomes are promising candidates for in vivo delivery of
anti-HIV nucleases to HIV-target cells.
Here, we optimized in-house exosome production using ultracentrifugation to isolate and load
cargo into functional exosomes. We established exosome functionality using our exosomes
carrying GFP DNA and mRNA to recipient 293T cells and the result is comparable to GFP
expression levels achieved with commercial 293-derived exosomes. However, K562 cells treated
with the same GFP-loaded exosomes did not express GFP, suggesting a cell-type dependent uptake
of exosomes. Exosomes can be engineered to express a specific ligand by fusion to an exosomal
membrane protein (Delcayre, 2005). To exploit exosomes as a delivery tool to target specific
tissues or cells, however, it is imperative to identify appropriate strategies to control their targeting.
The target cells in this study express the PD-1 receptor enriched in latently infected cells. We
created targeted exosomes by incorporating targeting ligand PD-L1, which is capable of binding
to these target cells, by fusion to exosomal membrane proteins Lactadherin or Lamp2b. This
technique has been previously used to direct exosomes to various immune cells as well as the
central nervous system, but has not been confirmed specifically for PD-1 expressing cells. These
engineered exosomes were then used to deliver plasmid DNA encoding GFP to PD-1
-
or PD-1
+

recipient 293T cells. Surprisingly, we observed that PD-1
+
cells exhibited a higher uptake of
56

exosomes than PD-1
-
cells. However, we also observed that PD-1 targeted exosomes do not deliver
preferentially to PD-1
+
cells. This finding should be confirmed on stable cell lines expressing PD-
1 to void undetermined effects of transfection. We interpreted these results to mean that the
majority of our engineered exosomes did not deliver to recipient cells in a receptor-mediated
manner. Taken together, our modified exosomes engineered to target PD-expressing cells did not
sufficiently operate as targetable delivery vehicles as hypothesized, but more comprehensive
studies are required to confirm or dismiss the use of exosomes in targeted gene therapy.

57

References
1. Fauci, A. & Folkers, G. Toward an AIDS-Free Generation. Jama 308, 343–344 (2012).

2. Fletcher, C. et al. Persistent HIV-1 replication is associated with lower antiretroviral drug
concentrations in lymphatic tissues. Proc Natl Acad Sci 111, 2307–2312 (2014).

3. Marin, B. et al. Non-AIDS-defining deaths and immunodeficiency in the era of
combination antiretroviral therapy. Aids 23, 1743 (2009).

4. Schackman, B. et al. The Lifetime Cost of Current Human Immunodeficiency Virus Care
in the United States. Med Care 44, 990 (2006).

5. Deeks, S. HIV: Shock and kill. Nature 487, 439–440 (2012).

6. Ho, Y.-C. et al. Replication-Competent Noninduced Proviruses in the Latent Reservoir
Increase Barrier to HIV-1 Cure. Cell 155, 540–51 (2013).

7. Shan, L. et al. Stimulation of HIV-1-Specific Cytolytic T Lymphocytes Facilitates
Elimination of Latent Viral Reservoir after Virus Reactivation. Immunity 36, 491–501
(2011).

8. Rasmussen, T. & Lewin, S. Shocking HIV out of hiding: where are we with clinical trials
of latency reversing agents? Curr Opin Hiv Aids 11, 394 (2016).

9. Kimata, J., Rice, A. & Wang, J. Challenges and strategies for the eradication of the HIV
reservoir. Curr Opin Immunol 42, 65–70 (2016).

10. Cary, D. & Peterlin, M. Targeting the latent reservoir to achieve functional HIV cure.
F1000research 5, 1009 (2016).

11. Karpinski, J. et al. Directed evolution of a recombinase that excises the provirus of most
HIV-1 primary isolates with high specificity. Nat Biotechnol 34, 401–409 (2016).

12. Massanella, M. & Richman, D. Measuring the latent reservoir in vivo. J Clin
Invest 126, 464–472 (2016).

13. Chun, Chadwick, Margolick & Siliciano. Differential susceptibility of naive and memory
CD4+ T cells to the cytopathic effects of infection with human immunodeficiency virus
type 1 strain LAI. J Virol 71, 4436–44 (1997).

14. Fletcher, C. et al. Persistent HIV-1 replication is associated with lower antiretroviral drug
concentrations in lymphatic tissues. Proc Natl Acad Sci 111,2307–2312 (2014).

15. Ruelas, D. & Greene, W. An Integrated Overview of HIV-1 Latency. Cell 155, 519–529
(2013).
58

16. Kinter, A. et al. The Common γ-Chain Cytokines IL-2, IL-7, IL-15, and IL-21 Induce the
Expression of Programmed Death-1 and Its Ligands. J Immunol 181, 6738–6746 (2008).

17. Chomont, N. et al. HIV reservoir size and persistence are driven by T cell survival and
homeostatic proliferation. Nat Med 15, 893–900 (2009).

18. Gaj, T., Gersbach, C. A. & Barbas, C. F. ZFN, TALEN, and CRISPR/Cas-based methods
for genome engineering. Trends Biotechnol. 31, 397–405 (2013).

19. Wyman, C. & Kanaar, R. DNA Double-Strand Break Repair: All’s Well that Ends
Well.Genetics 40, 363–383 (2006).

20. Ran, A. et al. In vivo genome editing using Staphylococcus aureus Cas9. Nature 520, 186–
191 (2015).

21. Sarkar, Hauber, Hauber & Buchholz. HIV-1 Proviral DNA Excision Using an Evolved
Recombinase. Science 316, 1912–1915 (2007).

22. Hauber, I. et al. Highly Significant Antiviral Activity of HIV-1 LTR-Specific Tre-
Recombinase in Humanized Mice. Plos Pathog 9, e1003587 (2013).

23. Qu, X. et al. Zinc-finger-nucleases mediate specific and efficient excision of HIV-1
proviral DNA from infected and latently infected human T cells.Nucleic Acids
Res 41, 7771–7782 (2013).

24. Ebina, H., Misawa, N., Kanemura, Y. & Koyanagi, Y. Harnessing the CRISPR/Cas9
system to disrupt latent HIV-1 provirus. Sci Reports 3, 2510 (2013).

25. Björndal et al. Coreceptor usage of primary human immunodeficiency virus type 1 isolates
varies according to biological phenotype. J Virol 71,7478–87 (1997).

26. Samson, M. et al. Resistance to HIV-1 infection in Caucasian individuals bearing mutant
alleles of the CCR-5 chemokine receptor gene. Nature 382, 722–725 (1996).

27. Holt, N. et al. Human hematopoietic stem/progenitor cells modified by zinc-finger
nucleases targeted to CCR5 control HIV-1 in vivo. Nat Biotechnol 28, 839–847 (2010)

28. Wang, J. et al. Homology-driven genome editing in hematopoietic stem and progenitor
cells using ZFN mRNA and AAV6 donors.Nat Biotechnol 33, 1256–1263 (2015).

29. Sakuma, T., Barry, M. & Ikeda, Y. Lentiviral vectors: basic to translational. Biochem
J 443, 603–618 (2012).

30. Mingozzi, F. & High, K. Immune responses to AAV in clinical trials. Curr Gene
Ther 11,321–30 (2011).
59

31. Duggan, S. & Keating, G. Pegylated Liposomal Doxorubicin. Drugs 71, 2531–2558
(2011).

32. Tan, A., Rajadas, J. & Seifalian, A. Exosomes as nano-theranostic delivery platforms for
gene therapy. Adv Drug Deliver Rev 65, 357–367 (2013).

33. Witwer, K. et al. Standardization of sample collection, isolation and analysis methods in
extracellular vesicle research. J Extracell Vesicles 2, (2013).

34. Simpson, R., Lim, J., Moritz, R. & Mathivanan, S. Exosomes: proteomic insights and
diagnostic potential. Expert Rev Proteomic 6, 267–83 (2009).

35. Andaloussi, S., Mäger, I., Breakefield, X. & Wood, M. Extracellular vesicles: biology and
emerging therapeutic opportunities. Nat Rev Drug Discov 12, 347–357 (2013).

36. Théry, C., Ostrowski, M. & Segura, E. Membrane vesicles as conveyors of immune
responses.Nat Rev Immunol 9, 581–593 (2009).

37. Batrakova, E. & Kim, M. Using exosomes, naturally-equipped nanocarriers, for drug
delivery. J Control Release 219, 396–405 (2015).

38. Johnsen, K. et al. A comprehensive overview of exosomes as drug delivery vehicles —
Endogenous nanocarriers for targeted cancer therapy. Biochimica Et Biophysica Acta Bba
- Rev Cancer 1846, 75–87 (2014).

39. Haney, M. et al. Exosomes as drug delivery vehicles for Parkinson’s disease therapy. J
Control Release 207, 18–30 (2015)

40. Bell, B., Kirk, I., Hiltbrunner, S., Gabrielsson, S. & Bultema, J. Designer exosomes as next-
generation cancer immunotherapy. Nanomedicine Nanotechnol Biology Medicine 12,163–
169 (2016).

41. Kooijmans, S. et al. Display of GPI-anchored anti-EGFR nanobodies on extracellular
vesicles promotes tumour cell targeting. J Extracell Vesicles 5,31053 (2016).

42. Delcayre, A. & Pecq, J.-B. Exosomes as novel therapeutic nanodevices. Curr Opin Mol
Ther8, 31–8 (2006).

43. Mulcahy, L., Pink, R. & Carter, D. Routes and mechanisms of extracellular vesicle uptake.
J Extracell Vesicles 3, (2014).

44. Rountree, R. et al. Exosome Targeting of Tumor Antigens Expressed by Cancer Vaccines
Can Improve Antigen Immunogenicity and Therapeutic Efficacy. Cancer Res 71, 5235–
5244 (2011).

60

45. Alvarez-Erviti, L. et al. Delivery of siRNA to the mouse brain by systemic injection of
targeted exosomes. Nat Biotechnol 29,341–345 (2011).

46. Ohno, S. et al. Systemically injected exosomes targeted to EGFR deliver antitumor
microRNA to breast cancer cells.Mol Ther J Am Soc Gene Ther 21,185–91 (2012).

47. Tauro, B. et al. Comparison of ultracentrifugation, density gradient separation, and
immunoaffinity capture methods for isolating human colon cancer cell line LIM1863-
derived exosomes. Methods 56, 293–304 (2012).

48. Zeringer, E., Barta, T., Li, M. & Vlassov, A. Strategies for Isolation of Exosomes. Cold
Spring Harb Protoc 2015,pdb.top074476 (2015).

49. Livshts, M. et al. Isolation of exosomes by differential centrifugation: Theoretical analysis
of a commonly used protocol. Sci Reports 5, 17319 (2015).

50. Théry, C., Amigorena, S., Raposo, G. & Clayton, A. Isolation and characterization of
exosomes from cell culture supernatants and biological fluids. Curr Protoc Cell Biology
Chapter 3, Unit 3.22 (2006).

51. Heinemann, M. et al. Benchtop isolation and characterization of functional exosomes by
sequential filtration. J Chromatogr A 1371, 125–135 (2014).

52. Rao, S., Gefroh, E. & Kaltenbrunner, O. Recovery modeling of tangential flow
systems. Biotechnol Bioeng 109,3084–3092 (2012).

53. Cooper, A. et al. Highly efficient large-scale lentiviral vector concentration by tandem
tangential flow filtration. J Virol Methods 177, 1–9 (2011).

54. Raposo, G. & Stoorvogel, W. Extracellular vesicles: Exosomes, microvesicles, and
friends. J Cell Biology 200, 373–383 (2013).

55. Lee, Y., Andaloussi, S. & Wood, M. Exosomes and microvesicles: extracellular vesicles
for genetic information transfer and gene therapy. Hum Mol Genet 21, R125–R134 (2012).

56. Lobb, R. et al. Optimized exosome isolation protocol for cell culture supernatant and
human plasma. J Extracell Vesicles4, 27031 (2015).

57. Sokolova, V. et al. Characterisation of exosomes derived from human cells by nanoparticle
tracking analysis and scanning electron microscopy. Colloids Surfaces B
Biointerfaces 87, 146–50 (2011).

58. Zöller, M. Tetraspanins: push and pull in suppressing and promoting metastasis. Nature
Reviews Cancer 9, 40–55 (2008).

61

59. Jeppesen, D. et al. Comparative analysis of discrete exosome fractions obtained by
differential centrifugation. J Extracell Vesicles 3, 25011 (2014).

60. Lötvall, J. et al. Minimal experimental requirements for definition of extracellular vesicles
and their functions: a position statement from the International Society for Extracellular
Vesicles. J Extracell Vesicles 3, 26913 (2014).

61. Xu, R., Greening, D., Rai, A., Ji, H. & Simpson, R. Highly-purified exosomes and shed
microvesicles isolated from the human colon cancer cell line LIM1863 by sequential
centrifugal ultrafiltration are biochemically and functionally distinct.Methods 87, 11–25
(2015).

62. Peterson, M., Otoc, N., Sethi, J., Gupta, A. & Antes, T. Integrated systems for exosome
investigation. Methods 87, 31–45 (2015).

63. Mineo, M. et al. Exosomes released by K562 chronic myeloid leukemia cells promote
angiogenesis in a src-dependent fashion. Nato Adv Sci Inst Se 15,33–45 (2012).

64. Tadokoro, H., Umezu, T., Ohyashiki, K., Hirano, T. & Ohyashiki, J. Exosomes Derived
from Hypoxic Leukemia Cells Enhance Tube Formation in Endothelial Cells. J Biol
Chem288, 34343–34351 (2013).

65. Quah, B. & O’Neill, H. The immunogenicity of dendritic cell-derived exosomes. Blood
Cells Mol Dis 35, 94–110 (2005).

66. Valadi, H. et al. Exosome-mediated transfer of mRNAs and microRNAs is a novel
mechanism of genetic exchange between cells. Nature Cell Biology 9, 654–659 (2007).

67. Taylor, D. & Gercel-Taylor, C. The origin, function, and diagnostic potential of RNA
within extracellular vesicles present in human biological fluids. Front Genetics 4, 142
(2013).

68. Ekström, K. et al. Characterization of mRNA and microRNA in human mast cell-derived
exosomes and their transfer to other mast cells and blood CD34 progenitor cells. J Extracell
Vesicles 1, (2012).

69. Tian, Y. et al. A doxorubicin delivery platform using engineered natural membrane vesicle
exosomes for targeted tumor therapy. Biomaterials 35,2383–2390 (2014).

70. Batagov, A. & Kurochkin, I. Exosomes secreted by human cells transport largely mRNA
fragments that are enriched in the 3′-untranslated regions. Biol Direct 8, 12 (2013).

71. Bolukbasi, M. et al. miR-1289 and ‘Zipcode’-like Sequence Enrich mRNAs in
Microvesicles. Mol Ther Nucleic Acids 1, e10 (2012).

62

72. Martin, K. & Ephrussi, A. mRNA Localization: Gene Expression in the Spatial
Dimension. Cell 136, 719–30 (2009).

73. Perez, E. et al. Establishment of HIV-1 resistance in CD4+ T cells by genome editing using
zinc-finger nucleases. Nat Biotechnol 26, 808–816 (2008).

74. Chevillet, J. et al. Quantitative and stoichiometric analysis of the microRNA content of
exosomes. Proc Natl Acad Sci 111, 14888–14893 (2014).

75. Kanada, M. et al. Differential fates of biomolecules delivered to target cells via
extracellular vesicles. Proc Natl Acad Sci 112,E1433–E1442 (2015).

76. Lin, D. et al. The PD-1/PD-L1 complex resembles the antigen-binding Fv domains of
antibodies and T cell receptors.Proc Natl Acad Sci 105, 3011–3016 (2008).

77. Oshima, K., Aoki, N., Kato, T., Kitajima, K. & Matsuda, T. Secretion of a peripheral
membrane protein, MFG‐E8, as a complex with membrane vesicles. Eur J
Biochem 269, 1209–1218 (2002).

78. Zeelenberg, I. et al. Targeting Tumor Antigens to Secreted Membrane Vesicles In vivo
Induces Efficient Antitumor Immune Responses. Cancer Res 68, 1228–1235 (2008).

79. Kooijmans, S., Vader, P., van Dommelen, S., van Solinge, W. & Schiffelers, R. Exosome
mimetics: a novel class of drug delivery systems. Int J Nanomed7, 1525–1541 (2012).

80. Fromentin, R. et al. CD4+ T Cells Expressing PD-1, TIGIT and LAG-3 Contribute to HIV
Persistence during ART. Plos Pathog 12, e1005761 (2016).

Abstract (if available)

Abstract Antiretroviral therapy is extremely effective against human immunodeficiency virus (HIV), but strategies to suppress the reactivation of the latent HIV reservoir are needed to obtain a drug-free control of HIV infection. Genome-editing nucleases are a promising approach, as they are effective at disrupting the latent proviral genome in vitro. The main hurdle, however, is ensuring delivery of the nucleases in a cell- or tissue-specific manner. Exosomes are biogenic nanovesicles capable of intercellular transport of nucleic acids and protein and have been engineered to target specific cells in vivo. As such, in this thesis, we examine the hypothesis that engineered exosomes are capable of delivering genome editing nucleases to latently infected cells in a targeted manner, using programmed death 1 (PD-1) receptor as a surrogate marker of HIV latency. First, we optimized a method to produce functional exosomes from the human embryonic kidney (HEK) 293T cell line using ultracentrifugation. Second, we compared methods of packaging our nuclease cargo into exosomes and determined that direct transfection into exosomes achieved the most efficient packaging of cargo into exosomes. For ease of testing, we use plasmid DNA encoding GFP as cargo. Finally, we engineered exosomes with targeting capability using PD-1 ligand 1 (PD-L1) that binds PD-1, fused to an exosomal membrane protein, either Lactadherin or Lamp2b. Our evaluation of the delivery of these engineered exosomes carrying GFP to PD-1 expressing cells led to the conclusion that although our untargeted exosomes achieved highly efficient gene delivery, further work is required to enhance their targeted delivery. Ultimately, our study evaluates a possible strategy for eradicating the latent HIV genome using an exosome-based delivery system. However, additional studies on targeting capability need to be performed and optimized in order to validate whether exosomes carrying genome-editing nucleases are capable of targeted delivery in vivo.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Pseudotyped viral vectors: HIV gene therapy applications and basic studies of SARS-COV-2

PDF

Improving adeno-associated viral vector for hematopoietic stem cells gene therapy

PDF

Molecular characterization of the HIV-1 Vpu protein and its role in antagonizing the cellular restriction factor BST-2/tetherin both in vitro and in vivo

PDF

Characterizing and manipulating homology-directed gene editing in human cells

PDF

Design and characterization of multiplex anti-HIV single domain antibodies for genome editing of the immunoglobulin locus

PDF

Genomic stability of transcriptionally targeted replication competent retroviral vectors

PDF

Rational selection of CRISPR/Cas9 guide RNAs for homology directed genome editing and its utility in the development of gene therapies

PDF

PRAS40 connects microenvironmental stress signaling to exosome-mediated secretion

PDF

Engineering viral vectors for T-cell immunotherapy and HIV-1 vaccine

PDF

Engineering B cells to express fully customizable therapeutic antibodies with enhanced Fc functions

PDF

Developing novel in vitro model systems to investigate therapeutic hypotheses for peroxisome biogenesis disorders

PDF

Placenta growth factor-miRNAs-lncRNAs axis in the regulation of ET-1 gene involved in pulmonary hypertension in sickle cell disease

PDF

Structural and biochemical analyses on substrate specificity and HIV-1 Vif mediated inhibition of human APOBEC3 cytidine deaminases

PDF

Hepatic differentiation in human naïve stem cells compared to human embryonic stem cells

PDF

Mapping arenavirus glycoprotein determinants that correlate to cell entry and virus pathogenicity

PDF

Behavioral choice assays and alcohol preference in Drosophila melanogaster

PDF

Disrupted ER-to-Golgi trafficking underlies anti-HIV drugs and alcohol induced cellular stress and hepatic injury

PDF

TLR8-transferred miR-192 acts as a tumor suppressor in neuroblastoma by inhibiting CTCF

PDF

Non-viral and viral hematopoietic progenitor cell gene therapy

PDF

Genetics and the environment: evaluating the role of noncoding RNA in autism spectrum disorder

Asset Metadata

Creator Gadon, Andrea Liza G. (author)

Core Title Using engineered exosomes and gene-editing to target latent HIV

School Keck School of Medicine

Degree Master of Science

Degree Program Biochemistry and Molecular Biology

Publication Date 11/29/2016

Defense Date 10/11/2016

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

Tag exosomes,gene editing,HIV,OAI-PMH Harvest,PD-1

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Hacia, Joseph (committee chair), Cannon, Paula (committee member), DePaolo, William (committee member)

Creator Email agadon@usc.edu,agadon10@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-324034

Unique identifier UC11214663

Identifier etd-GadonAndre-4947.pdf (filename),usctheses-c40-324034 (legacy record id)

Legacy Identifier etd-GadonAndre-4947.pdf

Dmrecord 324034

Document Type Thesis

Format application/pdf (imt)

Rights Gadon, Andrea Liza G.

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