Download by MICHAEL WING-YIN WU

ANTI-GD3 ANTIBODIES ARE TARGETING MOLECULES FOR DELIVERY OF
siRNA TO MELANOMA
by
MICHAEL WING-YIN WU
A thesis submitted to the Department of Pathology and Molecular Medicine
in conformity with the requirements for the degree of
Master of Science
Queen’s University
Kingston, Ontario, Canada
September, 2010
Copyright © Michael Wing-Yin Wu, 2010
ABSTRACT
Melanoma is the most deadly form of skin cancers, with an incidence increasing more rapidly
than any other malignant cancer in the past 40 years. Metastatic melanoma is resistant to
conventional treatments, such as chemotherapy and radiation therapy. Our lab has previously
demonstrated that Mcl-1 is a key contributor in protecting melanoma from therapy-induced cell
death. RNAi therapeutics was employed as a novel way to silence the anti-apoptotic protein by
using Mcl-1 mRNA sequence-specific siRNAs in vitro. In our hands, passive non-targeted
delivery of RNAi therapy into melanoma tumours has been shown to be neither effective, nor
selective in vitro and in vivo. Consequently, in this study, siRNA was linked to a delivery system
which expressed a ligand specifically targeting melanoma cells. Previously shown, melanoma
overexpresses the cell surface ganglioside GD3, thus it is my belief that the anti-GD3 R24
monoclonal antibody can function as a targeting molecule. The antibody was linked to coated
cationic liposomes (CCLs) carrying siRNA molecules. Our first step was to confirm R24 ligation
to CCLs. Untargeted CCLs showed insignificant values of antibody, whereas antibodyconjugated CCLs presented approximately 30 antibodies per liposome. I also confirmed that
siRNA was internalized within CCLs using spectrometry, with an encapsulation efficiency of
approximately 80%. Since liposomes need to be small to be effective in vitro and in vivo, CCLs
were confirmed to be less than 100nm in diameter. In vitro studies using fluorescent microscopy
demonstrated greater binding to melanoma cells with antibody-conjugated CCLs as compared to
untargeted CCLs. In vivo experiments showed specific localization of targeted CCLs to induced
subcutaneous mouse xenograft tumours. Western blotting demonstrated greater Mcl-1
knockdown using GD3-targeted CCLs. Taken together, these results suggest that anti-GD3
ii antibodies can serve as targeting molecules to deliver siRNA to melanoma cells and furthermore,
GD3-targeted CCLs can promote siRNA-mediated gene silencing.
iii ACKNOWLEDGEMENTS
First and foremost, I would like to thank my supervisor, Dr. Victor Tron, for giving me the
opportunity to undertake this research and degree. The guidance, support, and positivity he has
provided has made my time in his lab one of my greatest learning experience to date.
Thanks to my supervisory committee members, Dr. Harriet Feilotter, Dr. Brian Amsden, and Dr.
Xiaolong Yang, for their time, laboratory resources, and continued interest in my research
project.
Thank you past and present lab members and all lab technicians that have contributed to this
work. Thank you for the time, patience, moral support and expertise you have all shared with me.
A special thanks to my lab mates, Cindy Lentz, Geneviève Paré, Nick Zhang, Jiamin Chen and
Evan Rees; your thought-provoking questions, advice, instruction and technical assistance have
been invaluable and without your support, this research would not have advanced to this extent.
On a personal note, thank you to my friends and family for their emotional support,
encouragement, and faith in my abilities. Finally, I thank God for blessing me with this
incredible opportunity and experience to serve You in the study of science.
Funding for this project was provided by CIHR grant #394 780.
iv TABLE OF CONTENTS
ABSTRACT.................................................................................................................................... ii
ACKNOWLEDGEMENTS............................................................................................................ iv
LIST OF FIGURES...................................................................................................................... viii
LIST OF TABLES........................................................................................................................... x
ABBREVIATIONS........................................................................................................................ xi
CHAPTER 1 – INTRODUCTION.................................................................................................. 1
1.1.
Melanoma..................................................................................................................... 1
1.1.1.
Epidemiology of Melanoma................................................................................ 1
1.1.2.
Origin and Etiology of Melanoma....................................................................... 1
1.1.3.
Melanoma and Apoptosis.................................................................................... 2
1.1.4.
Progression of Melanoma.................................................................................... 3
1.1.5.
Treatments and Challenges of Melanoma........................................................... 5
1.2.
RNA Interference & Small Interfering RNA................................................................ 7
1.2.1.
The Mechanism of RNA Interference (small interfering RNA)..........................7
1.2.2.
siRNA Therapies................................................................................................. 9
1.2.3.
Challenges to siRNA Therapies........................................................................ 10
1.2.3.1. Specificity................................................................................................... 10
1.2.3.2. Potency: advantage of using Dicer-substrate siRNA (DsiRNA)................11
1.2.3.3. Stability...................................................................................................... 12
1.2.3.4. Size and Charge.......................................................................................... 13
1.2.3.5. Delivery...................................................................................................... 14
1.3.
Drug Delivery Systems............................................................................................... 14
1.3.1.
Introduction....................................................................................................... 14
1.3.2.
The Enhanced Permeability and Retention Effect.............................................15
1.3.3.
DDSs for siRNA Delivery: Non-Covalent Drug Complexes............................ 17
1.3.4.
DDSs for siRNA Delivery: Covalent Drug Complexes.................................... 19
1.3.5.
DDSs for siRNA Delivery: Liposomes............................................................. 21
1.3.6.
DDSs for siRNA Delivery: Challenges to Liposomal Therapies...................... 23
1.3.7.
DDSs for siRNA Delivery: Targeted Liposomal Therapies..............................25
1.3.8.
DDSs for siRNA Delivery: Potential Targeting Molecules for Melanoma.......27
1.4.
Summary and Goals.................................................................................................... 27
CHAPTER 2 – MATERIALS AND METHODS......................................................................... 29
2.1.
Preparation of (Untargeted and Targeted) Empty Coated Cationic Liposomes
(CCLs).............................................................................................................................. 29
2.1.1.
Chemicals, Reagents, and Materials.................................................................. 29
2.1.2.
Preparation of the anti-GD3 R24 antibody........................................................ 30
2.1.3.
Empty CCL Preparation Protocol...................................................................... 30
v 2.2.
Preparation of (Untargeted and Targeted) siRNA CCLs............................................ 31
2.2.1.
Chemical, Reagents, and Materials................................................................... 31
2.2.2.
Preparation of the anti-GD3 R24 Antibody.......................................................32
2.2.3.
Sequence and Modifications of siRNA............................................................. 32
2.2.4.
siRNA CCL Preparation Protocol..................................................................... 32
2.3.
Preparation of (Untargeted and Targeted) siRNA CCLs with Micelles..................... 33
2.3.1.
Chemicals, Reagents, and Materials.................................................................. 33
2.3.2.
Preparation of the anti-GD3 R24 Antibody.......................................................33
2.3.3.
Sequence and Modifications of siRNA............................................................. 35
2.3.4.
siRNA CCL Preparation Protocol with Micelles.............................................. 35
2.4.
Preparation of (Untargeted and Targeted) siRNA CCLs with Micelles, fused at
37°C.................................................................................................................................. 36
2.5.
Analysis of CCLs: Spectrometry and RNA Quantitation........................................... 36
2.6.
Analysis of CCLs: Denaturing Urea Polyacrylamide Gel.......................................... 36
2.7.
Analysis of CCLs: Dynamic Light Scattering............................................................ 37
2.8.
Analysis of CCLs: NanoDrop and Antibody Quantitation......................................... 38
2.9.
Analysis of CCLs: CBQCA Assay............................................................................. 38
2.10. Analysis of CCLs: Concentrating CCLs..................................................................... 39
2.11. Melanoma Cell Lines and Culture Conditions............................................................39
2.12. In Vitro Mcl-1 Silencing by CCLs.............................................................................. 40
2.13. In Vitro analysis by SDS-PAGE and Western blotting.............................................. 40
2.14. In vitro GD3-Specific Binding Assay for Flow Cytometry Analysis......................... 41
2.15. In vitro Melanoma-Specific Binding Assay for Confocal and Fluorescent Microscopy
Analysis............................................................................................................................ 42
2.16. Melanoma Xenograft SCID Mice............................................................................... 43
2.17. In vivo CCL Treatment and Harvest for Melanoma Xenografts in SCID mice..........44
2.18. Biophotonics Imaging through Biofluorescent Analysis (Biophotonics Box)........... 44
2.19. In vivo analysis by SDS-PAGE and Western blotting................................................ 45
2.20. In vivo analysis by Confocal Microscopy................................................................... 45
CHAPTER 3 – RESULTS............................................................................................................. 47
3.0.
3.1.
3.2.
3.3.
3.4.
3.5.
3.6.
3.7.
3.8.
Can anti-GD3 R24 monoclonal antibody serve as a targeting molecule on CCLs for
delivery of Mcl-1 siRNA to GD3-expressing melanoma?............................................... 47
Incorporating RNA into the CCLs (Integrity Verification)........................................ 48
Desired size of the CCLs............................................................................................ 50
Presence of protein on the surface of CCLs................................................................50
Confirming the binding of R24 antibody to cell surface GD3 of melanoma..............54
Ameliorating the CCL micelle fusion protocol...........................................................57
GD3-targeted CCLs bind melanoma cells while untargeted CCLs do not................. 59
GD3-targeted CCLs exhibit enhanced binding and internalization compared to
untargeted CCLs and decreases Mcl-1 expression in a dose dependent manner............. 59
GD3-targeted CCLs specifically localize, internalize and repress Mcl-1 expression in
LOX-IMVI xenograft tumours in a dose dependent manner following systemic
administration................................................................................................................... 62
vi 3.9.
GD3-targeted CCLs localize and induce Mcl-1 knockdown in Malme-3M xenograft
tumours following systemic administration......................................................................68
CHAPTER 4 – DISCUSSION....................................................................................................... 74
4.1.
Anti-GD3 R24 monoclonal antibody can serve as a targeting molecule on coated
cationic liposomes for delivery of Mcl-1 siRNA to GD3-expressing melanoma............ 74
4.1.1.
Evaluation of RNA encapsulation within GD3-targeted CCLs.........................74
4.1.2.
The Optimal Size of GD3-targeted CCLs......................................................... 75
4.1.3.
R24 Antibody and the Development of GD3-targeted CCLs............................
76
4.1.4.
In vitro Analysis of GD3-targeted and untargeted CCLs.................................. 79
4.1.5.
In vivo Analysis of GD3-targeted CCLs in Mouse Xenograft Models............. 80
4.2.
Future Directions........................................................................................................ 84
4.3.
Conclusions, Significance, and Clinical Relevance of this study............................... 86
REFERENCES.............................................................................................................................. 88
vii LIST OF FIGURES
Figure 1-1
The siRNA pathway of the mechanism of RNA interference in mammalian
cells...........................................................................................................................8
Figure 1-2
The enhanced permeability and retention (EPR) effect relating to drug delivery
systems within a solid tumour................................................................................ 16
Figure 2-1
Preparation of the untargeted siRNA CCLs........................................................... 34
Figure 3-1
High percentages of Mcl-1 siRNA was incorporated and encapsulated into the
CCLs throughout their preparation as quantified by spectrometry analysis.......... 49
Figure 3-2
Mcl-1 siRNA remained intact through the CCL preparation as analyzed via
denaturing urea polyacrylamide gel electrophoresis.............................................. 51
Figure 3-3
All CCLs were measured to be nanoparticles of less than 100nm in diameter as
sized by dynamic light scattering........................................................................... 52
Figure 3-4
Antibody was present on targeted micelles and GD3-targeted CCLs as measured
by the CBQCA assay..............................................................................................53
Figure 3-5
Anti-GD3 R24 monoclonal antibody binds and verifies cell surface GD3
expression on the melanoma cell lines via confocal microscopy...........................56
Figure 3-6
The 37°C overnight fusion of GD3-targeted micelles to Mcl-1 siRNAencapsulated CCLs retained antibody specificity for GD3, confirmed by flow
cytometry................................................................................................................58
Figure 3-7
GD3-targeted Mcl-1 siRNA-encapsulated CCLs enhance the binding and
internalization of Mcl-1 siRNA to Malme-3M and LOX-IMVI cells through
fluorescent microscopy.......................................................................................... 61
Figure 3-8
GD3-targeted Mcl-1 siRNA-encapsulated CCLs decrease Mcl-1 expression in
Malme-3M melanoma cells in vitro in a dose dependent manner......................... 63
Figure 3-9
GD3-targeted Mcl-1 siRNA-encapsulated CCLs specifically localized to the in
vivo mouse LOX-IMVI xenograft tumour following systemic administration via
tail vein injection.................................................................................................... 64
Figure 3-10
GD3-targeted Mcl-1 siRNA-encapsulated CCLs localized to the in vivo mouse
LOX-IMVI xenograft tumour following systemic administration via tail vein
injection in a dose dependent manner.................................................................... 66
viii Figure 3-11
Localization of the GD3-targeted Mcl-1 siRNA-encapsulated CCLs to the in vivo
mouse LOX-IMVI xenograft tumour following systemic administration via tail
vein injection, verified by confocal microscopy.................................................... 68
Figure 3-12
In vivo, GD3-targeted Mcl-1 siRNA-encapsulated CCLs decrease Mcl-1
expression in mouse LOX-IMVI xenograft tumours following systemic
administration via tail vein injection in a dose dependent manner........................ 69
Figure 3-13
GD3-targeted Mcl-1 siRNA-encapsulated CCLs specifically localized to the in
vivo mouse Malme-3M xenograft tumour following systemic administration via
tail vein injection.................................................................................................... 71
Figure 3-14
In vivo, GD3-targeted Mcl-1 siRNA-encapsulated CCLs decrease Mcl-1
expression in mouse Malme-3M xenograft tumours following systemic
administration via tail vein injection in a dose dependent manner........................ 72
ix LIST OF TABLES
Table 3-1
R24 binds multiple melanoma cell lines as verified by flow cytometry................ 55
Table 3-2
GD3-targeted CCLs bind multiple melanoma cell lines as verified by flow
cytometry................................................................................................................60
x ABBREVIATIONS
AGO2
AIDS
ApoB
Bcl-2
BLAST
BRAF
CBQCA
CCL
CD40
CNS
Cy3
DAPI
DDS
DNA
DOPE
DOTAP
dsRNA
DSPC
DSPE
DTIC
EGFP
EPR
Fab
FBS
Fc
FDA
HBS
HEPES
HIV
HPLC
malPEG-DSPE
MAPK/ERK
Mcl-1
MPS
mRNA
NCI
NRAS
OD
P-S
PBS
PEG
PEI
PSMA
RES
argonaute 2
acquired immune deficiency syndrome
apolipoprotein B
B-cell lymphoma 2
basic local alignment search tool
B-Raf proto-oncogene serine/threonine-protein kinase
3-(4-carboxybenzoyl)quinoline-2-carboxaldehyde
coated cationic liposome
cluster of differentiation-40
central nervous system
Cyanine-3
4’-6-diamidino-2-phenylindole
drug delivery system
deoxyribonucleic acid
dioleoylphosphalidylethanolanime
1,2-dioleoyl-3-trimethylammonium-propane
double-stranded RNA
1,2-distearoyl-sn-glycero-3-phosphocholine
1,2-distearoyl-sn-glycero-3-phosphoethanolamine
dacarbazine
enhanced green fluorescent protein
enhanced permeability and retention
fragment antigen-binding
fetal bovine serum
fragment crystallizable
Food and Drug Administration
HEPES-buffered saline
4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid
human immunodeficiency virus
high performance liquid chromatography
DSPE-N-[maleimide(polyethylene glycol)-2000]
mitogen-activated protein kinase / extracellular signal-regulated kinase
myeloid cell leukemia sequence-1
mononuclear phagocyte system
messenger RNA
National Cancer Institute®
neuroblastoma RAS (Rat Sarcoma) viral oncogene homolog
optical density
penicillin-streptomycin
phosphate-buffered saline
poly(ethylene glycol)
polyethylenimine
prostate-specific membrane antigen
reticuloendothelial system
xi RFV
RISC
RNA
RNAi
ROS
SARS
SCID
SDS-PAGE
siRNA
SNALP
TLR
UV
VEGF
relative fluorescence values
RNA-induced silencing complex
ribonucleic acid
RNA interference
reactive oxygen species
severe acute respiratory syndrome
severe combined immunodeficiency
sodium dodecyl sulphate polyacrylamide gel electrophoresis
small interfering RNA
stable nucleic acid-lipid particles
toll-like receptor
ultraviolet
vascular endothelial growth factor
xii CHAPTER 1 – INTRODUCTION
1.1 Melanoma
1.1.1 Epidemiology of Melanoma
The occurrence of malignant cutaneous melanoma incidence has increased more rapidly
than any other malignant cancer in the past 40 years1, 2. The lifetime risk and incidence in
Canada is 1 in 76, 1 in 90 in Europe and a staggering 1 in 32 in the United States3-5. Although
malignant melanoma only accounts for 5% of all dermatologic cancers, melanoma is responsible
for 80% of skin cancer-related deaths6. If melanoma is diagnosed at an early stage, it can be
cured by surgical removal of the tumour1, 7. This resection confers a 90% 5-year survival when
the surgery is treated with or without adjuvant chemotherapy8, 9. However, just as cancer
metastases are the cause of 90% of human cancer deaths, malignant melanoma also has a poor
prognosis1, 10, 11. Once malignant melanoma cells begin proliferation from the primary tumour,
the 5-year survival rate is less than 10%, with a median survival time of 6-9 months after
diagnosis1, 12, 13.
1.1.2 Origin and Etiology of Melanoma
Melanoma is a malignant tumour that arises from mutant melanocytes. Melanocytes
originate from the ectoderm (neural crest) and migrate predominantly to the skin. The main
function of these cells is to synthesize and secrete melanin, a pigment that serves as an
antioxidant and free-radical scavenger. Melanin is transferred to keratinocytes, where it absorbs
and dissipates ultraviolet (UV) light energy, protecting skin cells from carcinogenic mutations8,
14
. The development of melanoma is comprised of environmental stressors and genetic
predispositions.
Exposure to UV radiation is a major risk factor in the development of melanoma6. It is
common that a cutaneous lesion on sun-sensitive and sun-exposed skin can be identified as the
1 site of primary malignancy6, 8. Although chronic or low-grade exposures to UV light induce
protection against DNA damage, intense and intermittent exposures cause genetic damage6, 14.
This UV radiation damage can cause genetic mutations in the skin resulting in impaired
cutaneous immune function, increased local production of growth factors and induced formation
of DNA-damaging reactive oxygen species (ROS) that damage keratinocytes and melanocytes6,
14, 15
. Additional risk factors to melanoma include continued immunosuppression, family history
of the disease, the presence of multiple benign or atypical nevi, and a previous case of
melanoma6.
At the genetic and molecular level, an early step toward invasive melanoma involves the
oncogenic BRAF (B-Raf proto-oncogene serine/threonine-protein kinase) and NRAS
(neuroblastoma RAS viral oncogene homolog) mutations8. Both are exclusively mutated in the
mitogen-activated protein kinase / extracellular signal-regulated kinase (MAPK/ERK) pathway8.
The BRAF V600E activating mutation occurs in 30-70% of metastatic melanoma, while the
NRAS activating mutation appears in 5-15% of metastatic melanoma. The inactivating mutation
of phosphatase and tensin homolog (PTEN), a tumour suppressor gene involved in cell cycle and
its proliferation, is found in 30% of melanomas. In 20-40% of familial cases, the p16INK4A and
p14ARF inactivating mutations and CDK4 activating mutation contributes to cell cycle
dysfunction in melanoma.
1.1.3 Melanoma and Apoptosis
Metastatic malignant melanoma possesses a high resistance to chemotherapy, as response
rates are less than 10%. Because many cytotoxic drugs cause cell death by inducing apoptosis, an
imbalance of apoptotic regulatory proteins may contribute to chemotherapy resistance of
melanoma16-19. Accordingly, this cancer has demonstrated low levels of spontaneous apoptosis
2 and has been shown to be extremely resistant to apoptotic cell death16, 20-22. From these
observations, it has been revealed that Bcl-2-related apoptotic proteins play a critical role in the
chemoresistance in melanoma1, 16. An increased expression of the anti-apoptotic Bcl-2-related
protein, myeloid cell leukemia-1 (Mcl-1) has been observed in malignant melanoma compared to
benign nevi in an immunohistochemical comparison. Additionally, a significant difference in an
Mcl-1 score distinguished benign nevi to primary melanoma, primary melanoma to metastatic
melanoma and expression levels positively correlated with tumour depth. First reported in
differentiating myeloid cells, Mcl-1 levels were seen to decrease in further differentiated
progenitor cells, such as the common myeloid progenitor and the common lymphoid
progenitor23, 24. This research summarized that survival of bone marrow progenitors was
dependent on the expression of Mcl-1. Mcl-1 is also required to promote the survival of less
differentiated cells. Overexpression of Mcl-1 has been reported in other cancers as well,
including breast, lung, testicular, and leukemia24-28. As melanoma continues to acquire metastatic
potential, this cancer displays phenotypes of progressively less differentiated cells.
Our laboratory has previously used an siRNA approach to knockdown Mcl-1 as a means
to induce apoptosis1. This approach resulted in variable cell death. However, when siRNA was
used in conjunction with a small molecule inhibitor (Bcl-2/Bcl-X), cell death was significantly
increased. These experiments demonstrated the critical role that Mcl-1 plays in protecting
melanoma from therapy-induced cell death.
1.1.4 Progression of Melanoma
The Clark model of the progression of melanoma depicts the histological transformation
of normal melanocytes into malignant melanoma29. This accepted model describes the processes
involved in the formation of nevi to the subsequent development of dysplasia, proliferation,
3 invasion and metastasis6, 29. The first phenotypic change in melanocytes is the development of a
nevus, more commonly known as a birthmark or mole. These small pigmented lesions are
formed as the control of growth in melanocytes is disrupted, resulting in the abnormal, but
limited and contained, proliferation of melanocytes. A nevus rarely progresses to cancer. This
arrest in progression is likely due to oncogene-induced cell senescence, where growth stimulated
by the activation of oncogenic pathways is limited6, 30. The intracellular molecules that may be
involved in this phenomenon are those of the cell cycle-regulating MAPK/ERK signalling
pathway. Oncogenes BRAF and NRAS are examples of upstream proteins in this signalling
cascade that are common initiating mutations known to induce cell senescence6.
The next step toward melanoma is the development of dysplastic nevi, which may arise
from pre-existing benign nevi or as new lesions6. The dysplastic stage of benign melanoma is
characterized by random atypia of local cells. The molecular abnormalities at this stage affect
cell growth, DNA repair and the susceptibility to cell death. The frequency of nevi progression
toward malignancy is unknown. Also, accumulation of subsequent mutations within the
pigmented lesions, especially in tumour suppressor genes, can advance the benign nevi into the
radial growth phase6, 8.
In the radial growth phase, melanoma tumours begin to experience proliferation, growing
horizontally within the epidermis. These lesions transform to exhibit clonal proliferation as
differentiation between the cells decrease6. In addition with continuous atypia among the cells,
melanocytes become increasingly similar to stem cells in phenotype. In the radial growth phase,
a melanocytic tumour is considered malignant, but without the ability to metastasize.
Complete invasion into the dermis, with metastatic potential, is characteristic of the
vertical growth phase6. At this stage, melanoma proliferates intradermally as an expanding
4 nodule and acquires the potential to spread to distant sites. Upregulation of genes associated with
tumour angiogenesis has been shown6, 8, 31. Once malignant melanoma cells dissociate from the
primary lesion, migrate to the surrounding stroma, and obtain access to the blood or lymphatic
system, they invade these vessels and form a tumour at a distant site6, 8. These distant sites are
most commonly the liver, lung and brain, with renal metastases reflecting end-stage disease6, 32.
When melanoma metastases are formed, impairing normal functions of these organs, treatment is
beyond the scope of surgical resection. At this stage, prognosis is poor as melanoma represents
one of the most treatment refractory malignancies due to their aggressive behaviour and high
capacity to adapt to environmental stressors8, 13.
1.1.5 Treatments and Challenges of Melanoma
As mentioned earlier, melanoma can be cured through complete surgical resection of the
tumour if the cancer is diagnosed prior to metastasis1, 7. Primary melanoma is typically removed
with a margin of 1-2cm, including neighbouring subcutaneous fat, to achieve local control of the
tumour. In addition, the Breslow thickness of the primary melanoma is measured. The Breslow
thickness refers to the depth of invasion by the melanoma below the granular layer of the
epidermis. Patients found with melanomas with a Breslow thickness of 1mm or more will have a
sentinel lymph node biopsy in the associated drainage area. Such a procedure on the most
proximal lymph node to the primary melanoma site predominantly provides excellent prognostic
information, but also may provide surgical benefit to the patient. However, after melanoma has
metastasized, treatments are limited. Currently, the alkylating agent, dacarbazine (brand name:
DTIC), is the only FDA-approved chemotherapy for metastatic melanoma8. DTIC is the standard
treatment for advanced cases of melanoma, however, the patient response rate to this treatment is
only 5%, with a median survival of 8 months1, 33-35. The alkylating agents, Temozolomide and
5 Fotemustine were compared to DTIC as a single agent treatment. Results showed no differences
in overall survival, but showed a slight increase in progression-free survival (1.5 months to 1.9
months)8, 34, 36, 37. Another alternative treatment is the combination therapy known as
Dartmouth34. This regimen is comprised of the platinum-based cisplatin, estrogen antagonist
tamoxifen, and alkylating agents, carmustine and DTIC. It is widely prescribed outside of
clinical trials, even though it has proven to be more toxic and has not actually shown to be
superior to the single agent DTIC34, 38. Yet another treatment involves the cytokines, interleukin2 (IL-2) and interferon-α (INF-α) in combination with chemotherapy regimens. This treatment
has shown an improved response rate and progression free survival, but has no benefit to median
overall survival and has shown to only benefit a small subset of patients34, 39. Many other single
agents and combination regimens have been tested, yet none of them demonstrate superiority in
overall survival compared to DTIC8, 40.
Malignant melanoma is undoubtedly difficult to cure because of its rapid proliferation,
aggressive invasion into surrounding tissues, high tendency of metastasis to the brain, and
primary resistance to chemotherapy and radiation8, 41, 42. Furthermore, melanocytes may have a
natural resilience to therapies as they normally thrive in a hypoxic environment, relying on
growth factor signalling provided by neighbouring keratinocytes and inflammatory cells for their
survival8, 43. These cells of neuroectodermal origin are also recognized to migrate late during
prenatal development, and thus may have greater undifferentiating potential. For these reasons,
melanoma has this inherent fitness to migrate and survive cellular stress, giving melanoma cells
multiple mechanisms to invade, metastasize and resist therapy, providing rationale for
researchers to continue to look for novel therapeutic approaches. However, none have
significantly prolonged survival time1, 33, 44-47. Despite entering the fourth decade of clinical trials
6 involving cytotoxic chemotherapy drugs for the treatment of metastatic melanoma, there is still
no established effective standard first-line treatment.
1.2 RNA Interference & Small Interfering RNA
1.2.1 The Mechanism of RNA Interference (small interfering RNA)
RNA interference (RNAi) was first discovered by Fire and Mello in 1998 through the use
of double-stranded RNA in Caenorhabditis elegans48. Fire and Mello were awarded the 2006
Nobel Prize in Physiology or Medicine for this finding. In 2001, through the use of the
endogenous RNAi pathway, small interfering RNAs (siRNAs) were shown to specifically and
effectively silence gene expression without activating an interferon-immune response within
mammalian cells49. Following the in vitro success, the first in vivo evidence of RNAi-based
therapeutic efficacy was shown in an animal disease model in 2003, RNAi therapeutics have
represented a novel way to knock down targets that are otherwise ‘undruggable’ with existing
molecules50, 51. RNAi-mediated silencing encompasses multiple steps within the cell (Figure 11). The siRNA pathway begins with the introduction of double-stranded RNA (dsRNA) into the
cytoplasm of the cell52. Facilitated by the RNase III enzyme complex, Dicer, the dsRNA is
processed into a double-stranded complementary sequence of 19-21 nucleotides in length, also
known as small interfering RNA (siRNA)51, 52. The siRNAs are loaded into the multi-enzyme
complex, which includes argonaute 2 (AGO2) and the RNA-induced silencing complex (RISC),
where activation occurs as the sense passenger strand of the siRNA is cleaved. The activated
AGO2-RISC complex, now containing the single stranded antisense business (guide) RNA,
recognizes and binds messenger RNAs (mRNAs) that express a sequence complementary to the
antisense strand. If complementarity is less than perfect, the AGO2-RISC complex blocks
translation. If complementarity is perfect, the catalytic domain of AGO2 cleaves the mRNA
7 Figure 1-1. The siRNA pathway of the mechanism of RNA interference in mammalian cells
(adapted from de Fougerolles, 2007)
RNAi-mediated silencing via siRNA begins with the introduction of dsRNA localizing to the cell
cytoplasm. Facilitated by the RNase III enzyme complex, Dicer, the dsRNA is processed into
siRNA and these are incorporated into the AGO2-RISC complex. AGO2 cleaves and discards
the sense passenger strand, which produces the single-stranded antisense business (guide) strand
that activates RISC. Activated RISC recognizes complementary sequences on gene-specific
mRNAs to direct mRNA cleavage.
8 strand, specifically between nucleotides 10 and 11 of the guide strand to the 5’ end, silencing
protein expression of the particular gene. Gene silencing by mRNA cleavage is thought to be
particularly potent for two reasons: 1) mRNA cleavage products are rapidly degraded by
nucleases within the cell and 2) the activated AGO2-RISC complex is free to seek and destroy
another target mRNA53.
1.2.2 siRNA Therapies
The key advantage of developing siRNAs as drugs is their versatility to target any host
gene or pathogen provided their genetic sequence is known52. Furthermore, with present day
technologies, the therapeutic applications of these molecules can be tested relatively quickly and
modified with ease as required.
There are currently studies that employ direct delivery of naked siRNA into tissues such
as the eye, lung and central nervous system (CNS). Administered directly into its target tissue,
the naked siRNA is buffered in saline or other simple excipients, such as 5% dextrose51.
Effective ocular delivery of siRNA was first demonstrated in animal models of ocular
neovascularisation and scarring51, 54. Developed to treat wet age-related macular degeneration
(AMD), intravitreal injection of siRNAs specifically targeting the vascular endothelial growth
factor (VEGF) expression and the VEGF receptor-1 effectively reduced the area of
neovascularisation attributed to the disease51, 55, 56. . Naked siRNAs have also been delivered to
the less nuclease-rich lung to treat respiratory viruses, such as respiratory syncytial virus (RSV),
parainfluenza virus and severe acute respiratory syndrome (SARS). Intranasal administration of
siRNAs directed against viral genes reduced the viral load of RSV and parainfluenza virus. In a
non-human primate model of SARS corona virus infection, intranasally-administered siRNA
significantly reduced the progression of the disease and the viral load in the lung57. Examples of
9 other proteins that demonstrate the potential for targeted naked siRNAs are heme oxygenase 1
(HO1), angiopoeitin 2 (ANGPT2), and Discoidin Domain Receptor 1 (DDR1), all of which
significantly altered the phenotypes in their respective models through intranasal or intratracheal
administration58-60. Delivery to the CNS via intracerebroventricular, intrathecal or
intraparenchymal infusion has also been shown. These studies have results showing the silencing
of specific neuronal mRNA targets in multiple regions of the peripheral and central nervous
system61-64. Given these positive results, why these cells could directly internalize siRNAs into
the cytoplasm where the RNAi machinery operates, while others could not is unknown.
Although some siRNA therapies have proven effective, localization of siRNA to target
tissues has been the greatest barrier to the development of siRNA therapies52. Most mammalian
cells do not internalize siRNAs in ways that preserve their integrity for effective gene silencing.
In fact, the mechanism demonstrating this phenomenon is not fully understood. Despite these
successes with local injections of siRNA therapies, there are still many hurdles that must be
overcome with siRNA therapeutics.
1.2.3 Challenges to siRNA Therapeutics
From the initial sequencing of the double stranded RNA to the ultimate in vivo efficacy of
the siRNA therapy, complications may occur. There are a variety of challenges that accompany
the development of effective siRNA therapies, including specificity, potency, stability, size and
charge, and delivery.
1.2.3.1 Specificity
RNAi-mediated silencing of gene expression is highly specific. Contributing to its
specificity, the string of nucleotides in positions 2 to 8 from the 5’-end of the antisense business
siRNA strand are critical. This string is otherwise known as the “seed region” of the siRNA for
10 target mRNA recognition51, 65, 66. To minimize adverse silencing, the seed region of the siRNA
must be designed to specifically bind a section of the target mRNA that is unique to all other
genes. In order to find this target mRNA, a Basic Local Alignment Search Tool (BLAST) search
may be performed67. This resource identifies genes with a high degree of similarity so the target
gene may be found. Despite careful designs and validation of siRNAs through experimentation,
off-target effects may still take place. Although undesirable silencing is usually less than 3-fold,
there can be silencing of alleles when the siRNAs contain single nucleotide polymorphisms68-71.
Activation of immune and inflammatory responses have also been shown to be sequencespecific51, 52, 72. Mediated by the toll-like receptor (TLR) pathway, synthetic siRNA drugs must
be developed to evade TLR9 and TLR7, the TLRs involved in recognizing foreign RNAs.
Immunostimulation by TLR9 is mediated through the recognition of CpG motifs in antisense
oligonucleotides. TLR7 activation is mediated by Guanine/Uracil-rich sequences, as the receptor
demonstrates a specific binding affinity to these RNA nucleotides. Found largely in endosomes
of immune antigen-presenting cells, TLR7 and components of its pathway, trigger the production
of type I interferons and pro-inflammatory cytokines via nuclear factor-kappa-B (NF-κB)
activation73. If the candidate siRNA does not induce an adverse interferon immune response
when treating antigen-presenting plasmacytoid dendritic cells, acknowledged to be the gold
standard model for immunostimulatory activity, the siRNA may qualify for in vivo and clinical
therapeutic use.
1.2.3.2 Potency: advantage of using Dicer-substrate siRNA (DsiRNA)
Administering a minimal dose that provides therapeutic effect is the pinnacle goal of drug
therapy. Thus, optimizing the potency of selected siRNAs is important in developing RNAimediated silencing. Corresponding to the amounts of drug administered, adverse effects and
11 toxicity may arise if treatment concentrations are too high in vitro or in vivo. Potency of siRNA
therapeutics has been measured by in vitro activity at varying concentrations, usually in the
nanomolar range51. Yet, subnanomolar concentrations have been shown to produce gene
silencing74. Studies have demonstrated that longer synthetic siRNAs, 25-30 nucleotides in length,
are more effective at silencing than the 21-nucleotide siRNAs; for example, blunt 27-oligomer
dsRNA can be up to 100-fold more potent than traditional 21-oligomer duplexes74, 75. Although
the longer synthetic siRNAs are more difficult to synthesize, they have an advantage. Longer
siRNAs require Dicer-processing, which creates their substrate. This substrate is a direct link to
the production of siRNAs which aids incorporation into the AGO2-RISC complex. These are
known as Dicer-substrate small interfering RNAs (DsiRNAs)74, 76, 77. To further increase
potency, synthetic siRNAs are created asymmetrically, with a strand containing a 2 base pair 3’
overhang. This is done because the binding pocket of RISC is not symmetrical. The strand
containing the overhang becomes the recognized business antisense strand, binding the deep
pocket of RISC, as its 5’-end is more exposed for Dicer processing51, 74, 78. These modifications
to the traditional siRNA molecule greatly enhance its potency because once inside a cell, the
siRNA is incorporated and locked into RISC. This has shown the persistence of silencing in
some cell cultures for several weeks50, 52, 79, 80.
1.2.3.3. Stability
The half-life of unmodified and uncomplexed siRNAs in nuclease-rich human blood
plasma is only a few minutes, limiting their usefulness as drugs51, 81, 82. By adding modifications
to siRNA, one can stabilize the molecule and ameliorate its pharmacokinetic properties to
produce clinically relevant benefits. 2’-sugar modifications provide siRNA with the capacity to
evade endonuclease degradation and are considered to be more tolerable if located on the sense
12 strand. Examples of this type of modification are the 2’-fluoro and the 2’-O-methyl, the latter
being the more common modification. In addition to providing endonuclease evasion, the single
2’O-methyl modification at nucleotide 2 has shown to suppress off-target molecules without
losing its specificity for their target mRNA83, 84. Providing exonuclease resistance, the 3’-end
backbone of the siRNA is modified with phosphorothioate bonds (P=S) to replace the normal
phosphate bonds (P=O)51. To evaluate the effect of siRNA modifications on intracellular activity,
the fluorescence resonance energy transfer (FRET) technique has been used. This technique
optimizes the balance between siRNA stabilization and silencing efficacy85. Although naked
siRNAs do not exhibit longevity in mRNA silencing in nuclease-poor sites such as the lung,
nuclease-stabilized siRNAs have illustrated significantly enhanced pharmacokinetic properties in
vivo52, 81, 86, 87. These chemistries have shown to improve the persistence of intact siRNA,
stabilizing the molecules during transit to their target site and making use of the RNAi pathway
once within the cell.
1.2.3.4 Size and Charge
Clearance and adverse immune effects are directly linked to the final size of the siRNA.
If the siRNA drug is small, it will not bypass the renal filtration cut-off (approximately 50kDa)
and will consequently be eliminated by renal excretion51, 52. If the siRNA drug is too large, innate
immune cells may be activated, setting off defence systems used to combat foreign viral
pathogens. This upper limit in size is dictated by the length of the siRNA sequence, which is
recognized by serine/threonine-protein kinase receptors. An immune and inflammatory response
will be activated by dsRNAs 30 nucleotides or longer in length. Accordingly, optimal sizes of
potential siRNA drugs fall between 21-30 nucleotides in length with a molecular weight
exceeding 50kDa.
13 The charge of siRNA is also an important aspect to consider. Recognition by the body’s
reticuloendothelial system (RES), also known as the mononuclear phagocyte system (MPS), may
contribute to the elimination of siRNA drugs due to its charge. Localized to the liver and spleen,
these clearance systems phagocytise molecules with any non-neutral charge and protect normal
host cells from foreign infection51. In order to avoid undesired degradation and elimination, these
drugs must not demonstrate extensive positive or negative charges.
1.2.3.5 Delivery
The greatest obstacle to siRNA drug development is its effective delivery to target
diseased sites within the host and its intracellular cytoplasmic localization to utilize the RNAi
machinery51. Effective delivery of siRNA encompasses the challenges mentioned above,
including the drug’s stability, size and overall charge. In response to these multiple challenges,
drug delivery systems are developed to complex with siRNA molecules, assisting in its effective
systemic delivery.
1.3 Drug Delivery Systems
1.3.1 Introduction
Systemic administration of siRNAs is essential to target metastatic sites of cancer.
However, siRNA molecules rarely reach their intended site, and are not readily internalized into
cells. A drug delivery system is needed. In general, a drug delivery system (DDS) is a synthetic
nanoparticle carrier, effective once it has formed a complex with a drug88. Usually composed of
lipids and-or polymers, DDSs are designed to improve the pharmacological and therapeutic
properties of a drug when administered parenterally (intravenous, intramuscular, or
intraperitoneal). Consequent to their composition, DDSs alter the pharmacokinetics and
biodistribution of their associated drugs. There are a multitude of nanoparticle DDSs, all
14 categorized to have diameters of 200nm or less. Examples of these include polymer
microspheres, polymer-drug conjugates, lipid-drug complexes, lipid emulsions, micelles and
liposomes, all of which have shown to ameliorate drug delivery88-93. The main function of these
small-scale systems is to prolong a drug’s half-life in the blood plasma through drug
stabilization, optimizing the drug size, and negating the charges that may have accompanied the
free drug. DDSs produce extended circulation time and give drugs the opportunity to accumulate
at their intended site via the enhanced permeability and retention effect.
1.3.2 The Enhanced Permeability and Retention Effect
The enhanced permeability and retention (EPR) effect is the accumulation of circulating
drug at its intended sites (Figure 1-2). It is sometimes referred to as ‘passive targeting’88, 94. The
DDS-related biodistribution changes are a result of this mechanism; the greater the circulation
half-life, the greater the accumulation of drug at the target site via the EPR effect, especially as
peak concentration levels of drug do not occur until 1 to 3 days post-injection. This phenomenon
is the result of an increase in permeability of the vasculature and impaired lymphatic drainage
surrounding a diseased site95-97. In many pathological conditions, including the inflammatory
response, vascular remodelling takes place and leukocytes are free to enter the interstitial tissue98,
99
. DDSs are no exception; when complexed to drugs, they are able to extravasate and localize to
the target tissue interstitial space88, 95. Also, solid tumours display ‘leaky’ vasculature, causing
highly vascularised tumours to accumulate higher concentrations of DDS. Solid tumours also
employ the EPR effect from its neoangiogenesis. During their growth, tumours will demand a
greater supply of oxygen and nutrients than readily available, releasing cytokines and other
signalling molecules which initiates the formation of new blood vessels. The resulting
neovasculature, differing from the tight blood vessels in most normal tissues, will have gaps
15 Figure 1-2. The enhanced permeability and retention (EPR) effect relating to drug delivery
systems within a solid tumour (adapted from Torchilin, 2007)
The EPR effect, also referred to as ‘passive targeting’, is the accumulation and retention of
circulating drug to sites with increased vasculature permeability and impaired lymphatic
drainage. The result is improved biodistribution of a drug. While ‘free’ drug may be able to
permeate through the any endothelial cells of the vasculature (A), drug associated with drug
delivery systems will not be able to elicit these off-target effects (B). The drug delivery system
will circulate throughout the body until it can readily extravasate through ‘leaky’ vasculature,
entering the interstitial tissue (C). Pathological conditions, including the inflammatory response
and vascular remodelling within solid tumours, present a vasculature with increased
permeability. Within the interstitial space, release of the drug from the drug delivery system will
not present off-target effects, rather, facilitate its uptake by cells within the microenvironment
(D).
16 between adjacent endothelial cells as large as 600nm to 800nm in diameter, enabling DDS
accumulation. On the other hand, if the tumour is pre-angiogenic, or displays necrosis throughout
the tumour, DDSs accumulation will be poor. The accumulation of drug within the full mass of
the solid tumour is not homogenous, rather it is specific to areas with a proximity to
vasculature88, 100. A full explanation has not yet been achieved, but it has been elucidated that
DDSs of less than 100nm in diameter penetrate deepest into tumour tissues101. In summary,
many factors contribute to the EPR effect. These include the circulation half-life of the DDS, the
diameter of the DDS-complexed drug, the size of the blood vessel pores and the degree of
tumour vascularisation and angiogenesis. Together, these characteristics allow for the
accumulation of DDS-complexed drug within the tumour microenvironment, increasing drug
concentrations 10-fold or higher when compared to administration of the same dose of free
drug88, 96, 97.
1.3.3 DDSs for siRNA Delivery: Non-Covalent Drug Complexes
In our present day, there are many drug complexes and drug conjugates being researched
to ameliorate siRNA therapeutics. Non-covalent drug complexes refer to DDSs where the siRNA
molecules create stable complexes with the DDSs through non-covalent interactions. These
complexes make use of the negatively-charged siRNA backbone. By creating a DDS that is
positively charged, the formation of drug complexes is enabled. This results in a cationic DDS
which has different pharmacokinetic biodistribution properties than the siRNA molecule alone.
A polymer that has been used to create siRNA drug complexes, which has proven therapeutic
effects in vitro and in vivo, is polyethylenimine (PEI). PEI is a synthetic structure with large
cationic charge densities used to interact with the siRNA backbone51. These positively-charged
PEI-siRNA complexes create cationic polyplexes, which bind to the anionic cell surface through
17 electrostatic interactions, and are subsequently endocytosed into the cell. Within the endosome of
the cell, it is believed that the cationic nature of the DDS disrupts the low endosomal pH,
resulting in endosomal escape. Once released from the endosome, the complex will localize to
the cytoplasm of the cell where the RNAi mechanism can be initiated. Many studies have
demonstrated effective in vivo applications of PEI complexed with siRNAs. Ge et al. showed
antiviral effects in mice with induced influenza102. Geisbert et al. demonstrated partial protection
against the lethal Ebola infection in guinea pigs103. And others have shown anti-tumour activity
in mouse models104-106. A troublesome complication with these drug complexes is the extreme
toxicity of PEI seen at higher doses within animal models. Without a remedy, the DDSs of PEI
therapeutics are still in the pre-clinical stage.
The chitosan-siRNA complex is another non-covalent example of a DDS. Administered
intranasally, these nanoparticles show in vivo efficacy in lung epithelial cells, decreasing
transgenic enhanced green fluorescent protein (EGFP) expression by approximately 40%107.
Furthermore, intravenous administration of chitosan-siRNA complexes displays RNAi effect and
anti-tumour activity in subcutaneously xenografted breast cancer in mice108. Antibodies have
also been used in developing non-covalent drug complexes. For example, the protamineantibody fusion protein non-covalently binds its siRNA by having the protamine particle bind
both the oligonucleotides and the fragment antigen-binding (Fab) portion of the antibody109. This
approach displayed silencing in vitro on human immunodeficiency virus (HIV)-envelopeexpressing B16 melanoma cells and HIV-infected primary T cells in vitro with high efficiency.
These ligand-specific antibodies may also be directly conjugated to siRNA molecules, termed as
covalent drug complexes.
18 1.3.4 DDSs for siRNA Delivery: Covalent Drug Complexes
Covalently-bound drug complexes are another type of DDS that have been researched to
a great extent. These complexes show superior reproducibility and stability when compared to
the non-covalent drug complexes. It is common for the covalent conjugation to take place on the
passenger sense strand, in order to avoid disrupting the siRNA activity and adding unnecessary
modifications to the business antisense strand51. Covalently coupling cholesterol to the passenger
strand of the siRNA has been effective in vivo when delivered to the liver51, 110, 111. The conjugate
used was composed of an siRNA duplex, sequenced to knock out apolipoprotein B (ApoB), and
then conjugated to a cholesterol at the 5’-end of its passenger strand. Administered intravenously
into mice, this drug displayed uptake through ubiquitinously expressed cell-surface LDL
receptors, silencing ApoB expression in the liver and the jejunum by 55% and 50%, respectively.
The physiological effect that resulted from this knockdown was a reduction of serum cholesterol
in mice by more than 35%. This effect occurs because the mechanism that the cholesterol
incorporates itself into circulating lipoprotein particles within the blood plasma, significantly
prolonging their circulating half-life. Accordingly, these particles eventually localize to the liver
and to the lower alimental tract where it is endocytosed by cholesterol-receptor-expressing cells,
eliciting its mRNA silencing response. Generally speaking, protection from blood plasma
nucleases, especially degradation from exonucleases, will be acquired by coupling the siRNA to
any DDS. Furthermore, with the increased size of the entire DDS, the 50kDa cut-off for renal
filtration and elimination will not be significant; the drug will remain in circulation51, 52. The last
significant benefit to covalently conjugating a DDS to a siRNA molecule is its ability to target
specific cell populations. Cell populations naturally express an upregulated cell-surface receptor
during a diseased state. An example is illustrated in prostate cancer, which expresses a prostate-
19 specific membrane antigen (PSMA). A DDS was developed to target PSMA by aptamers,
nucleic acid binding species that are non-protein based alternatives to antibodies112, 113. These
aptamers, conjugated to siRNAs, create a chimeric RNA-based DDS that has shown to promote
specific cellular uptake by prostate cancer cells expressing human. In spite of showing evidence
for RNAi-mediated mRNA silencing in vitro and in vivo, these short aptamer conjugates are
rapidly cleared by the kidney, having a circulation half-life that is too short for systemic therapy.
Other researchers have created nanoparticles that include transferrin, folate, various
peptides and sugars, galactose and mannose, conjugated to their siRNA molecules51, 74, 105, 114, 115.
These examples have also demonstrated natural-receptor targeting to particular cell types
through in vivo applications. Protease-treated collagen, atelocollagen-conjugated siRNA have
also shown positive effects through systemic administration. In a subcutaneous tumour xenograft
model, Takei and colleagues demonstrated that targeting a model metastatic site, specifically the
tumour angiogenesis instead of a primary tumour, was possible116, 117.
Covalent and non-covalent drug complexes have both shown incredible progress as
DDSs, improving siRNA therapeutics in many aspects. However, these DDSs keep the siRNA
exposed to its surrounding environments, potentially decreasing their efficacy as a therapy.
Additionally, the molecules to which the siRNA molecules are interacting may be eliciting some
unknown adverse effects within the host, such as immune responses. Also, the toxicities
associated with these DDSs, especially with cationic molecules interacting with the siRNA, may
prove to be problematic. In order to solve some of these problems, conjugated-based approaches
can be coupled with other delivery system to enable a more targeted delivery of oligonucleotides.
An example of this is the liposome51, 105, 118-120
20 1.3.5 DDSs for siRNA delivery: Liposomes
Liposomes are a class of DDSs that are very versatile when compared to drug complexes.
While drug complexes rely on the release of the drug into the microenvironment and physically
binding a cell-surface ligand to be internalized, liposomes have the capacity to fuse with cell
membranes, enhancing drug delivery into a cell51, 52. Liposomes are self-assembling vesicles that
can be variable in size, lipid composition and encapsulated drug within its phospholipid bilayerenclosed aqueous compartment. Liposomes are composed of multiple lipids, including cationic
lipids, neutral lipids, cholesterol and polyethylene glycosolyated lipid.
Cationic lipids are added to the liposome formulation due to the functionality of their
positive charge. This positive charge is attracted to the negatively-charged cell surface molecules
and to interact with the naturally occurring anionic phospholipids found in the endosomal
membrane121, 122. These interactions form ion pairs, which adopt non-bilayer structures and
disrupt the endosomal membranes. The result is the liposomal drug release into the cell
cytoplasm, where the drug becomes activated and takes effect.
Similar to the cationic drug complexes, the major drawback of choosing a molecule with
a strong, constitutive positive charge is that they are hunted by serum proteins and rapidly taken
into the tissues of the MPS123. For these reasons, neutral lipids are added into the composition of
liposomes. Neutral lipids, such as 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC),
distributes the positive charges evenly within the liposome particle. Furthermore, these lipids
favour and give rise to the spherical unilamellar bilayer of liposomes as nanoparticles less than
200nm in diameter. Another neutral lipid, cholesterol, is used to stabilize the dynamic
membranes of a liposome. With its hydrophobicity and rigid structure, these cholesterol
molecules localize within the hydrophobic areas of the lipid bilayer to help preserve the shape
21 and size of the liposome. Poly(ethylene glycol) (PEG) molecules provide a similar functionality
to the cholesterol, as they help to stabilize the liposome51, 121. In addition to this, they also
prevent the aggregation between liposome particles and consistently control its particle size
formation. Most importantly, these long-chained hydrophilic molecules improve the
pharmacokinetic properties of a liposome. PEG shields liposomes’ positive surface charges that
would otherwise stick to negatively charged off-target cell membranes. PEG also prolongs the
circulation half-life by helping to evade MPS recognition and by increasing the size of the
liposomes to slow renal clearance following intravenous injection.
The last component of a liposome is the drug to be delivered. Hydrophilic drugs can be
readily encapsulated and stored within the aqueous interior as an inactive drug. Assuming that
the encapsulated hydrophilic drug is stable within the liposome and not prone to leakage or
degradation, the drug can remain in this latent state until its activation and release51, 88.
Presently, three approved liposomal chemotherapeutic agents employ these
chemistries124, 125. The first is liposomal daunorubicin (Daunosome®), approved in the USA and
Europe to treat acquired immune deficiency syndrome (AIDS)-related Kaposi’s sarcoma. The
second is liposomal doxorubicin (Myocet®), approved in Europe to treat metastatic breast cancer.
Lastly, PEGylated liposome doxorubicin (Doxil®/Caelyx®) is approved in the USA and Europe,
used to treat Kaposi’s sarcoma, refractory ovarian cancer and metastatic breast cancer. Many
other liposomal anticancer drugs, such as liposomal cisplatin and liposomal cytosine arabinoside
are in clinical trials88, 118.
The liposomal core may also house non-chemotherapeutic drugs, including hydrophilic
oligonucleotides. Through interactions between its negatively-charged backbone and the cationic
lipids, siRNAs have recently shown efficient encapsulation within the aqueous interior of a
22 liposome. Developed by Tekmira Pharmaceuticals, stable nucleic acid-lipid particles (SNALPs)
are examples of PEGylated liposomal siRNA therapeutics. Intravenously administered SNALPs
localize to the liver where it significantly silences ApoB in both mice and non-human primates
during pre-clinical studies80, 126. This occurs because SNALPs reduce uptake by scavenger cells
and enhance time spent in circulation, maximizing the EPR effect into Kupffer cells. The
silencing effect of SNALPs reduces ApoB mRNA in the liver by more than 90% with a single
dose. As a physiological outcome, serum cholesterol and low-density lipoproteins were reduced
by more than 65% and 85%, respectively. These values are much greater than the values
observed with the drug complex, cholesterol-siRNA111. The application of these nanoparticles
does not end as an anticancer drug. SNALP formulations have also effectively delivered siRNA
to the liver to reduce the replication of hepatitis B and to combat the Ebola virus infection in
animal models103, 127. SNALPs have one major setback preventing them from becoming a widely
used therapy. It is uncertain whether SNALPs can be used for delivering siRNA to organs other
than the liver51. Liposomes are a DDS with improved pharmacokinetic properties, yet there are
still many challenges accompany their use to deliver drugs to tumours.
1.3.6 DDSs for siRNA Delivery: Challenges to Liposomal Therapies
Challenges arise at every step of the development of a DDS. The formation of drugencapsulated liposomes has been thoroughly studied and its reproducibility does not pose as a
problem. Challenges arise once the DDS is administered to an in vitro or in vivo model, as the
bioavailability, biodistribution, toxicity, and delivery of liposomal drugs are not always
favourable.
The bioavailability of a drug is the percentage of an administered dose that reaches the
systemic circulation. Since liposomal therapies can be administered intravenously, this is of
23 minimal concern. PEGylation of the liposomes creates a hydrophilic region surrounding the
hydrophobic liposome bilayer, allowing for the DDS to solubilise in a polar solution. It is
assumed that drug solubility is highly efficient. By adopting a DDS, toxicity is theoretically
decreased8. Because liposomes are made with non-toxic and biodegradable ingredients, toxicities
associated with these carrier molecules would be mild88.
Examples where drug toxicities have decreased due to liposomal DDSs include the
liposomal anthracyclines, Daunosome®, Myocet®, and Doxil®/Caelyx®. Thousands of drug
molecules can be administered because drug potency is decreased by encapsulation128. Naturally,
if the carrier is used at unreasonably high quantities, problems with the liposome metabolism and
elimination may occur88. Toxicity effects are quite different when comparing liposomes and
siRNA-based therapies. Naked siRNAs display off-target effects and rapid clearance with
minimal toxicity, while siRNA-encapsulated liposomes have a very different effect. When
repeatedly administered, new toxicities are observed with liposomes through the induction of
immune responses and complement activation129-132. Additionally, liposomes have shown the
natural tendency to localize and accumulate at the MPS, particularly the spleen, preventing the
drug from reaching its target site88. For these reasons, the EPR effect may not be a major factor
for liposomal siRNA biodistribution. This aside, drug release rates are also points of concern. In
liposomal systems, the drug is inactive while encapsulated and failure to release the drug from its
interior may result in a reduced therapeutic effect133. On the contrary, if the drug is too rapidly
released in the presence of plasma proteins, the biodistribution of the liposomal drug will
approach those of the free drug, with high toxicities and no localization to the diseased site134.
Theoretically, the optimal rate of release of the drug is between these two extremes88. Slow drug
release rates allow the liposomal siRNAs to remain encapsulated and protected and to localize to
24 the diseased site through the EPR effect, where it may then slowly be released into the tumour
microenvironment for uptake and effect. Factors that contribute to the final biodistribution of
liposomal siRNAs are its size and its charge. During the formation of the nanoparticle, size can
vary between 50 and 250nm in diameter. It is possible for the manufacturer to manipulate
liposome to an optimal size. Optimal liposome size is small enough to benefit from the EPR
effect and permeate deep into tumour tissues, yet large enough to evade renal filtration and
elimination. Also, by optimizing liposome composition, net positive surface charges can be
minimized, thus uptake and clearance via the MPS will diminish.
Identical to siRNA therapeutics, the greatest challenge to the liposome DDS is its
effective targeted delivery. To solely rely on the combination of a prolonged circulation half-life
and EPR effect for localization of liposomal siRNA to the diseased site has sufficed and shown
positive results in certain primary tumours or liver-specific therapies. However, to target a
specific tumour cell population, regardless of their tumour development or location within the
host, the addition of a targeting moiety is essential for site-specific siRNA delivery.
1.3.7 DDSs for siRNA Delivery: Targeted Liposomal Therapies
PEGylated liposomes have already been approved as cancer treatments in many counties
around the world. The next step in the development of liposomal therapeutics is to have these
drugs specifically target diseased sites. The greatest progress in targeted liposomal siRNA
treatments has been the development of the SNALP. They have demonstrated long lasting
effective silencing at a practical siRNA dose (90% knockdown with 11 days of ≤3 mg/kg/day) in
vivo. In the last decade, there have only been a few examples which have employed targeted
liposomal siRNA treatments.
25 Cationic liposomes associated with transferrin and complexed with siRNAs were
delivered to primary cultures of luciferase-expressing cortical neurons135. Targeting the
transferrin receptor, efficient uptake of these targeted liposomal Cyanine-3 (Cy3)-labelled
siRNAs was observed in vitro through confocal microscopy. To control for these results, uptake
was reduced by blocking the transferrin receptor with excess transferrin, a competitive binding
assay. Using anti-luciferase and anti-c-Jun siRNAs, gene silencing in vitro was up to 50%, 48
hours after transfection without significant toxicity. In vivo results were similar with minimal
toxicity. A 40% reduction in luciferase activity was observed in the striatum of luciferase mice.
In addition, fluorescent microscopy illustrated extensive local distribution of transferrin-targeted
liposomal Cy3-labelled siRNAs among the neuronal cells of the mice.
The use of antibodies as targeting molecules has also shown positive results. Specifically
targeting the dendritic cell marker DEC-205, the NLDC-145 monoclonal antibody was
conjugated to the liposomal CD40-targeted siRNA136. Through fluorescent microscopy, bone
marrow-derived dendritic cell-specific binding and internalization was observed after a 4°C
incubation of 30 minutes, where the control L292 cell line did not show either. When
administered intravenously into mice, these targeted liposomal siRNAs accumulated in the liver
and the spleen, organs rich in dendritic cells, and showed no unspecific localization to the kidney
or lung. Approximately 60% of gene silencing was observed with this treatment, eliciting a
significant decrease in surface expression of CD40 and mRNA levels in dendritic cells. Control
treatments and other immune cells, such as B cells, were not affected in this experiment.
Additionally, silencing of CD40 was exhibited for 12 days, much longer than the published
maximum of 7 days when injecting naked siRNA. This phenomenon is attributed to how
dendritic cells do not go through cell division to dilute internalized siRNA.
26 1.3.8 DDSs for siRNA Delivery: Potential Targeting Molecules for Melanoma
Upon review of the literature, a number of potential melanoma surface molecules have
been identified as drug delivery targets. Tyrosine preferentially accumulates in melanoma cells
and is essential for synthesis of melanin137. In fact, in vitro and in vivo melanoma growth is
reduced when tyrosine is restricted138-143. Although increased in melanoma, receptors for the
amino acid tyrosine exist in many cell types. Another melanoma surface molecule is alphamelanocyte-stimulating hormone (α-MSH), a small tridecapeptide involved in the control of skin
pigmentation144. The biological activity of α-MSH is mediated through interactions with the
melanocortin-1 receptor (MC1R)145. The MC1R has been identified both on human and mouse
melanoma cells. In fact, more than 80% of human metastatic melanoma tumours bear MC1R146.
Yet another surface molecule is GD3, a surface ganglioside whose expression is largely limited
to tumours of neuroectodermal origin, such as melanoma and neuroblastoma147, 148. Antibodies
directed against GD3, and a similar ganglioside GD2, are both capable of binding to, and
internalizing into melanoma cells149, 150. GD2 has been shown to be more characteristic of
neuroblastoma, while GD3 has demonstrated a greater specificity to melanoma42, 151-153.
1.4 Summary and Goals
Despite the multitude of challenges that arise from the delivery of RNAi-mediated gene
silencing, literature has shown ways to overcome each of these. As naked siRNA drug is
degraded rapidly in circulation, a drug delivery system is used. To protect the siRNA and evade
renal elimination, the drug delivery system of liposomes is formed to encapsulate the drug.
Conjugated to liposomes are molecules that not only shield the nanoparticle from untoward
uptake by the MPS and its associated organs, but aid in its delivery to specific diseased sites.
Studies using targeted liposomal siRNA treatments have demonstrated localization of the therapy
27 at primary tumours and metastatic sites, which express a tumour cell-specific cell surface
marker. The goal of my research was to create a drug delivery system that would improve the
efficacy of siRNA delivery to melanoma. The DDS was developed to specifically bind
melanoma therefore cell surface markers for this malignancy need to be recognized. My research
has focused on the melanoma cell surface marker, disialoganglioside GD3. To target this
molecule, anti-GD3 R24 monoclonal antibody has been employed. Using this targeting moiety,
liposomes were developed and evaluated for their efficacy in vitro and in vivo. For these studies,
Mcl-1 siRNA was employed as the payload. As mentioned earlier, the Tron Laboratory has
previously demonstrated a need for Mcl-1 knockdown in vitro.
I hypothesized that anti-GD3 antibodies are targeting molecules for delivery of siRNA to
melanoma.
Furthermore, my research has involved continual optimization of the coated cationic
liposome (CCL) formulations, including the attachment of its targeting molecule. With flexibility
and variability in the composition of the liposomes and their chemistries, the progression in the
development of my liposomal siRNA therapy will be illustrated. These advances were all aimed
toward localizing the siRNA to the cell cytoplasm in order to induce effective RNAi-mediated
gene silencing in melanoma.
Accordingly, my objectives are as follows:
1. To create CCLs with encapsulated Mcl-1 DsiRNA and targeted anti-GD3 antibody.
2. To deliver targeted CCLs in vitro containing DsiRNA.
3. To deliver targeted CCLs containing DsiRNA to mouse melanoma xenograft tumours.
28 CHAPTER 2 – MATERIALS AND METHODS
2.1 Preparation of (Untargeted and Targeted) Empty Coated Cationic Liposomes (CCLs)
2.1.1 Chemicals, Reagents, and Materials
Cholesterol, 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), 1,2-distearoyl-snglycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000] (mPEG-DSPE) and
1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[maleimide(polyethylene glycol)-2000]
(malPEG-DSPE) were purchased from Avanti Lipids Inc. (Alabaster, AL). The Whatman
nucleopore Track-Etch membrane (Piscataway, NJ) polycarbonate filters used during extrusion
were purchased from Fisher Scientific Co. Ltd. (Ottawa, ON). Sephadex G-50 (Sigma-Aldrich;
St. Louis, MO) sepharose beads for column purification were dissolved in a 0.2M NaOH
solution for storage. All of the other reagents of biochemical and molecular biology grade were
obtained from Sigma-Aldrich (St. Louis, MO). HEPES-buffered saline was composed of 25mM
HEPES (Sigma-Aldrich) and 140mM HCl (Sigma-Aldrich), brought to a pH 7.5, and
deoxygenated with nitrogen gas for 20 minutes. All glassware used in the preparation of CCLs
was baked at 150°C for 4 hours in order for them to become RNase-free. An additional reagent
used in this protocol was Traut’s reagent (Thermo Scientific; Rockford, IL). Traut’s reagent is a
water soluble compound that reacts with primary amines to introduce sulfhydryl (thiol) groups
while maintaining similar charge properties to the original amino group. It was used to create
thiol groups on the melanoma-specific targeting molecule, anti-GD3 R24 monoclonal antibody.
The R24 monoclonal antibody was acquired from the NCI (Frederick, MD). For the antibody
purification steps, pre-packed PD10 columns (GE Healthcare; Buckinghamshire, UK) were
purchased. Also, to verify the presence of antibody during its purification, the spot test, Bio-Rad
Reagent (Bio-Rad; Hercules, California) was purchased.
29 2.1.2 Preparation of the anti-GD3 R24 Antibody
To prepare the R24 antibody, it was first thiolated using Traut’s reagent. This allowed for
the R24 antibody to conjugate to the maleimide group on malPEG-DSPE. First, a 15mM stock
solution of Traut’s reagent was created by dissolving 15µmoles (2.06mg) of Traut’s reagent in
1mL of HBS buffer. Immediately upon hydrating the Traut’s reagent, it was added to
approximately 2.045mg of antibody at molar ratio of 20:1, Traut’s reagent to protein (8.9µL of
stock solution to 1mg of antibody). This was allowed to incubate at room temperature for 1 hour,
rocking. To separate the thiolated R24 antibody from unconsumed Traut’s reagent, a desalting
PD10 column was used with HBS buffer. After the column, the antibody solution was subjected
to the spot test using BioRad reagent. Aliquots showing with the presence of antibody were
collected, pooled together and stored at 4°C. The concentration of thiolated antibody was
analyzed via NanoDrop Spectrophotometer ND-1000 (Wilmingdon, DE).
2.1.3 Empty CCL Preparation Protocol
Empty CCLs were created and used as control treatments to siRNA-encapsulated CCLs.
Stored at -20°C in a chloroform solution, DSPC, cholesterol, mPEG-DSPE and malPEG-DSPE
were mixed at a molar ratio of 40:20:2:0.2 for a total of approximately 20µmol of lipids within a
15mL Kimble Chase glass culture tube (Sigma). After adding 1000µL of RNase-free water, the
mixture was sonicated for 10 minutes in the VWR Aquasonic 50D Ultrasonic Cleaner (West
Chester, PA). The mixture dried as the chloroform evaporated by thin film rotary evaporation
using the Heidolph Laborota 4010-digital (Shire Hill, UK). Once the liquid turned to a white gellike film, toothpaste-like, another 1000µL of RNase-free water was added. Rotary evaporation
continued until the white solution became homogenous and translucent, without noticeable
solids, signifying that all lipids had successfully inverted to form liposomes. This resulting
30 solution was then extruded through 200nm, 100nm and double 100nm filters to progressively
decrease the size of the liposomes. Extrusion was performed at pressures between 150 to 300psi
in the extruder, VWR Signature™ Heating Circulator (West Chester, PA). Finally, the liposomes
were exchanged into HEPES-buffered saline by chromatography through a 10mL column packed
with Sephadex G-50 (Sigma) beads. Immediately after the collection of the liposomes from the
column, the liposomes were split into two volumes: liposomes that would acquire the R24
antibody as a targeting molecule and the other, untargeted. To the first aliquot, the R24 antibody
stock concentration was added at a 1:10, antibody to malPEG-DSPE molar ratio. Approximately
0.003µmol (0.450mg) of antibody was added. The untargeted CCLs would receive the same
volume of HBS buffer and was designated as untargeted liposomes. Each aliquot rocked
overnight at room temperature. Following overnight incubation, both targeted CCLs and
untargeted CCLs were stored at 4°C until analyzed or until they were required for treatment. It is
important to note that the duration for reactivity of the maleimide on the malPEG-DSPE
molecule to the thiol group of the R24 antibody is only 3 hours, therefore this preparation must
be completed within this time limit, otherwise conjugation between the malPEG-DSPE and R24
antibody is not guaranteed.
2.2 Preparation of (Untargeted and Targeted) siRNA CCLs
2.2.1 Chemicals, Reagents, and Materials
The chemicals and materials from the preparation of siRNA CCLs are identical to those
from the empty CCLs, with the single addition of the lipid 1,2-dioleoyl-3-trimethylammoniumpropane (DOTAP), also purchased from Avanti Lipids Inc. (Alabaster, AL).
31 2.2.2 Preparation of the anti-GD3 R24 Antibody
The antibody preparation to the anti-GD3 R24 antibody used to create targeted siRNA
CCLs is identical to the preparation of the empty CCLs.
2.2.3 Sequence and Modifications of siRNA
The siRNA encapsulated into the liposome core was specific for a short nucleotide
sequence in the coding region of the Mcl-1 mRNA. The siRNA was purchased from RNA-TEC
(Belgium), presently IDT. The 25-nucleotide sequence of the sense strand was 5’CCCGCCGAAUUCAUUAAUUUACUG*T-3’. Modifications to the sense strand of Mcl-1
siRNA included a 3’ thiol group found on the terminal thymidine and a phosphorothioate bond
marked by the ‘*’ between the GT bases at the 3’end. Furthermore, these GT nucleic acids were
modified to chimeric DNA bases, while the remainder of the sequence was composed of RNA
bases. The sequence of the antisense business strand was 5’ACAGUAAAUUAAUGAAUUCGGCGGG*U*A -3’. The 27-nucleotide antisense was
composed of RNA bases that contained two phosphorothioate bonds, marked by the ‘*’ between
the GU and UA bases at the 3’ end. The RNA must be completely desalted and HPLC purified
when purchased. Upon delivery, it was resuspended in RNase-free water for storage at -20°C.
The molecular weight of this duplex is 16820.4g/mol.
2.2.4 siRNA CCL Preparation Protocol
In a 15mL Kimble Chase glass culture tube (Sigma), 800µg of Mcl-1 siRNA and 200µg
of Tex 613™ DS Transfection Control (Integrated DNA Technologies) RNA was solvated in
RNase-free water to bring its final volume to 1mL. To acquire a 1:1 positive to negative charge,
an amount equivalent in charge of DOTAP was added. Chloroform was added to the DOTAP to
bring its volume to 500µL. 1000µL of methanol was added to the RNA-DOTAP mixture. With
32 intermittent vortexing, methanol (Sigma) was pipetted into the solution, 100µL at a time, until
the milky white solution became translucent. The purpose of the methanol was to create a
monophasic solution in order for the siRNA and DOTAP to complex with one another. On
average, the total volume of methanol added was 1.8mL. Once monophasic, the DOTAP-siRNA
solution was incubated, undisturbed, for 30 minutes at room temperature. Post-incubation,
500µL RNase-free water and 500µL chloroform was added to revert the mixture into a biphasic
solution, where the DOTAP-siRNA complexes would concentrate in the chloroform. The
solution was vortexed and was left to sit for 2-3 minutes where a 2-phase separation began.
When separation was noticeable, the mixture was centrifuged with the Thermo Scientific Sorvall
Legend T Plus centrifuge (Waltham, MA) for 7 minutes at 800 x g. After centrifugation, a
distinct phase separation could be seen. The upper aqueous phase was removed and transferred
to a scintillation vial for RNA analysis, while the bottom organic phase remained in the Kimble
Chase tube. The remainder of this preparation follows the protocol to create empty liposomes,
including the addition of neutral lipids, sonication, rotary evaporation, extrusion and
chromatography. The untargeted siRNA CCL preparation is illustrated in Figure 2-1. Targeted
siRNA CCLs also had the addition of the targeting anti-GD3 R24 antibody, while untargeted
siRNA CCLs had an equal volume of HBS buffer added as an untargeted control.
2.3 Preparation of (Untargeted and Targeted) siRNA CCLs with Micelles
2.3.1 Chemicals, Reagents, and Materials
There are no new chemicals, reagents or materials in this preparation.
2.3.2 Preparation of the anti-GD3 R24 Antibody
The antibody preparation of the anti-GD3 R24 antibody used to create targeted siRNA
CCLs with micelles is identical to the preparation of the empty CCLs.
33 Figure 2-1. Preparation of the untargeted siRNA CCLs (adapted from Stuart & Allen, 2000)
The preparation begins with a biphasic mixture represented by the upper water aqueous phase
containing the siRNA and the bottom organic chloroform phase containing the cationic lipid (A).
When methanol is added, the mixture becomes monophasic, allowing for non-covalent
interactions between the negatively charged siRNA and positively charged cationic lipid to take
place (B). Following incubation and centrifugation, the mixture returns to being biphasic,
however, the siRNA molecules are pulled down to the organic phase by the cationic lipid (C).
Neutral lipids and water are added and the mixture is sonicated to begin forming spherical lipid
structures (D). Rotary evaporation removes the chloroform and transforms the mixture into a gel
phase (E). By adding buffer, the spherical lipid structures invert to encapsulate an aqueous core
to form liposomes (F). The resulting liposome mixture undergoes extrusion and purification via
chromatography before storage at 4°C.
34 2.3.3 Sequence and Modifications of siRNA
The sequence and modification of the Mcl-1 siRNA used to create targeted siRNA CCLs
with micelles is identical to those from the preparation of siRNA CCLs.
2.3.4 siRNA CCL Preparation Protocol with Micelles
Micelles are small lipid vesicles that are similar to liposomes but much smaller in size.
First, targeted or untargeted micelles were created, which were later fused with siRNAencapsulated liposomes. The preparation of micelles began with the addition of the lipids mPEGDSPE and malPEG-DSPE at a molar ratio of 0.8:0.2 – values that were relative to the neutral
lipids of the liposome preparation. These lipids were completely dried by rotary evaporation,
after which 1mL of HBS buffer was added to solvate the dried lipid film. The lipid resuspension
was divided into 2 aliquots. Exactly 0.003µmol (0.450mg) of antibody was dissolved in 500µL
HBS buffer (degassed) and immediately added to the first aliquot (targeted micelles). This
addition resulted in a 1:10 antibody to malPEG-DSPE molar ratio, which ensured conjugation
between the two functional groups. 500µL HBS buffer was added to the second aliquot of
micelles (untargeted micelles). These final mixtures were incubated at room temperature and
rocked overnight. The liposomes were prepared similar to those of the siRNA CCLs. The
changes to this procedure include the exclusion of malPEG-DSPE in the addition of neutral
lipids, the balancing of molar ratios between the neutral lipids DSPE:cholesterol:mPEG-DSPE to
40:20:1.2 for an approximate total of 22µmol lipids, and the halting of the protocol after size
exclusion chromatography, one step prior to the addition of antibody. When the liposomes had
been prepared and the overnight incubation of micelles was passed, these were added together
and allowed to incubate at 65°C for 1 hour to facilitate fusion of the micelles to the liposomes.
Following the fusion, the CCLs were purified in another G50 Sephadex column to remove all
35 unbound antibody and to fully exchange the solvent to HBS buffer. The CCLs were stored at 4°C
until analysis or treatment use.
2.4 Preparation of (Untargeted and Targeted) siRNA CCLs with Micelles, fused at 37°C
Targeted CCLs did not demonstrate ligand-specific binding, thus the fusion protocol was
subjected to change. The preparation was similar to the siRNA CCL preparation protocol with
micelles, with the sole change of the incubation conditions for the micelles and liposomes from 1
hour incubation at 65°C to 37°C overnight, before being subjected to purification via
chromatography.
2.5 Analysis of CCLs: Spectrometry and RNA Quantitation
Spectrometry was performed to quantify the initial amount of RNA and the amount of
RNA found within a CCL solution. Permission was granted from the Yang lab to use their BioRad SmartSpecTMPlus spectrophotometer (Hercules, CA), which measured the optical density
(OD) of the sample solution. The CCLs were diluted 1 to 100 in methanol to take the reading.
Methanol was used as a control solvent because it breaks the spherical CCL structure, allowing
for RNA release and proper quantification. OD readings were measured at λ= 260nm (260nm
wavelength) and RNA concentrations were calculated with the conversation factor A260 1 =
40µg/mL. Readings were always performed in triplicates.
2.6 Analysis of CCLs: Denaturing Urea Polyacrylamide Gel
RNA integrity was confirmed through denaturing gel electrophoresis of RNA using ureapolyacrylamide gels. Due to the small size of the siRNAs used in these experiments, 27 base pair
antisense and 25 base pair passenger, these gels were used to replace the conventional agarose
gels to visualize a better distinction between these two bands. To create the 15% gel, 6.3g urea
was first mixed and dissolved in 7.5mL of acrylamide, 1.5mL of 10X Tris-buffered EDTA
36 (TBE) (Sigma; Fluka BioChemika, Switzerland) and 1.5mL of RNase-free water. Immediately
following the addition of 50µL of 10% ammonium persulfate (APS) and 10µL
tetramethylethylenediamine (TEMED), the gel was loaded and solidified into a vertical gel
electrophoresis apparatus. To prepare the samples, each sample received a volumetric ratio of 1:2
of CCLs to RNA sample loading buffer, including the 0.1-1.0kb RNA ladder. Before loading the
sample into the solidified gel, 1X TBE was prepared as the running buffer and the RNA mixtures
were heated to 65°C for 10 minutes in order to denature the RNA secondary structure. Gels were
run at approximately 120 volts for 2 hours. Reagents used in the denaturing urea polyacrylamide
gel were acquired from Sigma-Aldrich unless noted otherwise. These gels were visualized under
transilluminescent UV light using the Alpha Innotech FluorChem™ 8900 (Santa Clara, CA) with
the program, AlphaEaseFC Software™, version 3.2.3 (Alpha Innotech / Cell Biosciences).
2.7 Analysis of CCLs: Dynamic Light Scattering
The dynamic light scattering analyzer DynaPro™ and Temperature Controlled
Microsampler (ProteinSolutions, presently Wyatt; Santa Barbara; CA) were used to measure the
final diameters of the CCLs. Using the software, Dynamics v5.25.44, 20µL of sample was
loaded into the cuvette and inserted into the machine for analysis. For consistent readings, the
quantity of particles was established, to be approximately 5000. Subsequently, the duration of the
particle size analysis was left for 20 seconds before the average particle diameter was taken. If
the 20µL sample included more than 8000 particles (particle overload), 20µL of HBS buffer was
added to the cuvette, and the analysis was restarted. For accurate measurements, the program
settings were fixed at Solvent: Aqueous Buffer, MW model: Ave. 149-370kDa, and Hardware
Settings Sensitivity: 100%.
37 2.8 Analysis of CCLs: NanoDrop and Antibody Quantitation
The Thermo Scientific NanoDrop Spectrophotometer ND-1000 (Wilmingdon, DE) was
used to quantify the initial antibody and thiolated antibody concentrations. Samples were always
well mixed before being loaded onto the lower pedestal. 2µL of antibody was used to ensure a
proper column formation within the apparatus and consistent concentration readings. The
downloaded operating software used was NanoDrop v3.3.0. Protein A280 was selected as
molecule type, with sample type changed to IgG to account for its molecular weight. The
spectrophotometer was cleaned and maintained as suggested by the NanoDrop 1000 user
manual.
2.9 Analysis of CCLs: CBQCA Assay
The CBQCA protein quantitation kit from Molecular Probes (Eugene, Oregon) was used
to measure the amount of protein bound to samples of micelles and-or liposomes through
fluorescence. This was possible as the ATTO-TAGTM CBQCA (3-(4-carboxybenzoyl)quinoline2-carboxaldehyde) reagent reacted with primary amines, in the presence of cyanide, to form
fluorescent conjugates. Using a 96-well plate, each control and sample well had 10µL of 5mM
CBQCA-dimethyl sulfoxide (DMSO) solution and 5µL of 20mM potassium cyanide (KCN)
added, with 1.5µL 10% Triton-X used to improve the sensitivity of the assay. The thiolated
antibody control and CCL-micelle samples were allowed to incubate for approximately 90
minutes at room temperature with rocking. The samples were then read in the BioTek uQuant
Microplate Spectrophotometer (Winooski, VT) fluorescence microplate reader and analyzed with
the application SoftMaxPro version 5.3 (MDS Analytical Technologies (US) Inc.; Sunnyvale,
CA).
38 2.10 Analysis of CCLs: Concentrating CCLs
When the concentration of CCLs to be used for in vitro or in vivo treatments was too low,
the CCL solution was concentrated using CentrisartI (Satorius Stedim Biotech, Goettingen,
Germany). If the concentration was too low, the chance of cell toxicity and contamination in
vitro increased greatly. For use, 2mL of the CCL solution was poured into the outer centrifuge
tube and the inner floater tube was reinserted to slide down and sit at the surface of the sample.
Capping the tubes, the samples were spun at 2500 x g for 5 minutes at a time until the desired
CCL solution volume was reached. At this time, the RNA concentration was remeasured by
spectrometry.
2.11 Melanoma Cell Lines and Culture Conditions
The in vitro studies employed the use of the well characterized NCI60 melanoma cell
lines, Malme-3M, SK-MEL-2, SK-MEL-5, SK-MEL-28 and LOX-IMVI, with the additional cell
line MeWo. All cells lines were purchased from ATCC (Manassas, VA). This research
predominantly used Malme-3M cells and LOX-IMVI cells. Malme-3M cells were used because
of their high GD3 expression and relatively low Mcl-1 expression. LOX-IMVI cells were used
because of their high Mcl-1 expression. All cells were maintained at 37°C with a 5% CO2 in the
Forma Series II Water Jacketed CO2 Incubator (Thermo Electron Corporation; Marietta, OH).
They were cultured in HyClone RPMI 1640 media (HyClone Laboratories Inc., Thermo
Scientific; Logan, Utah) with 10% fetal bovine serum (FBS) (Sigma) and 1% penicillinstreptomycin antibiotics (P-S) (Sigma). When confluency reached 95%, cells were washed with
phosphate-buffered saline (PBS), and split 1:4 in P100 plates.
39 2.12 In vitro Mcl-1 Silencing by CCLs
For in vitro experiments where Mcl-1 knockdown were analyzed, 250,000 cells were
plated into a P60 plate, 24 hours prior to transfection. This was done to allow confluency to
reach 60% by the time of transfection. Media used during this preparation and at the time of
transfection did not contain FBS or P-S antibiotics in order to eliminate any interference that
could prevent CCL binding. Although treatment concentrations for each study varied, the
siRNA-lipofectamine control was kept constant at 10nM concentrations for all experiments.
Treatment conditions were added drop-wise to their respective plates and incubated for 4 hours
before the cells were washed and fresh media was added (10% FBS and 1% P-S antibiotics).
Cells were harvested at a 48 hour time point, where they were washed, centrifuged, and the
resulting cell pellet was stored at -80°C until further analysis.
2.13 In vitro analysis by SDS-PAGE and Western blotting
Cells were removed from -80°C storage and solubilised by lysis buffer containing a
protease inhibitor cocktail (Roche Diagnostics; Mannheim, Germany). Protein content was
determined using the Bio-Rad protein quantitation kit (Bio-Rad) with the BioTek uQuant
Microplate Spectrophotometer (BioTek) and analyzed by the software, Gen5 version 1.02.8
(BioTek Instruments Inc.). 20-40µg of protein was resuspended in 4x loading buffer, composed
of filtered 10% SDS, 12% glycerol, 12.5% 2-β-mercaptoethanol, and 1% bromophenol blue, and
then denatured at 95°C for 10 minutes before being loaded into 12% polyacrylamide gel. After
electrophoresis, the proteins were transferred onto a polyvinylidene fluoride (PVDF) membrane
(Millipore Corporation; Billerica, Massachusetts) by electrophoresis at 100 volts for 1 hour at
4°C. Following amido black staining and a 4% non-fat milk block (1 hour at room temperature),
the membrane was incubated with primary antibody. The primary antibodies included α-rabbit
40 anti-Mcl-1 antibody (Cell Signalling, Beverly, MA) used at a 1 to 1000 dilution in 5% bovine
serum albumin (BSA) (Sigma) in Tris-Buffered Saline Tween-20 (TBST) overnight at 4°C, αrabbit anti-cleaved caspase 3 (Cell Signalling) used at a 1 to 1000 dilution in milk overnight at
4°C, and α-mouse anti-γ-tubulin (Sigma) used at a 1:10000 dilution in milk for 1 hour at room
temperature. A 1:1000 dilution of secondary antibody in milk was selected accordingly and was
rocked for 1 hour at room temperature. The blot was washed three times for 5 minutes with
TBST, consisting of 20mM Tris-HCl, 150mM NaCl, and 1% Tween -20, and was then exposed
onto Fuji medical x-ray film (FUJIFILM Corporation; Tokyo, Japan) using electrogenerated
chemical luminescence (ECL) solutions containing coumaric acid, luminal and hydrogen
peroxide. Densitometry analysis of the exposed film was performed using Quantity One version
4.6.3 (Bio-Rad) and percent expression and knockdown was calculated relative to the control
treatments. Reagents used in the Western blot analysis were acquired from Sigma unless noted
otherwise.
2.14 In vitro GD3-Specific Binding Assay for Flow Cytometry Analysis
Flow cytometry experiments were performed to verify the expression of GD3 on
melanoma cells lines and to determine whether antibody-targeted CCLs bind to melanoma cells.
For these studies, melanoma cells were washed, harvested, and aliquoted into 15mL plastic
conical tubes (1 tube per treatment group). Firstly, for the antibody study, 0.05mg of primary
R24 antibody was added to the tube and was allowed to incubate at 4°C for 45 minutes.
Following a PBS wash, 20µL of a 1:50 dilution of α-mouse AlexaFluor488 secondary antibody
(Invitrogen;Carlsbad, CA) was incubated with the cells for 30 minutes at 4°C. The anti-mouse
secondary antibody was the practical choice for the murine isotype R24 IgG3 antibody. Cells
were washed in PBS before being resuspended in 5mL Falcon tubes (Becton Dickinson and
41 Company; Franklin Lakes, NJ) for analysis. Between each step, the samples were centrifuged
with the Sorvall Legend T Plus centrifuge (Thermo Scientific; Waltham, MA) for 10 minutes at
1000rpm. After each spin, the supernatant was discarded and the pellet was resuspended for the
next treatment. In the studies involving CCLs, CCLs were used instead of primary antibody. The
rest of the protocol remained the same. At a volume equivalent to 0.05mg of antibody expression
on CCLs, as measured by the CBQCA assay, CCLs were incubated with the cells for 60 minutes
at 4°C before their PBS wash and subsequent treatments. Prepared samples were analyzed by the
Queen’s Cancer Research Institute using the Beckman Coulter EPICS Altra HSS flow cytometer
(Brea, CA).
2.15 In vitro Melanoma-Specific Binding Assay for Confocal and Fluorescent Microscopy
Analysis
Preparation for fluorescent microscopy analysis began by dropping Fisherbrand 18 circle
1D microscope cover slips (Thermo Fisher Scientific; Waltman, MA) into a 6-well plate and
plating 175,000 melanoma cells into each well. After an overnight incubation, the cells were at
approximately 70% confluency, ideal for microscopy. After a PBS wash, the cover slips
containing cells were fixed with 10% formalin for 10 minutes at room temperature. After fixing
the cells to the cover slip, and between each of the following steps, three 5 minute washes with
PBS were performed. The permeabilization step was completed with a 5 minute 0.3% Triton-X100 (Ricca Chemical Company; Arlington, TX) treatment. Unspecific binding of subsequent
targeting molecules was blocked by 10% goat serum in PBS for 30 minutes. R24 primary
antibody, used at a concentration of 1:200 in 1% goat serum, incubated with the cells for 1 hour
at room temperature, which was followed by α-mouse AlexaFluor488 secondary antibody in 1%
goat serum for 1 hour in darkness at room temperature. The cover slips were mounted onto
42 double frosted microscope slides (Thermo Fisher Scientific) using Vectashield Hard Set™
mounting medium with 4’-6-diamidino-2-phenylindole, more commonly known as DAPI
(Vector Laboratories, Inc.; Burlingame, CA). DAPI is a fluorescent stain that binds strongly to
DNA, employed as a visual control for the nuclei of cells. To observe binding of CCLs through
fluorescent microscopy, various CCL treatment concentrations were given at the ideal cell
confluency and were incubated for 24 hours before the cells were washed, fixed, permeabilized
and blocked with 4% non-fat milk (1 hour at room temperature),. Because the CCLs contained
antibody, secondary antibody was put on following the block. The CCL-treated cover slips were
then mounted and sealed with Vectashield Hard Set™ mounting medium with DAPI. These
slides were viewed at the Queen’s Cancer Research Institute using the Leica TCS SP2
multiphoton confocal laser scanning microscope. Images were taken using oil immersion under
various filters at 100x magnification and analyzed by the basic Leica operating software, Image
Pro Plus. Slides were also viewed under the Nikon Eclipse TE2000-U fluorescent microscope
(Tokyo, Japan) with permission from the Yang lab. Images were taken with various filters and
exposure times using the Imaging Retiga 2000R FAST 1394 camera (Surrey, BC) and the
program SimplePCI version 5.3.1.081004 (Hamamatsu; Sewickley, PA).
2.16 Melanoma Xenografts in SCID Mice
Female mice with severe combined immunodeficiency (SCID) were ordered from
Charles River (Wilmington, MA) and requested to be 3-4 weeks old upon arrival. The mice were
housed in the Queen’s Animal Care Facility, and kept in a hepa-filtered unit known as the
‘BioBubble’ (Rm 263), and given acidified water during care. Two types of melanoma xenograft
tumours were induced: LOX-IMVI and Malme-3M. To create the xenograft, melanoma cells
were cultured, washed, collected, and counted in order for a 100µL volume to be injected. To
43 prepare the mice, the hair on the right flank was shaved, and in that location, 100µL of
melanoma cells were injected subcutaneously with a 27 gauge ½ needle (Becton Dickinson and
Company) on a 1mL syringe (Becton Dickson & CO). To induce a LOX-IMVI tumour in the
mice, 2.5 million cells were injected. It took approximately 5 days for the tumour to reach an
approximate volume of 5mm3. To induce Malme-3M tumours in the mice, 10 million cells were
injected and grew for 8-10 weeks before becoming 5mm3 in size.
2.17 In vivo CCL Treatment and Harvest for Melanoma Xenografts in SCID Mice
When tumours reached a volume of approximately 5mm3, prepared CCLs were
administered intravenously by tail vein injection. Treatment groups for these experiments
included siRNA CCL concentrations of 5mg/kg and 1mg/kg, both referring to the amount of
siRNA relative to the weight of mouse. 5mg/kg and 1mg/kg of empty CCLs were injected in
volumes identical to those of the siRNA CCLs. A PBS control group was also used, using the
highest injected volume. Dosing schedules varied for each experiment. For the LOX-IMVI mice,
treatments were given on day 1 and day 2, and the mice were sacrificed on day 4. With the
Malme-3M mice, treatments were given on day 1, day 3, and day 5, and the mice were sacrificed
on day 7 for organ harvest and subsequent analyses. The tissues harvested from the LOX-IMVI
mice were the tumour, liver, lung and kidney. With the Malme-3M mice, the spleen was also
collected. Upon removal, the tissues were cut in half. Half of each tissue sample was flash frozen
(liquid nitrogen) and stored at -80°C for future biophotonics box and Western blot analysis. The
remaining half was prepared for analysis via microscopy.
2.18 Biophotonics Imaging through Biofluorescent Analysis (Biophotonics Box)
The flash frozen halves of tissue samples were brought to the Queen’s Cancer Research
Institute for biofluorescent imaging. This technique allows for imaging of the whole, undigested
44 tissue sample, in order to determine the localization of fluorescent siRNA CCLs. The frozen
samples were thawed and all the tissues from a single mouse were imaged at once. To determine
whether the different treatment groups led to different localizations of fluorescent siRNA CCLs,
an image of a selected organ from all the mice was taken. This was repeated for all the organs
harvested; tumour, liver, lung, kidney, and if available, spleen. The images were captured by a
Hamamatsu B/W ORCA ER digital camera as part of the Pan-A-See-Ya Panorama imaging
system. For analysis, the relative fluorescence value (RFV) was taken of all tissue samples and
measured using constant settings. RFVs were adjusted for tumour size and differences in
fluorescence were gauged relative to the PBS control tissues. Analysis of the images captured
was completed by Quantity One version 4.6.3 (Bio-Rad) and the relative fluorescence value
(RFV) of each tissue was measured and calculated relative to the set control PBS-treated
samples.
2.19 In vivo analysis by SDS-PAGE and Western blotting
Following biophotonics imaging, the frozen tumour samples were homogenized with the
use of PowerGen 125 Homogenizer (Fisher Scientific Co Ltd.). Once the in vivo tumour samples
had been homogenized, they were then lysed and analyzed for protein quantitation. The
remainder of the analysis follows the in vitro analysis by SDS-PAGE gel electrophoresis and
Western blotting, described in section 2.13.
2.20 In vivo analysis by Confocal Microscopy
The tissues harvested from the LOX-IMVI mice were the tumour, liver, lung and kidney.
With the Malme-3M mice, the spleen was also collected. Upon removal, the tissues were cut in
half. While half of each tissue sample was prepared for biophotonics box and Western blot
analysis, the remaining half was prepared for as slides for analysis via microscopy. The slides of
45 the in vivo samples were first placed in fresh 4% paraformaldehyde for 2 hours at room
temperature and then transferred into 30% sucrose overnight at 4°C. The next day, samples were
frozen in Optimal Cutting Temperature (O.C.T.) compound (Sakura Finetek; Torrance, CA,
USA) within a cryomold (Sakura Finetek) and frozen at -80°C until use. These samples were
then cut into 10µm sections using the Leica CM1850 UV cryostat (Wetzlar, Germany).
Afterwards, a cover slip was placed on top to seal the tissue sample onto a slide after applying
10µL of DAPI (Fisher Scientific). These slides were analyzed under confocal microscopy.
Similar to in vitro samples in section 2.15, confocal microscopy analysis was performed at the
Queen’s Cancer Research Institute with the Leica TCS SP2 multiphoton confocal laser scanning
microscope. Images were taken using oil immersion under various filters at 20x, 63x and 100x
magnification. To analyze the images, the basic Leica operating software was used, Image Pro
Plus.
46 CHAPTER 3 – RESULTS
3.0 Can anti-GD3 R24 monoclonal antibody serve as a targeting molecule on CCLs for
delivery of Mcl-1 siRNA to GD3-expressing melanoma?
GD3 is a widely expressed cell surface ganglioside whose expression is largely limited to
cells of neuroectodermal origin147, 148. It is the major ganglioside of embryonic and
undifferentiated tissues, decreasing progressively during development154. Expressing at least
10,000 molecules of GD3 on melanocytes, the ganglioside has been shown to be a major
component of the cell membrane of melanomas147, 155-158. Also, a five- to ten-fold increase of
GD3 expression is observed following malignant transformation156, 159. The ganglioside exhibits
upregulation as melanoma progresses from the primary to metastatic stage, playing an important
role in proliferation, growth and metastasis160.
For the purpose of this study, large amounts of specific antibody directed against GD3
were required. Fortunately, the R24 antibody is available at clinical grade from the NCI in the
United States of America. The IgG3 mouse monoclonal antibody has been revealed to bind the
GD3-specific acetylated trisaccharide structure found in a terminal position of the molecule156.
Analysis of cultured cells and neoplastic tissues with the antibody showed that R24 reacts with
astrocytes, astrocytomas, melanocytes, and melanomas, cell types that originate from the neural
crest161-163. This antibody thus shows significant specificity for melanoma, a requirement for this
study.
Antibody-targeted liposomal treatments have demonstrated therapeutic effect on solid
tumours. Cationic liposomes targeted with a monoclonal antibody against the GD2
disialoganglioside, containing antisense DNA oligonucleotides specific for c-myb, elicited
selective inhibition of the proliferation with GD2-positive neuroblastoma and melanoma cells in
47 vitro150, 164. In the same study, MZ2-MEL xenograft tumours induced in nude mice were treated
with 2.5mg/kg of antisense DNA oligonucleotides encapsulated within antibody-targeted
liposomes, once a week for 4 weeks. Results from silencing the c-myc gene interrupted its
signalling pathways, activating p53 and inhibiting Bcl-2 proteins, leading to extensive tumour
cell apoptosis and consequently, exhibiting a significant reduction in tumour growth site. It is
clear in this study that antibody attachment to gangliosides leads to internalization and
subsequent internal effects. This provides proof that targeting a similar ganglioside can be used
to deliver a nucleic acid payload to melanoma cells.
3.1 Incorporating RNA into the CCLs (Integrity Verification)
Many methods have been developed to incorporate siRNA into liposomes, including the
passive encapsulation techniques of RNA hydration of the dried lipid film and reverse-phase
evaporation, but my method followed one that optimizes the negative-positive charge interaction
between siRNAs and the cationic lipid123. Amounts of RNA were measured from a sample of
preparation via spectrometry to give a total RNA reading (Figure 3-1). At the start of the
preparation, 1000µg of RNA was used. 94.4% of the initial RNA was still present following
analysis of the organic phase. In the organic phase, the RNA present had been noncovalently
complexed with the cationic lipid DOTAP via electrostatic interactions, whereas 5.6% of the
RNA was not complexed and removed with the aqueous phase. The addition of neutral lipids,
rotary evaporation, and extrusion further decreased RNA incorporation to 86.8% and 85.8%,
respectively. Following the final steps of the fusion between prepared liposomes with micelles,
and purification by chromatography, 78.5% of the initial RNA was measured within the final
CCL product.
48 Figure 3-1. High percentages of Mcl-1 siRNA was incorporated and encapsulated into the
CCLs throughout their preparation as quantified by spectrometry analysis
At various stages of the CCL preparation, a 10µL volume of sample was taken for A260
spectrometry analysis. The measured absorbance reading was converted to an RNA
concentration, which was multiplied by the samples’ total volume, to determine the total quantity
of siRNA that existed during that particular stage of the CCL preparation. Starting with 100%
RNA, 5.6% of this was lost at the organic phase, after the removal of the aqueous phase. As the
preparation continued, RNA incorporation measurements taken immediately following the rotary
evaporation and extrusion stages were 86.8% and 85.8%, respectively. At the end of the
preparation, 78.5% of the initial RNA still remained within the CCL sample. (n=4)
49 To verify that the RNA encapsulated within the liposomes were intact and undamaged,
the resulting CCLs from two separate preparations were run on a denaturing urea polyacrylamide
gel (Figure 3-2). A 0.1-1.0kb RNA ladder and the Mcl-1 siRNA control were loaded, serving as
markers for the antisense and sense strands of 27bp and 25bp. Two separate bands are seen on
the CCLs containing RNA, suggesting that the siRNA strands were indeed viable. Duplexed
siRNA was not found in empty targeted CCLs, as no bands were displayed.
3.2 Desired size of the CCLs
Much debate has centered on the optimal diameter and size of liposomal therapies that
would produce the most efficient targeting and delivery to tumours101. The initial CCLs that were
formed following rotary evaporation through my protocol fell between 200-300nm (data not
shown). After extrusion, targeted CCLs with encapsulated Mcl-1 siRNA were measured to be
approximately 85nm in size, with no readings greater than 100nm. Empty targeted CCLs,
without siRNA, had diameters between 50nm and 70nm in size. Found between these two
readings, the untargeted siRNA CCLs averaged a liposome size of 75nm (Figure 3-3). Generally,
the size of siRNA-encapsulated CCLs measured to be at least 10nm larger than the empty CCLs
with their hollow aqueous core.
3.3 Presence of protein on the surface of CCLs
The CBQCA assay was used to confirm the presence of antibody on the surface of CCLs
(Figure 3-4). Samples were taken from CCLs before their fusion with micelles, from the targeted
and untargeted micelles, and from the final purified CCLs (targeted and untargeted). Prior to
fusion with micelles, the extruded and column-purified CCLs showed negligible protein content
on its exterior, as expected. Also, no protein was detected on the untargeted micelle sample.
Upon fusion of these two, the untargeted CCLs gave no protein reading, while the targeted
50 Figure 3-2. Mcl-1 siRNA remained intact through the CCL preparation as analyzed via
denaturing urea polyacrylamide gel electrophoresis
Final purified CCLs were run on a denaturing urea polyacrylamide gel to confirm that the RNA
encapsulated within the CCLs remained undamaged. The 0.1kb-1.0kb RNA ladder was run in the
first lane, with pure Mcl-1 siRNA serving as a control in lane 2. CCLs that were created with
encapsulated siRNA (siRNA CCL) exhibited two distinct bands at 25bp and 27bp, representative
of the intact sense-antisense RNA duplex. With samples that were created as empty CCLs
(Empty CCL), lane 5 and 6, the RNA bands are not present. There were no significant
differences between the CCLs that were GD3-targeted (Ab) and untargeted (non) revealed in this
assay.
51 Figure 3-3. All CCLs were measured to be nanoparticles of less than 100nm in diameter as
sized by dynamic light scattering
An aliquot of final purified CCLs was taken for dynamic light scattering analysis to determine
their average diameter. These nanoparticles all were measured to be less than 100nm in size.
Targeted empty CCLs were smallest, averaging just less than 60nm in diameter. Untargeted
CCLs with encapsulated siRNA were approximately 75nm, and when a targeting moiety was
conjugated to its surface, its size increased by approximately 10nm, to 84.47nm. (n=5)
52 Figure 3-4. Antibody was present on targeted micelles and GD3-targeted CCLs as
measured by the CBQCA assay
Approximately 5µL was taken from these respective samples for protein quantitation analysis via
the CBQCA assay. Protein detected from this assay was synonymous to the presence of antibody
in the sample. The results from this assay show that all CCLs prior to the fusion with micelles
exhibited no surface-bound antibody. Accordingly, no protein was detected with the untargeted
micelles and untargeted CCLs. On the contrary, the initial amount of thiolated antibody was
great, nearly 2000µg per 5µL sample. Coupling the antibody to the micelles exhibited 1452.5µg
of protein present on its surface. Following fusion, the protein reading of the purified final
targeted CCL sample measured 871.31µg of antibody. (n=3)
53 micelles displayed nearly 1.5mg of antibody, averaging approximately 10.0nmoles of antibody
per 5µL sample. Although some antibody was lost in the fusion process, antibody-targeted CCLs
expressed approximately 900µg of protein, translatable to 6.0nmoles of antibody or 30 antibodies
per liposome.
3.4 Confirming the binding of R24 antibody to cell surface GD3 of melanoma
The anti-mouse R24 antibody received from the NCI is specific for ganglioside GD3155,
156, 161
. Before going ahead with further experiments, confirmation of GD3 expression on
melanoma cell lines was necessary. This was performed by flow cytometry and confocal
microscopy. Malme-3M displayed the greatest percent binding and fluorescent intensity at
99.97% and 655.2, respectively (Table 3-1). Fluorescent intensity is the fold-greater change in
absolute fluorescent value, as measured by the flow cytometer, compared to the blank PBS
sample. SK-MEL-28, MeWo and SK-MEL-5, and SK-MEL-2 exhibited the next highest
readings and LOX-IMVI cells displayed the lowest R24 binding at 31.43%. In total, over 18,000
cells were gated as representative of each melanoma cell line (data not shown).
These flow cytometry results were reflected in the fluorescence images captured by
confocal microscopy. The green fluorescent anti-mouse AlexaFluor488 was used as a secondary
antibody for the murine R24 primary antibody. Malme-3M and SK-MEL-28 displayed the
greatest fluorescence at their cell surfaces (Figure 3-5). The Malme-3M cells expressed GD3
greatest at its cell surface, with clusters of GD3 molecules located throughout its cytoplasm as
well. Fluorescently-marked GD3 outlined the SK-MEL-28 cells, with expression observed to be
evenly distributed at the cell surface. GD3 staining on MeWo was found at its cell surface and
scattered throughout the cytoplasm of cells. SK-MEL-5 displayed little GD3-related fluorescence
with no clear pattern. Expression of GD3 on the SK-MEL-2 cells seemed to outline a few cells,
54 Melanoma Cell Line
% Binding of R24 antibody
Fluorescent Intensity
Malme-3M
99.97
655.2
SK-MEL-28
99.95
330.8
MeWo
97.89
319.1
SK-MEL-5
97.15
291.6
SK-MEL-2
96.76
257.1
LOX-IMVI
31.43
53.3
Table 3-1. R24 binds multiple melanoma cell lines as verified by flow cytometry
As measured from the CBQCA assay, approximately 0.05mg anti-GD3 R24 monoclonal
antibody was incubated with the melanoma cell lines for 45 minutes in preparation for flow
cytometry analysis. After incubation with secondary antibody and multiple washes, the cells
were resuspended in PBS and analyzed for antibody binding via flow cytometry. Of the 6 cell
lines, Malme-3M displayed the greatest percent binding and fluorescent intensity at 99.97% and
655.2, respectively. Fluorescent intensity is the fold-greater change in absolute fluorescent value,
as measured by the flow cytometer, compared to a blank PBS sample. The following four
melanoma cell lines also exhibited binding over 95% with high fluorescent values: SK-MEL-28
presented the second highest values with 99.95% binding and an intensity of 330.8; MeWo
displaying 97.89% and 319.1 fluorescence; SK-MEL-5 at 97.15% and 291.6; and SK-MEL-2
showing 96.76% and 257.1. Dissimilar to the other melanoma cell lines, LOX-IMVI displayed
the low R24 binding, 31.43%, with fluorescent intensity of 53.5. (n=2)
55 Figure 3-5. Anti-GD3 R24 monoclonal antibody binds and verifies cell surface GD3
expression on the melanoma cell lines via confocal microscopy
Melanoma cells were cultured on a cover slip, which was prepared for immunofluorescent
confocal microscopy. The cells were fixed with 10% formalin, permeabilized with Triton-X-100,
blocked in 10% goat serum, and incubated for 45 minutes with 1:100 R24 primary antibody and
AlexaFluor488 secondary antibody, successively. Mounted with the nucleus-staining DAPI,
these slides were qualitatively imaged at 100x magnification. Cell surface expression is evident
in Malme-3M (A) and SK-MEL-28 (B), while MeWo (C), SK-MEL-5 (D) and SK-MEL-2 (E)
displayed expression of GD3 to a lesser extent. Minimal fluorescence was detected from the cell
line, LOX-IMVI (F), which was not unexpected considering the lesser GD3-targeted R24
binding as shown in previous flow cytometry results (Table 3-1).
56 however, similar to SK-MEL-5, this cell line displayed sparse expression of GD3. Lastly,
minimal green fluorescence, which correlated with little GD3 expression, was detected with the
LOX-IMVI cells.
3.5 Ameliorating the CCL-micelle fusion protocol
In order to conjugate the R24 antibody targeting moiety onto the exterior of CCLs with
greater efficiency, the original protocol required improvement. Upon the introduction of micelle
fusion, the protocol involved siRNA-encapsulated CCLs and targeted or untargeted micelles to
fuse at 65°C for 1 hour, as suggested by previous research123, 165. When these treatments were
given to Malme-3M cells in vitro, there was no difference observed in Mcl-1 protein knockdown
between the untargeted CCLs and antibody-targeted CCLs (data not shown). The assumption
was that the targeting antibody was being denatured due to the high fusion temperatures. This
thought led to the modification of the protocol to include an overnight fusion incubation at
physiological temperatures, 37°C. The resultant CCLs were incubated with Malme-3M cells and
flow cytometry was performed to determine percent binding of the CCLs to the melanoma cells
using both fusion conditions. CCLs fused at 65°C for an hour resulted in 1.7% of the cells
expressing fluorescence with near autofluorescent intensity of 23.6 (Figure 3-6). On the other
hand, GD3-targeted CCLs that were fused with micelles at 37°C overnight produced 98.7%
binding, a value similar to the binding efficiency of R24 antibody to melanoma cells illustrated
in Table 3-1. The fluorescent intensity measured was 73.4. It is evident that at a fusion
temperature of 37°C (overnight) binding of CCLs to the melanoma cell lines is highly efficient,
while the 65°C 1 hour incubation was not. More than 9,500 cells were gated in each treatment
group.
57 Figure 3-6. The 37°C overnight fusion of GD3-targeted micelles to Mcl-1 siRNAencapsulated CCLs retained antibody specificity for GD3, confirmed by flow cytometry
GD3-targeted micelles were fused with Mcl-1 siRNA-encapsulated CCLs at 65°C for 1 hour and
for 37°C overnight. Post-fusion, both GD3-targeted CCLs samples were purified and incubated
with Malme-3M cells for 1 hour in preparation for flow cytometry analysis. After subsequent
incubation with secondary antibody and washes, the cells were resuspended in PBS and
analyzed. The 65°C-fused GD3-targeted CCLs showed minimal binding to the melanoma cells,
1.7% with a near-autofluorescent intensity of 23.6 (A). In contrast, 98.7% of the gated Malme3M cells were detected to express binding of the 37°C-fused GD3-targeted CCLs (B). A
fluorescence reading of 73.4 was measured from the fluorescence-positive cells.
58 3.6 GD3-targeted CCLs bind melanoma cells while untargeted CCLs do not
To determine whether R24 antibody-targeted CCLs were binding melanoma cells, cells
were first treated with targeted and untargeted CCLs, in vitro, and incubated for an hour before
proceeding onto flow cytometry analysis. Results from this experiment were positive, as the
R24-antibody targeting GD3-targeted CCLs served a role similar to the primary antibodies for
the fluorescent anti-mouse AlexaFluor 488 secondary antibody. Each cell line was represented
by over 9,000 gated cells. 99.92% of the Malme-3M and SK-MEL-28 cells demonstrated binding
of CCLs (Table 3-2). MeWo showed 99.43% binding of GD3-targeted CCLs and SK-MEL-5
showed a little less, at 98.02%. Measured below 90% binding, 79% of the SK-MEL-2 cells
emitted fluorescence when gated through the flow cytometer. Only 8.60% of LOX-IMVI cells
displayed binding of both GD3-targeted CCLs. Generally, these results followed the trend of
R24 expression as found in Table 3-1. Percent binding of untargeted CCLs was very low. All
melanoma cell lines exhibited the baseline autofluorescence with less than 5% binding of
untargeted CCLs. To further support these positive results, binding of the CCLs via R24 should
be measured in competitive binding experiments.
3.7 GD3-targeted CCLs exhibit enhanced binding and internalization compared to untargeted
CCLs and decreases Mcl-1 expression in a dose dependent manner
To measure the ability of GD3-targeted CCLs and untargeted CCLs to bind and
internalize into melanoma cells, Malme-3M and LOX-IMVI were treated with GD3-targeted
CCLs or untargeted CCLs and fluorescent images were captured in vitro. The red fluorescence
observed had been encapsulated within CCLs, thus, its localization is representative of the Mcl-1
siRNA (Figure 3-7). Malme-3M demonstrated the great distinction between the GD3-targeted
CCL treatment and the untargeted CCL treatment. After a 24-hour incubation, the Mcl-1 siRNA
59 Melanoma Cell Line % Binding of GD3-targeted CCLs % Binding of untargeted CCLs
Malme-3M
99.92
4.56
SK-MEL-28
99.92
2.85
MeWo
99.43
2.51
SK-MEL-5
98.02
2.79
SK-MEL-2
79.00
2.33
LOX-IMVI
8.60
2.92
Table 3-2. GD3-targeted CCLs bind multiple melanoma cell lines as verified by flow
cytometry
As measured by the CBQCA assay, the equivalent to 0.05mg of anti-GD3 R24 monoclonal
antibody of GD3-targeted CCLs was incubated with the melanoma cell lines for 1 hour in
preparation for flow cytometry analysis. After secondary antibody incubation and multiple
washes, the cells were resuspended in PBS and analyzed by flow cytometry for CCL binding. Of
the 6 cell lines, Malme-3M and SK-MEL-28 demonstrated the greatest binding of targeted CCLs
with 99.92%. Closely following at 99.43% and 98.02% were the MeWo and SK-MEL-5
melanoma cell lines, respectively. SK-MEL-2 exhibited lesser binding of targeted CCLs at
79.00% and LOX-IMVI displayed the least with 8.60%. Percent binding of untargeted CCLs was
highest at 4.56% with Malme-3M cells, though still a relatively insignificant percent binding
value. The remaining cell lines displayed the baseline autofluorescent percent binding of this
analysis, exhibiting less than 3% binding of untargeted CCLs.
60 Figure 3-7. GD3-targeted Mcl-1 siRNA-encapsulated CCLs enhance the binding and
internalization of Mcl-1 siRNA to Malme-3M and LOX-IMVI cells through fluorescent
microscopy
Malme-3M and LOX-IMVI cells were cultured and treated with GD3-targeted or untargeted
CCLs for 4 hours before being washed and adding fresh media. Mcl-1 siRNA within the CCLs
was tagged with red fluorescence for visualization under the microscopy, so localization of red
fluorescence represents the localization of the siRNA-encapsulated CCLs. After 24 hours, the
cells were imaged at 20x magnification, captured with 10 second exposures. GD3-targeted CCLs
showed greatest binding and internalization in the high GD3-expressing Malme-3M cells (A)
relative to the fluorescence observed with untargeted CCLs (B). Comparing the two images
taken of the LOX-IMVI cells, GD3-targeted CCLs (C) and untargeted CCLs (D) show
approximately the same fluorescence within the their respective cells. However, an observable
conclusion would be that there is less red fluorescence in the background of the GD3-targeted
image as compared to the untargeted CCLs.
61 of the GD3-targeted CCLs is clearly found within the cells, with great contrast of colour between
the fixed cells and the darkened background. This contrast signifies the uptake of CCLs from the
surrounding treatment media. Untargeted CCL treatments of the Malme-3M cells only depicted a
small amount of localization of the red fluorescence within cells with a relatively high red
fluorescence in the background. Comparing the two images taken of the LOX-IMVI cells, GD3targeted CCLs and untargeted CCLs show approximately the same fluorescence. However, it is
apparent that there is less red fluorescence in the background of the GD3-targeted images as
compared to the untargeted CCLs. These images were taken at 20x magnification with an
exposure time of 10 seconds.
Mcl-1 siRNA encapsulated within GD3-targeted CCLs decreased Mcl-1 expression in a
dose dependent manner when Malme-3M cells were treated in vitro for 48 hours, as analyzed by
SDS-PAGE gel electrophoresis and Western blotting (Figure 3-8). Bands of γ-tubulin were used
as a loading control for the normalization of the Mcl-1 bands. Treatment groups were compared
to the untreated cells, which was standardized to zero Mcl-1 knockdown. The lipofectamine with
Mcl-1 siRNA was used as a positive control. Despite the relatively low knockdown compared to
the lipofectamine with siRNA treatment, a trend can be observed from this analysis suggesting
that Mcl-1 expression decreases in a GD3-targeted CCL dose dependent manner.
3.8 GD3-targeted CCLs specifically localize, internalize and repress Mcl-1 expression in LOXIMVI xenograft tumours in a dose dependent manner following systemic administration
In vivo experiments were undertaken to determine whether GD3-targeted CCLs could
repress Mcl-1 expression within mouse melanoma xenografts. Fluorescence from the collected
tissue samples was indicative of the encapsulated RNA within the CCLs as captured by the
biophotonics box (Figure 3-9). Measurements of fluorescence of tissue samples from the mice
62 Figure 3-8. GD3-targeted Mcl-1 siRNA-encapsulated CCLs decrease Mcl-1 expression in
Malme-3M melanoma cells in vitro in a dose dependent manner
A dose response experiment was performed on Malme-3M melanoma cells with GD3-targeted
CCLs, with encapsulated Mcl-1 siRNA. For 4 hours, CCLs of concentrations 1nM, 10nM, 50nM
and 100nM were left to incubate with the cells before being replaced with fresh media. After 48
hours, the cells were harvested and all fractions were normalized for equal protein concentration.
Equal protein amounts were subjected to SDS-PAGE gel electrophoresis under reducing
conditions on a 12% polyacrylamide gel. Western blotting was performed with antibodies
against Mcl-1 and loading control, γ-tubulin. Mcl-1 expression decreased in a dose dependent
manner. Although there was no knockdown observed at 1nM, 16% and 29% knockdown was
detected in the 10nM and 50nM concentrations, respectively, with a maximum knockdown of
37% at the 100nM concentration.
63 64 Figure 3-9. GD3-targeted Mcl-1 siRNA-encapsulated CCLs specifically localized to the in
vivo mouse LOX-IMVI xenograft tumour following systemic administration via tail vein
injection
Mouse LOX-IMVI xenograft tumours were created by subcutaneously injecting 2.5 million cells
into the right flank. LOX-IMVI tumours grew to 5mm3 in 5 days, which was when CCL
treatments were injected systemically through the tail vein on day 1 and day 2. The mice were
sacrificed on day 4 and their liver, lung, kidney and tumour were harvested and flash frozen for
qualitative bioluminescence analysis. Allowed to thaw in order to minimize background
fluorescence, the imaging settings were calibrated to capture the red fluorescence that would be
emitted by the red-tagged RNA and each RFV was adjusted for tissue size. Comparing the liver
samples, the CCL-treated tissue (A) displayed lesser fluorescence than the PBS-treated tissue
(B), with a calculated RFV of 0.862. The same trend was observed with the lung tissues, 0.797
(C and D). The fluorescence found in the CCL-treated kidney (E) exhibited slightly greater
fluorescence than the PBS-treated kidney (F), with a RFV ratio of 1.022. The RFV ratio between
the CCL-treated (G) and PBS-treated (H) tumour tissues was 2.618, illustrating that there is
significant localization of treatment to the tumour tissue with the 5mg/kg GD3-targeted CCL
treatment.
64 treated with 5mg/kg GD3-targeted CCLs were made relative to those from the tissues samples of
mice injected with PBS. Comparing the liver samples, the CCL-treated tissue displayed
approximately the same fluorescence as the PBS-treated tissue, with a calculated RFV ratio of
0.862. The same trend is applicable to the lung tissues as the RFV ratio was calculated to be
0.797. Again, the kidney shows a value significantly similar to the PBS control, indicating, little,
if any fluorescence. When comparing the tumours tissues between CCL treatment and PBS
treatment, a large distinction is observed with the 5mg/kg GD3-targeted CCL treatment tumour
tissue illustrating high amounts of fluorescence at the corners and ends of the tumour. The
calculated RFV ratio between the tumour samples was 2.618. From these images, it is evident
that the GD3-targeted CCLs did not localize to organs of degradation or excretion, but rather,
fluorescence from the CCLs was specifically delivered to the subcutaneous xenograft tumour
tissue.
In addition to the fluorescence being evaluated in the organ tissues (liver, lung, kidney),
the tumour tissues of each treatment group was compared (Figure 3-10). Made relative to the
tumour samples of the PBS-treated tumour, it is apparent that the 5mg/kg GD3-targeted CCLtreated mouse elicited the greatest fluorescence with a 2.618 fold increase of fluorescence as
represented by the RFV ratio. The tumour sample from the 1mg/kg GD3-targeted CCL treatment
group exhibited a 1.061 RFV value as compared to the PBS-treated control. The empty GD3targeted CCL treatments displayed slightly less fluorescence, with a RFV ratio of 0.935. The
RFV ratios were compared to the PBS-treated tumour sample, hence its RFV ratio was 1.
A closer look of these LOX-IMVI xenograft tumours was taken using fluorescent
confocal microscopy. The tumour tissue samples were prepared through fixing, preservation,
cutting from frozen blocks, and were mounted onto slides with nucleus-staining DAPI. The PBS-
65 RVF ratio
Figure 3-10. GD3-targeted Mcl-1 siRNA-encapsulated CCLs localized to the in vivo mouse
LOX-IMVI xenograft tumour following systemic administration via tail vein injection in a
dose dependent manner
Mouse LOX-IMVI xenograft tumours were created by subcutaneously injecting 2.5 million cells
into the right flank. LOX-IMVI tumours grew to 5mm3 in 5 days, which was when CCL
treatments were injected systemically through the tail vein on day 1 and day 2. The mice were
sacrificed on day 4 and their tumours were harvested and flash frozen for qualitative
bioluminescence analysis. Allowed to thaw in order to minimize background fluorescence, the
camera settings were calibrated to capture the red fluorescence that would be emitted by the redtagged RNA. Adjusted for volume, these tumour samples produced a RFV standardized to the
background darkness. By comparing the RFV of a tumour sample to the PBS-treated tumour
sample, the 5mg/kg GD3-targeted liposome treatment accumulated the greatest amount of red
fluorescence with a 2.618 fold increase (A). A RFV ratio of 1.061 represented slightly greater
fluorescence found in the 1mg/kg GD3-targeted liposome treatment (B). The empty GD3targeted liposomes displayed slightly less fluorescence, with a RFV ratio of 0.935 (C). The RFV
ratios were compared to the PBS-treated tumour sample, thus, the RFV ratio of this sample was 1
(D).
66 treated tumour did not show any red fluorescence, as expected (Figure 3-11). Empty CCLs that
were GD3-targeted, also did not show red fluorescence. On the contrary, tumours of the 5mg/kg
GD3-targeted CCLs treatment group displayed areas of red fluorescence embedded between
areas of non-red fluorescence. These images were captured at 20x and 100x magnification,
displaying a section of tumour tissue that had been treated with 5mg/kg GD3-targeted CCLs.
Due to the nature of laser confocal microscopy, only a single layer of the tumour tissue is
analyzed for fluorescence. Taking this into account, with red fluorescence surrounding bluestained nuclei, it can be concluded that the fluorescence is within the cell. Thus, GD3-targeted
CCLs have not only delivered their contents to the tumour site, this treatment has internalized
into cells, potentially initiating the RNAi mechanism for Mcl-1 downregulation.
Subcutaneous LOX-IMVI xenograft tumours were analyzed by SDS-PAGE gel
electrophoresis and Western blotting. Used to normalize the Mcl-1 bands, the γ-tubulin band was
used as a loading control (Figure 3-12). The average knockdown of the PBS-treated tumours was
used to compare the knockdown of all other tumour samples. The five tumours from the 5mg/kg
GD3-targeted CCL treatment group elicited Mcl-1 knockdown averaging 38.8%, with greatest
knockdown found in tumour #5 at 66%. Averaging 33.5% knockdown, the 1mg/kg GD3-targeted
CCL treatment showed consistent knockdown. In the tumours treated with the empty GD3targeted CCL, despite the large variance observed, Mcl-1 knockdown averaged to 25.67%.
3.9 GD3-targeted CCLs localize and induce Mcl-1 knockdown in Malme-3M xenograft
tumours following systematic administration
With successes observed in the LOX-IMVI xenografts, subcutaneous Malme-3M
xenograft tumours in SCID mice were induced. The rationale behind using Malme-3M cells was
because of its high cell surface expression of GD3, much greater than that of LOX-IMVI, as
67 Figure 3-11. Localization of the GD3-targeted Mcl-1 siRNA-encapsulated CCLs to the in
vivo mouse LOX-IMVI xenograft tumour following systemic administration via tail vein
injection, verified by confocal microscopy
Mouse LOX-IMVI xenograft tumours were created by subcutaneously injecting 2.5 million cells
into the right flank. LOX-IMVI tumours grew to 5mm3 in 5 days, which was when CCL
treatments were injected systemically through the tail vein on day 1 and day 2. The mice were
sacrificed on day 4 and their tumours were harvested. Half of the tumour tissue was fixed with
4% paraformaldehyde, preserved in 30% sucrose, frozen in O.C.T, and cut into 10µm slices to be
mounted with DAPI onto slides for qualitative imaging at 20x (A, B, and C) and 100x (D)
magnification. As expected, the treatment groups of PBS (A) and GD3-targeted empty liposomes
(B) did not display any red fluorescence within their tumours. On the other hand, the 5mg/kg
GD3-targeted liposome tumours (C and D) displayed consistent red fluorescence throughout,
however, it cannot yet be concluded that the siRNA has localized within the cytoplasm of the
cells to activate the RNAi mechanism.
68 Figure 3-12. In vivo, GD3-targeted Mcl-1 siRNA-encapsulated CCLs decrease Mcl-1
expression in mouse LOX-IMVI xenograft tumours following systemic administration via
tail vein injection in a dose dependent manner
Mouse LOX-IMVI xenograft tumours were created by subcutaneously injecting 2.5 million cells
into the right flank of SCID mice. LOX-IMVI tumours grew to 5mm3 in 5 days, which was when
CCL treatments were injected systemically through the tail vein on day 1 and day 2. The mice
were sacrificed on day 4 and their tumours were harvested. Half of the tumour sample was flash
frozen, homogenized, lysed and normalized for equal protein concentration. Equal protein
amounts were subjected to SDS-PAGE gel electrophoresis under reducing conditions on a 12%
polyacrylamide gel. Western blotting was performed with antibodies against Mcl-1 and loading
control, γ-tubulin. Percent knockdown of Mcl-1 was made relative to the average of the three
PBS control samples. Decrease in Mcl-1 protein was greatest when the values from the 5mg/kg
treatment group are averaged to 38.8%, with highest knockdown of 66%. This was followed by
the average of the 1mg/kg treatment group at 33.8%, with values ranging from 22% to 48%
knockdown. The empty liposome treatment averaged 25.67% knockdown. Within the empty
liposome values, 46% knockdown in the first tumour sample of this group lessens the
significance and validity of these results.
69 illustrated in Figure 3-5A. Tracking tumour growth for 10 weeks, tumours grew to 5mm3 in size.
The biofluorescence of tissues from the 5mg/kg GD3-targeted CCLs and control PBS treatment
groups were measured via the biophotonics box, as described in the methods and materials. The
liver tissue, lung tissue, kidney tissue and spleen tissue of both treatments (5mg/kg and PBS)
showed minimal delivery of the CCLs, as represented by red fluorescence (Figure 3-13). Lastly,
the tumour tissue samples of both treatment groups displayed a small amount of fluorescence.
The 5mg/kg GD3-targeted CCL-treated tumour showed slightly greater fluorescence throughout
the tissue, whereas the PBS-treated tumour displayed few localized bright spots of fluorescence.
The calculated RFV ratio between the tumour tissues was 1.195. Comparing the tissues within
the same mouse, there was no obvious accumulation anywhere, which was unexpected because
of the positive results observed in the LOX-IMVI mice (Figure 3-11).
After harvesting the organs of the mice, the other half of the tumour tissues were fixed,
preserved, cut from frozen blocks, and mounted onto slides with nuclear-staining DAPI (data not
shown). After thoroughly scanning the slides, no red fluorescence could be found in any of the
treatment groups as imaged at 100x magnification.
The flash frozen samples were analyzed by SDS-PAGE gel electrophoresis and Western
blotting. Protein expression was measured relative to the loading control γ-tubulin (Figure 3-14).
Mcl-1 expression was compared to the average knockdown observed between the three PBStreated tumours. The 5mg/kg GD3-targeted CCLs showed the greatest knockdown with two
samples producing 67% and 80% knockdown. Averaging 73.5%, this was a five times greater
decrease than the 16.25% average knockdown found in the 1mg/kg GD3-targeted CCL
treatment. The empty GD3-targeted CCL treatment, injected at the 5mg/kg volume, displayed an
average of 15.33% knockdown. However, the values averaged to produce this number were two
70 71 Figure 3-13. GD3-targeted Mcl-1 siRNA-encapsulated CCLs specifically localized to the in
vivo mouse Malme-3M xenograft tumour following systemic administration via tail vein
injection
Mouse Malme-3M xenograft tumours were created by subcutaneously injecting 10 million cells
into the right flank. Malme-3M tumours grew to 5mm3 in approximately 9 weeks, which was
when CCL treatments were injected systemically through the tail vein on day 1, day 3, and day 5.
The mice were sacrificed on day 7 and their liver, lung, kidney, spleen and tumour were
harvested and flash frozen for qualitative bioluminescence analysis. Allowed to thaw in order to
minimize background fluorescence, the imaging settings were calibrated to capture the red
fluorescence that would be emitted by the red-tagged RNA. Each RFV reading was measured
relative to the background darkness and adjusted to their respective tissue volumes. The CCLtreated liver sample (A) did not display greater fluorescence than the PBS-treated liver tissue
(B), with an RFV value of 0.618. Comparing the CCL-treated lung (C) to the PBS-treated lung
(D), there was slightly greater fluorescence found with the 5mg/kg GD3-targeted CCL treatment
(RFV of 1.048). A similar RFV ratio was observed between the CCL-treated kidney (E) and the
PBS-treated kidney (F), 1.036 and the CCL-treated spleen (G) and the PBS-treated spleen (H),
1.041. The RFV ratio of 1.195 is representative of the tumour treated with 5mg/kg GD3-targeted
liposomes (I), displaying slightly more fluorescence than the PBS-treated tumour (J).
71 Figure 3-14. In vivo, GD3-targeted Mcl-1 siRNA-encapsulated CCLs decrease Mcl-1
expression in mouse Malme-3M xenograft tumours following systemic administration via
tail vein injection in a dose dependent manner
Mouse Malme-3M xenograft tumours were created by subcutaneously injecting 10 million cells
into the right flank. Malme-3M tumours grew to 5mm3 in approximately 9 weeks, which was
when CCL treatments were injected systemically through the tail vein on day 1, day 3, and day 5.
The mice were sacrificed on day 7 and their tumours were harvested. Half of each tumour
sample was flash frozen, and eventually homogenized, lysed and normalized for equal protein
concentration. Equal protein amounts were subjected to SDS-PAGE gel electrophoresis under
reducing conditions on a 12% polyacrylamide gel. Western blotting was performed with
antibodies against Mcl-1 and loading control, γ-tubulin. Percent knockdown of Mcl-1 was made
relative to the average of the three PBS control samples. Decrease in Mcl-1 protein was greatest
in the 5mg/kg treatment group (average = 73.5%). Between the four samples of 1mg/kg GD3targeted CCL treatment group, an average of 16.25% was observed. The averages of 15.33% and
9.25% were observed in the empty liposome treatments of 5mg/kg and 1mg/kg, respectively.
However, in each of the empty CCL treatment groups, the maximum knockdown observed was
46% and 35%. Taking into account high knockdown values found within these negative controls,
the validity of the experiment may be questioned.
72 samples showing no Mcl-1 knockdown and a final sample displaying 46% knockdown.
Similarly, the empty GD3-targeted CCL with the 1mg/kg injected volume showed an average of
9.25% knockdown, also with two samples showing no Mcl-1 downregulation, and the two others
at 7% and 30%. Taking the entire blot into account, the overall trend of each treatment groups’
knockdown resulted better than anticipated, despite seeing minimal red fluorescence in the
biophotonics box and confocal microscopy results.
73 CHAPTER 4 – DISCUSSION
Delivery of siRNA is the greatest challenge of its use as small-molecule drugs to treat
disease, especially cancers. In light of this, my research has focussed on the development of drug
delivery systems (DDSs) to enhance siRNA delivery by employing a targeting moiety toward a
cell surface marker of melanoma cells, disialoganglioside GD3. To specifically target this
molecule, the anti-GD3 R24 monoclonal antibody was selected. Using this targeting moiety,
coated cationic liposomes (CCLs) were developed and evaluated for their efficacy in vitro and in
vivo.
4.1 Anti-GD3 R24 monoclonal antibody can serve as a targeting molecule on coated
cationic liposomes for delivery of Mcl-1 siRNA to GD3-expressing melanoma
In the following studies, I employed the R24 antibody to target the GD3 expression of
melanoma cell lines in vitro and in vivo using mouse xenograft tumours. Beyond the examination
and insight of the observed results, progress and challenges throughout the development and
optimization of my liposomal therapy will be explained with additional suggested approaches
and suggestions for future experiments.
4.1.1 Evaluation of RNA encapsulation within GD3-targeted CCLs
As shown in Figures 3-1 and 3-2, RNA remained intact within the CCL and remained
highly incorporated throughout the CCL procedure. The 78.5% incorporation of RNA fell
slightly short of the 80%-100% reported by Stuart and colleagues, developers of the
procedure123. Then again, my result was more efficient than the protocols presented by Leonetti,
Ropert, and Thierry in the past decade, producing only 3%, 10% and 50%-70% siRNA
incorporation, respectively166-169. The reason as to why the decrease in RNA incorporation
occurred prior to micelle-liposome fusion may have been because these steps included the
74 translocation of DOTAP-siRNA complexes to incorporate within the CCLs with the addition of
neutral lipids.
However, there was no absolute verification that the RNA had been incorporated within
the CCLs. There is a slight possibility that the RNA may have leaked out of the CCLs or had not
been encapsulated within CCLs, rather simply solubilised within the CCL solution even after
purification via chromatography. Two methods that could have confirmed the formation of
liposomes and encapsulation of RNA were mass spectrometry and electron microscopy.
Concerning mass spectrometry, analysis of CCLs with encapsulated RNA would have provided a
different mass-charge ratio than CCLs without RNA, thus revealing whether the DDS was
formed successfully. Regarding electron microscopy, heavy metal particles (lead atoms) are
commonly used and would be complexed to specific molecules of the CCL or to the lipids that
form the liposome or to the incorporated RNA. The molecules chosen to bind these dense
particles would appear dark under the electron microscope. For example, if the RNA is
complexed, then one would hope to see multiple distinct spheres of grey or black, representative
of RNA encapsulation within CCLs. Sparse display of grey would signify unincorporated RNA.
The other example would be to complex the dense particles to membrane-forming lipids,
resulting in spherical rings viewed under the microscope for successfully formed CCLs.
4.1.2 The Optimal Size of GD3-targeted CCLs
Liposomes have been extensively studied as DDSs to deliver siRNA to tumour sites.
Furthermore, the optimal size of these nanoparticles has been through much discussion. More
applicable to in vivo studies than in vitro studies, liposome size plays a critical role in their
ability to extravasate through leaky tumour blood vessel walls, contributing to drug accumulation
within the tumour microenvironments. With small diameters, liposomes are able to penetrate
75 deeper into the interstitial tissues of the tumour mass, to areas where blood vessels may not be
present. Additionally, it has been shown that smaller liposome sizes reduce their recognition by
complements and the MPS in the blood101, 170, 171. More specifically, liposomes with a diameter
less than 100nm have shown to circulate longer in the blood by avoiding hepatic and splenic
uptake and thus, may be favourable in terms of optimizing in vivo drug release101, 172, 173. Also,
smaller liposomes allow for unchallenged extravasation through the larger leaky vessels,
contributing to drug accumulation in the tumour microenvironment.
By performing multiple passes through the 200nm, 100nm and double 100nm
polycarbonate filters during extrusion, the CCLs become smaller in size. After 3-4 passes
through each filter, the CCLs were analyzed by dynamic light scattering (Figure 3-3). Although
GD3-targeted CCLs with encapsulated Mcl-1 siRNA presented the largest size (average =
84.47nm), all CCLs were measured to be less than 100nm, thus benefiting from all the
advantages previously mentioned.
Initially, dynamic light scattering was an unfavourable method for sizing my liposomes,
especially as mass spectrometry and electron microscopy would have been able to provide size
analysis, in addition to validating RNA encapsulation within CCLs. However, many other
liposomal studies have used the light scattering technique to determine nanoparticle size128, 174,
175
. This technique was simple and was not time consuming, thus subsequently, employing
analysis by dynamic light scattering has been effective in my work by providing specific values
for the liposome sizes by using a small aliquot of each sample.
4.1.3 R24 Antibody and the Development of GD3-targeted CCLs
Before conjugating the R24 antibody as a targeting moiety to the exterior of CCLs, its
capacity to bind to multiple cell lines was verified via flow cytometry and confocal microscopy.
76 GD3 expression profiles of melanoma cells lines could not be found, so in addition to the percent
of cells that had bound R24 antibody, the fluorescent intensity measured represents the relative
quantity of GD3 expressed (Table 3-1). Confocal microscopy was also employed to estimate the
expression and visualize the localization of GD3 on the melanoma cell lines (Figure 3-5). Each
cell line demonstrated cell surface expression of GD3 except for LOX-IMVI, a cell line that
previously showed to express fewer GD3 molecules via flow cytometry. These results aside, the
use of a different GD3-specific epitope or molecule to quantify and visualize GD3 expression
would eliminate bias generated from R24 use and further verify the trends observed between
melanoma cell lines and their respective GD3 expression.
After confirming that R24 antibody binds GD3-expressing cells, conjugating the
molecule to liposomes was the subsequent step. There were two familiar methods that could
have been used to couple the R24 antibody to the liposome exterior: the hydrazide coupling
method and the maleimide coupling method176. The hydrazide coupling method involves
coupling the antibody to the exposed PEG molecule through their fragment crystallizable (Fc)
region. The resulting reaction would produce consistent orientation of the conjugated antibodies,
with their epitope (antigen-binding region) exposed for target interaction. The other benefit from
this coupling method would be the reduced clearance by the immune system, as the Fc region
would not be exposed for complement binding. The major disadvantage of this linkage chemistry
is that it has not presented high reaction efficiency (approximately 17%)177. On the other hand,
the maleimide coupling method readily reacts with exposed thiol groups, but because antibodies
do not readily express thiols, random thiolation of the antibody may lead to random orientation
on the liposome, possibly inducing an Fc-mediated complement immune response. On a positive
77 note, as presented in many studies, the epitopes are left unaffected by thiolation and are still able
to bind their target ligands129, 176.
My studies employed the maleimide coupling method by thiolating the R24 antibody
with Traut’s reagent and conjugating the resultant molecule with the malPEG-DSPE, which
naturally incorporates itself into the liposome bilayer. Because preparation of the GD3-targeted
CCLs preceded the use of micelles, a 3-hour time limit was placed on these experiments
coinciding with untoward oxidation of maleimide. Oxidation of maleimide weakens its reactivity
with the thiolated antibody. When the protocol describing micelle fusion was introduced and
optimized, the 3 hour limit of preparation time for the CCLs was removed178. The post-insertion
technique of forming targeted micelles to fuse with the pre-formed siRNA-encapsulated
liposomes was performed at 65°C for an hour, as suggested by previous research165. Also, an
assay was discovered that measures the presence of protein in a CCL solution, the CBQCA
assay. By analyzing 5µL of each CCL preparation, the total amount of antibody was measured at
multiple stages (Figure 3-4).
While the pre-fusion CCLs, untargeted micelles and untargeted CCLs showed a
negligible mass of antibody, their targeted counterparts exhibited a decreasing trend of antibody,
an expected result as purification would rid the excess unbound antibody from the fully-formed
liposomal drug post-fusion. With 871.31µg of protein and a known amount of lipid found on
end-stage targeted CCLs, approximately 30 antibodies were expressed on the each formed
liposome123, 177.
To further confirm that thiolated R24 antibody was bound to the exterior of CCLs, flow
cytometry was performed using GD3-targeted CCLs in place of the R24 primary antibody (Table
3-2). Contrary to my expectations, negative results were obtained from this analysis (Figure 3-
78 6A), signifying that the GD3-targeted CCLs fused as 65°C for an hour were not specifically
binding melanoma cells. In search of why binding was not observed, the conditions at which
micelles were fused with the CCLs were questioned. Previously concluded, the optimal
incubation to facilitate maximum fusion of the micelles to the liposomes was shown to take place
after 1 hour at 60°C179. Leakage of molecules from within the micelles and liposomes were also
minimal with these incubation conditions. Inferred from the same study, fusion performed at
37°C for over 24 hours displayed near-identical results to other researchers179. For this reason,
incorporation of targeted and untargeted micelles with liposomes was incubated overnight at
37°C. This change prevented the degradation of the antibody epitope. After this modification to
the fusion protocol, melanoma-specific binding of GD3-targeted CCLs was observed at 98.7%
binding by flow cytometry (Figure 3-6B), comparable to those values observed from R24
antibody binding (Table 3-1).
An approach that could improve the melanoma-specific targeting molecule would be to
use the Fab region of the R24 antibody only, the region of the antibody specific for GD3. By not
including the Fc region, antibody-associated immune responses during in vivo trials would
lessen, allowing for reduced drug-associated adverse effects. However, steps toward their use in
my experiments have not been taken. Nevertheless, with the successful addition of melanomaspecific targeting ligands expressed on liposomal nanoparticles, the following experiments were
hopeful to show enhanced siRNA delivery and effectiveness in vitro and in vivo88.
4.1.4 In vitro Analysis of GD3-targeted and untargeted CCLs
Binding and internalization of the CCLs was visualized through fluorescent microscopy
in the Malme-3M and LOX-IMVI melanoma cell lines (Figure 3-7). These two cell lines were
chosen to be studied extensively because of the greatly overexpressed Mcl-1 in LOX-IMVI cells
79 and because of the highly expressed GD3 in Malme-3M cells. As expected, GD3-targeted
binding and internalization was high in the Malme-3M cell line and was minimal in the LOXIMVI cell line (Table 3-1). Means by which this microscopic analysis could have been improved
would have been to perform a time course experiment to illustrate localization of the red
fluorescence at 1, 2, 4, etc. hours post-transfection. Also, siRNA does not readily degrade within
cells when stabilized. Its silencing effects have been shown to last for up to 11 days80. Therefore,
if the red fluorescent protein could be tagged to the siRNA, I would expect to observe
fluorescence at late time points, demonstrating the continuous silencing effect of the RNA51.
A time course experiment demonstrating the Mcl-1 knockdown at selected CCL
concentrations would have also benefited my in vitro analyses. The GD3-targeted CCL dose
response experiment showed initial knockdown of 16% (10nM), which increased to 37%
(100nM) (Figure 3-8). The maximum knockdown observed was not as high as expected,
especially since the lipofectamine with siRNA control displayed a clear 81% knockdown.
However, the effectiveness of CCL treatments in vivo is greater than in vitro due to the longer
duration at which the drug is incubated within the model122. For this reason, my final
experiments involved studying the effects of GD3-targeted CCLs in mouse xenograft tissues.
4.1.5 In vivo Analysis of GD3-targeted CCLs in Mouse Xenograft Models
Mouse xenograft models were used to mimic the formation of metastatic sites. Xenograft
tumours were induced by injecting melanoma cells subcutaneously into SCID mice. The location
of the site for tumour growth serves as a good representation of melanoma metastases, favoured
over other xenograft methods, because it is the common area whereby melanoma cells would
extravasate, localize, and establish a microenvironment180. However, not a true representation of
metastatic melanoma, the xenograft tumours were allowed to grow to 5mm3 in size. In clinical
80 diagnoses, when melanoma presents itself at volumes greater than 1mm3, tumours are considered
thick, high risk melanoma24, 150. For my in vivo studies, using a larger tumour allowed a greater
number of analyses to be performed on a single mouse.
Dissimilar dosing schedules were used between the LOX-IMVI- and Malme-3M-induced
tumours. The mice with LOX-IMVI melanoma tumours were put on a dosing schedule that had
been shown to elicit in vivo RNAi-mediated silencing181. Treatments were administered on day 1
and 2, and the mice were sacrificed on day 4; 48 hours after their final injection181. Alternatively,
Malme-3M melanoma tumours followed a dosing schedule previously proposed by our lab.
Injections were performed on days 1, 3, and 5, sacrificing the mice on day 7. The tissues
harvested from the mice consisted of the liver, lung, kidney and tumour, with the addition of the
spleen in the Malme-3M in vivo experiment. Of these organ tissues analyzed via biofluorescence,
the greatest accumulation of fluorescence was observed in the tumour tissue from the LOXIMVI tumours (Figure 3-9G & Figure 3-10A). Even though these results were positive, the
preparation of samples for biofluorescent analysis possessed many variables that may have
skewed fluorescence measurements and subsequent results. The first was the possibility of
external light entering into the box during times of measurements. Light from a certain position
would brighten tissue samples in its proximity, making them seem to fluoresce. Also, tissue
samples that had not been completely thawed for readings would appear to fluoresce, as ice
refracted light to produce measureable emission. Furthermore, unless fluorescence found within
tissues of the mouse could be measured in real-time, the red fluorescence encapsulated within the
CCLs may have degraded as the experiment progressed. I suspect that fluorescence degradation
was evident in the Malme-3M tumours, as doses were not administered on consecutive days
prior to harvest and tumours were not harvested until day 7. The Malme-3M tumours were
81 anticipated to glow brightly with red fluorescence through microscopy. However, observing no
red fluorescence within the Malme-3M tumour was unexpected because these cells demonstrated
a much higher cell surface GD3 expression than the LOX-IMVI cells. Lastly, all organ samples
were halved, one section for fixing and the other for flash freezing. It is possible that the
fluorescence amassed in the other half of the organ sample. Concerning the results surrounding
the biofluorescence analysis, it is important to note that the RFV measured from the Malme-3M
tumour was only 1.195, while the LOX-IMVI tumour had a calculated RFV of 2.618. The
discrepancy observed between these two tumour types may not have been because there was
lesser targeting of the Malme-3M tumour, but rather the red fluorescent RNA had degraded
following the lengthy incubations within these in vivo models.
A similar development was observed in the images produced by confocal microscopy. No
fluorescence was detected within any of the Malme-3M tumour sample (data not shown).
Because the red fluorescence was not actually complexed with the Mcl-1 siRNA, it is likely that
the fluorescent protein covalently bonded to Tex 613™ DS transfection control RNA was
degraded. Thus, it is possible that targeting of the Malme-3M tumour did occur, but my
techniques were simply not able to visualize or quantify binding or internalization. In contrast,
abundant fluorescence was observed in the LOX-IMVI tumour samples treated with 5mg/kg
GD3-targeted CCLs (Figure 3-11D). Red fluorescence was observed to surround the blue DAPIstained nuclei of cells. From this, I can conclude that the RNA and non-fluorescent Mcl-1 siRNA
has been internalized into the cells, possibly activating RNAi-mediated silencing.
In both the LOX-IMVI and Malme-3M tumours, Mcl-1 silencing was measured to be
dose-dependent. Western blot analysis of the LOX-IMVI tumours illustrated that at 5mg/kg,
Mcl-1 knockdown was 38.8% (Figure 3-12). The Malme-3M tumours demonstrated an Mcl-1
82 knockdown average of 73.5% for the 5mg/kg dose (Figure 3-14). Although there were only two
samples representing the 5mg/kg GD3-targeted CCL treatment group in mice with Malme-3M
tumours, an average 73.5% decrease in Mcl-1 expression could be attributed to the make-up of
the tumour. The Malme-3M tumour grew to the 5mm3 size in 9 weeks. During this period of
development, the tumour likely became self-sustaining, creating a strong vasculature to nourish
its proliferating mass, a phenomenon not probable with LOX-IMVI tumours. Despite these
striking results and trends, the PBS-treated control tumours showed a range of bands which
appeared to have differences in the levels of Mcl-1 expression, depicted clearly in Figure 3-14.
Furthermore, the variance between the knockdown percentages was large in some samples. For
example, the empty GD3-targeted CCLs of the LOX-IMVI tumour samples had high variance
between values. They averaged 25.67% knockdown but ranged from 2%-46% knockdown.
Because the LOX-IMVI tumours grew to 5mm3 in diameter within 5 days, tumour necrosis may
have caused of the disparity between expression and knockdown of Mcl-1. To quantify the level
of necrosis and verify the melanoma characteristics of the tumour, immunohistochemistry should
have been performed on these samples in order to evaluate the expression of ganglioside GD3.
Microarrays could have also been run on each tumour sample. This technique would have
illustrated the possible heterogeneity of the LOX-IMVI and Malme-3M melanoma xenograft
tumours.
As expected, knockdown of Mcl-1 by CCLs was shown to be greater in vivo than in vitro.
Despite challenges pertaining to the techniques employed to analyze the effects of the GD3targeted CCLs, progress has been made. To advance from the preliminary ground work
demonstrated here, it is crucial to increase all samples sizes as they are presently low, and to
employ cell line and CCL controls in future in vivo experiments.
83 4.2 Future Directions
My future directions include the following:
1. Labelling neutral lipids with radioactive isotopes (tritium, 3H)
2. Tagging the red fluorescence to the encapsulated Mcl-1 siRNA
3. Verifying endosome escape via immunofluorescent confocal microscopy
4. Adding the fusogenic lipid, DOPE
5. Providing the combination therapy of small molecule inhibitor ABT-737 and Mcl-1
siRNA
Previous studies have used radioactivity to quantify the amount of lipid involved in
creating CCLs and to determine the localization of radioactive CCLs within an in vivo model182,
183
. Similarly, future experiments should employ the radioactive isotope, tritium (3-Hydrogen,
3H), in the form of the neutral lipid 3H-cholesterol. The 3H isotope would be selected because of
its low-grade radioactivity, making it safer to use than other isotopes, including 32-Phosphorus
and 14-Carbon184, 185. Incorporating 3H-cholesterol into the preparation would enable lipid
concentrations in the CCLs to be calculated. Additionally, measured radioactive counts-perminute readings of dissected organ tissue samples would provide information on the localization
of CCL treatments within a mouse model.
My current liposomal therapy involves the use of a red fluorescent molecule tagged to an
unmodified factory-scrambled RNA. This RNA is not protected by modifications and thus, most
prone to rapid intracellular degradation. To improve my experiments, future work should
conjugate the red fluorescent molecule directly to the encapsulated Mcl-1 siRNA to better
facilitate the determination of siRNA localization. Studies have demonstrated that liposomal
84 siRNA treatments can be better visualized in this way under fluorescent and confocal
microscopy, using tags such as Cy3 and EGFP181.
During my research, I was not able to visualize the localization of the drug complexes. I
had to rely on protein knockdown to assume my drug had been delivered to the cytoplasm of the
cell. Localization of molecules can be determined through the use of cell organelle-specific
antibodies. Drug delivery systems have been shown to internalize through the endosomal
pathway, where drug degradation ensues as endosomes mature to lysosomes186. To visualize
endosome escape, endosome- and lysosome-specific antibodies can be used in an
immunofluorescent confocal microscopy study, where overlay between the fluorescent images
(endosome and tagged RNA) would represent trapped CCLs. Many researchers have found
liposomes getting trapped within endosomes once internalized; hence the following approach
addresses this challenge.
In addition to the various types of lipids incorporated into liposomes (described earlier),
fusogenic lipids are another category in CCL development. Fusogenic lipids help facilitate the
internalization of the CCLs through the cellular membrane and promote endosome escape. After
endosomal escape, the CCLs localize in the cell cytoplasm and are able to initiate the RNAimediated silencing pathway. Dioleoylphosphalidylethanolanime (DOPE) has been previously
shown to serve as a proficient fusogenic lipid with its positively charged and conical 3dimensional structure187. However, liposomes involving DOPE as the predominant lipid have
shown to lead to cytotoxicity in vitro and limited success in vivo. Their use has caused severe
adverse effects187. In future experiments, adding a low molar amount of the fusogenic lipid,
DOPE, may improve fusogenicity and increase knockdown effects188. Our lab has performed
preliminary in vitro experiments incorporating DOPE into the chemical composition of the
85 CCLs. Results from these trials have shown consistently greater Mcl-1 knockdown compared to
both untargeted and empty CCL control treatments (data not shown). This provides promise in
the continual development of our liposomal therapy.
Lastly, it is important to note that Mcl-1 silencing alone is not enough to lead melanomas
to apoptosis. Our lab has previously shown that the small molecule inhibitor ABT-737, in
combination with Mcl-1 knockdown, represents a promising strategy for melanoma treatment1.
Consequently, transforming my liposomal therapy to enable the encapsulation of small molecule
inhibitors (ABT-737) along with gene-specific siRNA (Mcl-1) would produce a strong
apoptosis-inducing melanoma-specific drug.
4.3 Conclusions, Significance, and Clinical Relevance of this study
My research has employed the R24 antibody as a targeting moiety to deliver siRNA to
melanoma. Coated cationic liposomes were used as siRNA drug delivery systems and results of
their use have illustrated that these targeted delivery systems ameliorate binding, internalization
and initiation of RNAi-mediated gene silencing in vitro and in vivo in melanoma. The
conjugation of GD3-targeted R24 antibody to the chemically complex liposomes demonstrated
their prowess as a potential drug for therapeutic use. Focussing on the development of my
liposomal therapy, multiple steps were involved in order to produce a positive end-result. Firstly,
the challenge of incorporating siRNA into the aqueous core of the lipid sphere has been
overcome, achieving approximately 80% efficiency. Next, conjugation chemistry of a targeting
moiety to the exterior of the delivery system was successful, selecting the thiol-maleimide
chemistry to create a stable bond at consistently high reaction rates. Lastly, my research refined
the lipid composition and the process by which my final coated cationic liposomes were created,
establishing a strong foundation whereby later work may better initiate the RNAi pathway.
86 The ultimate goal of my research was to develop and optimize clinically viable therapies
for malignant melanoma. Once metastasized, malignant melanoma is essentially incurable, and
thus, bold new therapeutic approaches are urgently needed. Treatments that have the potential for
clinical use would be optimized as a one-time dose, disease-specific and disease-targeted drugs
with minimal side effects and minimal toxicity. My research project was to advance the
development of an adaptable ligand-targeted siRNA drug delivery system, specific for melanoma
markers. However, by selecting a targeting moiety specific to another cancer or disease allows
this research to be translatable into other fields of medicine. Additionally, the siRNA therapy
encapsulated within the liposome can be directed against other overexpressed oncogenes and
their respective mRNAs, sequestering their protein expression, thus cellular function. The
versatility of this drug delivery system is dependent on the thorough understanding and precise
execution of their creation, so investigations on their preparation have been completed and
optimized. I believe that liposomal therapy research will be a major forefront contributor in the
development of personalized therapies.
87 REFERENCES
1. Keuling, A. M. et al. RNA silencing of Mcl-1 enhances ABT-737-mediated apoptosis in
melanoma: role for a caspase-8-dependent pathway. PLoS One 4, e6651 (2009).
2. Chudnovsky, Y., Khavari, P. A. & Adams, A. E. Melanoma genetics and the development of
rational therapeutics. J. Clin. Invest. 115, 813-824 (2005).
3. Canadian Cancer Society's Steering Committee. Canadian Cancer Statistics 2009. (2009).
4. de Vries, E. & Coebergh, J. W. Melanoma incidence has risen in Europe. BMJ 331, 698
(2005).
5. American Cancer Society. Cancer Facts & Figures 2009. (2009).
6. Miller, A. J. & Mihm, M. C.,Jr. Melanoma. N. Engl. J. Med. 355, 51-65 (2006).
7. Gray-Schopfer, V., Wellbrock, C. & Marais, R. Melanoma biology and new targeted therapy.
Nature 445, 851-857 (2007).
8. Amaravadi, R. K. & Flaherty, K. T. Targeted therapy for metastatic melanoma. Clin. Adv.
Hematol. Oncol. 5, 386-394 (2007).
9. Balch, C. M. et al. Prognostic factors analysis of 17,600 melanoma patients: validation of the
American Joint Committee on Cancer melanoma staging system. J. Clin. Oncol. 19,
3622-3634 (2001).
10. Hanahan, D. & Weinberg, R. A. The hallmarks of cancer. Cell 100, 57-70 (2000).
11. Sporn, M. B. The war on cancer. Lancet 347, 1377-1381 (1996).
12. Barth, A., Wanek, L. A. & Morton, D. L. Prognostic factors in 1,521 melanoma patients with
distant metastases. J. Am. Coll. Surg. 181, 193-201 (1995).
13. Chin, L., Garraway, L. A. & Fisher, D. E. Malignant melanoma: genetics and therapeutics in
the genomic era. Genes Dev. 20, 2149-2182 (2006).
14. Gilchrest, B. A., Eller, M. S., Geller, A. C. & Yaar, M. The pathogenesis of melanoma
induced by ultraviolet radiation. N. Engl. J. Med. 340, 1341-1348 (1999).
15. Thompson, J. F., Scolyer, R. A. & Kefford, R. F. Cutaneous melanoma. Lancet 365, 687-701
(2005).
16. Tang, L. et al. Expression of apoptosis regulators in cutaneous malignant melanoma. Clin.
Cancer Res. 4, 1865-1871 (1998).
17. Hickman, J. A. Apoptosis induced by anticancer drugs. Cancer Metastasis Rev. 11, 121-139
(1992).
18. Kerr, J. F., Winterford, C. M. & Harmon, B. V. Apoptosis. Its significance in cancer and
cancer therapy. Cancer 73, 2013-2026 (1994).
19. Fisher, D. E. Apoptosis in cancer therapy: crossing the threshold. Cell 78, 539-542 (1994).
20. Harmon, B. V., Takano, Y. S., Winterford, C. M. & Gobe, G. C. The role of apoptosis in the
response of cells and tumours to mild hyperthermia. Int. J. Radiat. Biol. 59, 489-501
(1991).
21. Staunton, M. J. & Gaffney, E. F. Tumor type is a determinant of susceptibility to apoptosis.
Am. J. Clin. Pathol. 103, 300-307 (1995).
22. Mooney, E. E., Ruis Peris, J. M., O'Neill, A. & Sweeney, E. C. Apoptotic and mitotic indices
in malignant melanoma and basal cell carcinoma. J. Clin. Pathol. 48, 242-244 (1995).
23. Opferman, J. T. et al. Obligate role of anti-apoptotic MCL-1 in the survival of hematopoietic
stem cells. Science 307, 1101-1104 (2005).
88 24. Zhuang, L. et al. Mcl-1, Bcl-XL and Stat3 expression are associated with progression of
melanoma whereas Bcl-2, AP-2 and MITF levels decrease during progression of
melanoma. Mod. Pathol. 20, 416-426 (2007).
25. Wuilleme-Toumi, S. et al. Mcl-1 is overexpressed in multiple myeloma and associated with
relapse and shorter survival. Leukemia 19, 1248-1252 (2005).
26. Song, L., Coppola, D., Livingston, S., Cress, D. & Haura, E. B. Mcl-1 regulates survival and
sensitivity to diverse apoptotic stimuli in human non-small cell lung cancer cells. Cancer.
Biol. Ther. 4, 267-276 (2005).
27. Sano, M. et al. Overexpression of anti-apoptotic Mcl-1 in testicular germ cell tumours.
Histopathology 46, 532-539 (2005).
28. O'Driscoll, L. et al. Expression and prognostic relevance of Mcl-1 in breast cancer.
Anticancer Res. 24, 473-482 (2004).
29. Clark, W. H.,Jr et al. A study of tumor progression: the precursor lesions of superficial
spreading and nodular melanoma. Hum. Pathol. 15, 1147-1165 (1984).
30. Braig, M. & Schmitt, C. A. Oncogene-induced senescence: putting the brakes on tumor
development. Cancer Res. 66, 2881-2884 (2006).
31. Nyormoi, O. & Bar-Eli, M. Transcriptional regulation of metastasis-related genes in human
melanoma. Clin. Exp. Metastasis 20, 251-263 (2003).
32. Klatte, T., Rao, J. Y., Ribas, A. & Pantuck, A. J. Metastatic melanoma to the kidney
presenting with renal vein tumor thrombus. Urology 69, 982.e7-982.e9 (2007).
33. Stein, J. A. & Brownell, I. Treatment approaches for advanced cutaneous melanoma. J.
Drugs Dermatol. 7, 175-179 (2008).
34. Queirolo, P. & Acquati, M. Targeted therapies in melanoma. Cancer Treat. Rev. 32, 524-531
(2006).
35. Eggermont, A. M. & Kirkwood, J. M. Re-evaluating the role of dacarbazine in metastatic
melanoma: what have we learned in 30 years? Eur. J. Cancer 40, 1825-1836 (2004).
36. Avril, M. F. et al. Fotemustine compared with dacarbazine in patients with disseminated
malignant melanoma: a phase III study. J. Clin. Oncol. 22, 1118-1125 (2004).
37. Middleton, M. R. et al. Randomized phase III study of temozolomide versus dacarbazine in
the treatment of patients with advanced metastatic malignant melanoma. J. Clin. Oncol.
18, 158-166 (2000).
38. Chapman, P. B. et al. Phase III multicenter randomized trial of the Dartmouth regimen versus
dacarbazine in patients with metastatic melanoma. J. Clin. Oncol. 17, 2745-2751 (1999).
39. Eton, O. et al. Sequential biochemotherapy versus chemotherapy for metastatic melanoma:
results from a phase III randomized trial. J. Clin. Oncol. 20, 2045-2052 (2002).
40. Huncharek, M., Caubet, J. F. & McGarry, R. Single-agent DTIC versus combination
chemotherapy with or without immunotherapy in metastatic melanoma: a meta-analysis
of 3273 patients from 20 randomized trials. Melanoma Res. 11, 75-81 (2001).
41. Hamamura, K. et al. Ganglioside GD3 promotes cell growth and invasion through p130Cas
and paxillin in malignant melanoma cells. Proc. Natl. Acad. Sci. U. S. A. 102, 1104111046 (2005).
42. Furukawa, K., Arita, Y., Satomi, N., Eisinger, M. & Lloyd, K. O. Tumor necrosis factor
enhances GD3 ganglioside expression in cultured human melanocytes. Arch. Biochem.
Biophys. 281, 70-75 (1990).
43. Bedogni, B. et al. The hypoxic microenvironment of the skin contributes to Akt-mediated
melanocyte transformation. Cancer. Cell. 8, 443-454 (2005).
89 44. Ketcham, A. S. & Balch, C. M. Classification and staging system. In Cutaneous melanoma:
clinical management and treatment results worldwide, Balch CM and Milton GW (eds). ,
55-62 (1985).
45. Pagnan, G. et al. GD2-mediated melanoma cell targeting and cytotoxicity of liposomeentrapped fenretinide. Int. J. Cancer 81, 268-274 (1999).
46. Eberle, J., Kurbanov, B. M., Hossini, A. M., Trefzer, U. & Fecker, L. F. Overcoming
apoptosis deficiency of melanoma-hope for new therapeutic approaches. Drug Resist
Updat 10, 218-234 (2007).
47. Lorigan, P., Eisen, T. & Hauschild, A. Systemic therapy for metastatic malignant melanoma-from deeply disappointing to bright future? Exp. Dermatol. 17, 383-394 (2008).
48. Fire, A. et al. Potent and specific genetic interference by double-stranded RNA in
Caenorhabditis elegans. Nature 391, 806-811 (1998).
49. Elbashir, S. M. et al. Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured
mammalian cells. Nature 411, 494-498 (2001).
50. Song, E. et al. RNA interference targeting Fas protects mice from fulminant hepatitis. Nat.
Med. 9, 347-351 (2003).
51. de Fougerolles, A., Vornlocher, H. P., Maraganore, J. & Lieberman, J. Interfering with
disease: a progress report on siRNA-based therapeutics. Nat. Rev. Drug Discov. 6, 443453 (2007).
52. Dykxhoorn, D. M. & Lieberman, J. Knocking down disease with siRNAs. Cell 126, 231-235
(2006).
53. Hutvagner, G. & Zamore, P. D. RNAi: nature abhors a double-strand. Curr. Opin. Genet.
Dev. 12, 225-232 (2002).
54. Nakamura, H. et al. RNA interference targeting transforming growth factor-beta type II
receptor suppresses ocular inflammation and fibrosis. Mol. Vis. 10, 703-711 (2004).
55. Tolentino, M. J. et al. Intravitreal injection of vascular endothelial growth factor small
interfering RNA inhibits growth and leakage in a nonhuman primate, laser-induced
model of choroidal neovascularization. Retina 24, 660 (2004).
56. Shen, J. et al. Suppression of ocular neovascularization with siRNA targeting VEGF receptor
1. Gene Ther. 13, 225-234 (2006).
57. Li, B. J. et al. Using siRNA in prophylactic and therapeutic regimens against SARS
coronavirus in Rhesus macaque. Nat. Med. 11, 944-951 (2005).
58. Zhang, X. et al. Small interfering RNA targeting heme oxygenase-1 enhances ischemiareperfusion-induced lung apoptosis. J. Biol. Chem. 279, 10677-10684 (2004).
59. Bhandari, V. et al. Hyperoxia causes angiopoietin 2-mediated acute lung injury and necrotic
cell death. Nat. Med. 12, 1286-1293 (2006).
60. Matsuyama, W. et al. Suppression of discoidin domain receptor 1 by RNA interference
attenuates lung inflammation. J. Immunol. 176, 1928-1936 (2006).
61. Makimura, H., Mizuno, T. M., Mastaitis, J. W., Agami, R. & Mobbs, C. V. Reducing
hypothalamic AGRP by RNA interference increases metabolic rate and decreases body
weight without influencing food intake. BMC Neurosci. 3, 18 (2002).
62. Thakker, D. R. et al. Neurochemical and behavioral consequences of widespread gene
knockdown in the adult mouse brain by using nonviral RNA interference. Proc. Natl.
Acad. Sci. U. S. A. 101, 17270-17275 (2004).
63. Thakker, D. R. et al. siRNA-mediated knockdown of the serotonin transporter in the adult
mouse brain. Mol. Psychiatry 10, 782-9, 714 (2005).
90 64. Dorn, G. et al. siRNA relieves chronic neuropathic pain. Nucleic Acids Res. 32, e49 (2004).
65. Jackson, A. L. et al. Widespread siRNA "off-target" transcript silencing mediated by seed
region sequence complementarity. RNA 12, 1179-1187 (2006).
66. Birmingham, A. et al. 3' UTR seed matches, but not overall identity, are associated with
RNAi off-targets. Nat. Methods 3, 199-204 (2006).
67. Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. Basic local alignment
search tool. J. Mol. Biol. 215, 403-410 (1990).
68. Jackson, A. L. et al. Expression profiling reveals off-target gene regulation by RNAi. Nat.
Biotechnol. 21, 635-637 (2003).
69. Lin, X. et al. siRNA-mediated off-target gene silencing triggered by a 7 nt complementation.
Nucleic Acids Res. 33, 4527-4535 (2005).
70. Qiu, S., Adema, C. M. & Lane, T. A computational study of off-target effects of RNA
interference. Nucleic Acids Res. 33, 1834-1847 (2005).
71. Schwarz, D. S. et al. Designing siRNA that distinguish between genes that differ by a single
nucleotide. PLoS Genet. 2, e140 (2006).
72. Judge, A. D. et al. Sequence-dependent stimulation of the mammalian innate immune
response by synthetic siRNA. Nat. Biotechnol. 23, 457-462 (2005).
73. Hornung, V. et al. Sequence-specific potent induction of IFN-alpha by short interfering RNA
in plasmacytoid dendritic cells through TLR7. Nat. Med. 11, 263-270 (2005).
74. Kim, D. H. et al. Synthetic dsRNA Dicer substrates enhance RNAi potency and efficacy.
Nat. Biotechnol. 23, 222-226 (2005).
75. Rose, S. D. et al. Functional polarity is introduced by Dicer processing of short substrate
RNAs. Nucleic Acids Res. 33, 4140-4156 (2005).
76. Reynolds, A. et al. Induction of the interferon response by siRNA is cell type- and duplex
length-dependent. RNA 12, 988-993 (2006).
77. Kim, D. H. et al. Interferon induction by siRNAs and ssRNAs synthesized by phage
polymerase. Nat. Biotechnol. 22, 321-325 (2004).
78. Siolas, D. et al. Synthetic shRNAs as potent RNAi triggers. Nat. Biotechnol. 23, 227-231
(2005).
79. Palliser, D. et al. An siRNA-based microbicide protects mice from lethal herpes simplex
virus 2 infection. Nature 439, 89-94 (2006).
80. Zimmermann, T. S. et al. RNAi-mediated gene silencing in non-human primates. Nature
441, 111-114 (2006).
81. Layzer, J. M. et al. In vivo activity of nuclease-resistant siRNAs. RNA 10, 766-771 (2004).
82. Choung, S., Kim, Y. J., Kim, S., Park, H. O. & Choi, Y. C. Chemical modification of
siRNAs to improve serum stability without loss of efficacy. Biochem. Biophys. Res.
Commun. 342, 919-927 (2006).
83. Jackson, A. L. et al. Position-specific chemical modification of siRNAs reduces "off-target"
transcript silencing. RNA 12, 1197-1205 (2006).
84. Fedorov, Y. et al. Off-target effects by siRNA can induce toxic phenotype. RNA 12, 11881196 (2006).
85. Raemdonck, K. et al. In situ analysis of single-stranded and duplex siRNA integrity in living
cells. Biochemistry 45, 10614-10623 (2006).
86. Bartlett, D. W. & Davis, M. E. Insights into the kinetics of siRNA-mediated gene silencing
from live-cell and live-animal bioluminescent imaging. Nucleic Acids Res. 34, 322-333
(2006).
91 87. Morrissey, D. V. et al. Activity of stabilized short interfering RNA in a mouse model of
hepatitis B virus replication. Hepatology 41, 1349-1356 (2005).
88. Allen, T. M. & Cullis, P. R. Drug delivery systems: entering the mainstream. Science 303,
1818-1822 (2004).
89. Liu, M., Kono, K. & Frechet, J. M. Water-soluble dendritic unimolecular micelles: their
potential as drug delivery agents. J. Control. Release 65, 121-131 (2000).
90. Moghimi, S. M., Hunter, A. C. & Murray, J. C. Long-circulating and target-specific
nanoparticles: theory to practice. Pharmacol. Rev. 53, 283-318 (2001).
91. LaVan, D. A., McGuire, T. & Langer, R. Small-scale systems for in vivo drug delivery. Nat.
Biotechnol. 21, 1184-1191 (2003).
92. Duncan, R. The dawning era of polymer therapeutics. Nat. Rev. Drug Discov. 2, 347-360
(2003).
93. Allen, T. M. Ligand-targeted therapeutics in anticancer therapy. Nat. Rev. Cancer. 2, 750763 (2002).
94. Maeda, H., Wu, J., Sawa, T., Matsumura, Y. & Hori, K. Tumor vascular permeability and the
EPR effect in macromolecular therapeutics: a review. J. Control. Release 65, 271-284
(2000).
95. Hashizume, H. et al. Openings between defective endothelial cells explain tumor vessel
leakiness. Am. J. Pathol. 156, 1363-1380 (2000).
96. Jain, R. K. Transport of molecules across tumor vasculature. Cancer Metastasis Rev. 6, 559593 (1987).
97. Northfelt, D. W. et al. Doxorubicin encapsulated in liposomes containing surface-bound
polyethylene glycol: pharmacokinetics, tumor localization, and safety in patients with
AIDS-related Kaposi's sarcoma. J. Clin. Pharmacol. 36, 55-63 (1996).
98. Schiffelers, R., Storm, G. & Bakker-Woudenberg, I. Liposome-encapsulated
aminoglycosides in pre-clinical and clinical studies. J. Antimicrob. Chemother. 48, 333344 (2001).
99. Edens, H. A. et al. Neutrophil transepithelial migration: evidence for sequential, contactdependent signaling events and enhanced paracellular permeability independent of
transjunctional migration. J. Immunol. 169, 476-486 (2002).
100. Netti, P. A., Roberge, S., Boucher, Y., Baxter, L. T. & Jain, R. K. Effect of transvascular
fluid exchange on pressure-flow relationship in tumors: a proposed mechanism for tumor
blood flow heterogeneity. Microvasc. Res. 52, 27-46 (1996).
101. Nagayasu, A., Uchiyama, K. & Kiwada, H. The size of liposomes: a factor which affects
their targeting efficiency to tumors and therapeutic activity of liposomal antitumor drugs.
Adv. Drug Deliv. Rev. 40, 75-87 (1999).
102. Ge, Q. et al. Inhibition of influenza virus production in virus-infected mice by RNA
interference. Proc. Natl. Acad. Sci. U. S. A. 101, 8676-8681 (2004).
103. Geisbert, T. W. et al. Postexposure protection of guinea pigs against a lethal ebola virus
challenge is conferred by RNA interference. J. Infect. Dis. 193, 1650-1657 (2006).
104. Grzelinski, M. et al. RNA interference-mediated gene silencing of pleiotrophin through
polyethylenimine-complexed small interfering RNAs in vivo exerts antitumoral effects in
glioblastoma xenografts. Hum. Gene Ther. 17, 751-766 (2006).
105. Schiffelers, R. M. et al. Cancer siRNA therapy by tumor selective delivery with ligandtargeted sterically stabilized nanoparticle. Nucleic Acids Res. 32, e149 (2004).
92 106. Urban-Klein, B., Werth, S., Abuharbeid, S., Czubayko, F. & Aigner, A. RNAi-mediated
gene-targeting through systemic application of polyethylenimine (PEI)-complexed
siRNA in vivo. Gene Ther. 12, 461-466 (2005).
107. Howard, K. A. et al. RNA interference in vitro and in vivo using a novel chitosan/siRNA
nanoparticle system. Mol. Ther. 14, 476-484 (2006).
108. Pille, J. Y. et al. Intravenous delivery of anti-RhoA small interfering RNA loaded in
nanoparticles of chitosan in mice: safety and efficacy in xenografted aggressive breast
cancer. Hum. Gene Ther. 17, 1019-1026 (2006).
109. Song, E. et al. Antibody mediated in vivo delivery of small interfering RNAs via cellsurface receptors. Nat. Biotechnol. 23, 709-717 (2005).
110. Biessen, E. A. et al. Targeted delivery of oligodeoxynucleotides to parenchymal liver cells
in vivo. Biochem. J. 340 ( Pt 3), 783-792 (1999).
111. Soutschek, J. et al. Therapeutic silencing of an endogenous gene by systemic administration
of modified siRNAs. Nature 432, 173-178 (2004).
112. McNamara, J. O.,2nd et al. Cell type-specific delivery of siRNAs with aptamer-siRNA
chimeras. Nat. Biotechnol. 24, 1005-1015 (2006).
113. Chu, T. C., Twu, K. Y., Ellington, A. D. & Levy, M. Aptamer mediated siRNA delivery.
Nucleic Acids Res. 34, e73 (2006).
114. Hu-Lieskovan, S., Heidel, J. D., Bartlett, D. W., Davis, M. E. & Triche, T. J. Sequencespecific knockdown of EWS-FLI1 by targeted, nonviral delivery of small interfering
RNA inhibits tumor growth in a murine model of metastatic Ewing's sarcoma. Cancer
Res. 65, 8984-8992 (2005).
115. Merdan, T., Kopecek, J. & Kissel, T. Prospects for cationic polymers in gene and
oligonucleotide therapy against cancer. Adv. Drug Deliv. Rev. 54, 715-758 (2002).
116. Takei, Y., Kadomatsu, K., Yuzawa, Y., Matsuo, S. & Muramatsu, T. A small interfering
RNA targeting vascular endothelial growth factor as cancer therapeutics. Cancer Res. 64,
3365-3370 (2004).
117. Takeshita, F. & Ochiya, T. Therapeutic potential of RNA interference against cancer.
Cancer. Sci. 97, 689-696 (2006).
118. Torchilin, V. P. Recent approaches to intracellular delivery of drugs and DNA and organelle
targeting. Annu. Rev. Biomed. Eng. 8, 343-375 (2006).
119. Li, W. & Szoka, F. C.,Jr. Lipid-based nanoparticles for nucleic acid delivery. Pharm. Res.
24, 438-449 (2007).
120. Longmuir, K. J., Robertson, R. T., Haynes, S. M., Baratta, J. L. & Waring, A. J. Effective
targeting of liposomes to liver and hepatocytes in vivo by incorporation of a Plasmodium
amino acid sequence. Pharm. Res. 23, 759-769 (2006).
121. Maclachlan, I. siRNAs with guts. Nat. Biotechnol. 26, 403-405 (2008).
122. Semple, S. C. et al. Rational design of cationic lipids for siRNA delivery. Nat. Biotechnol.
28, 172-176 (2010).
123. Stuart, D. D. & Allen, T. M. A new liposomal formulation for antisense
oligodeoxynucleotides with small size, high incorporation efficiency and good stability.
Biochim. Biophys. Acta 1463, 219-229 (2000).
124. Pastorino, F. et al. Ligand-targeted liposomal therapies of neuroblastoma. Curr. Med.
Chem. 14, 3070-3078 (2007).
125. Soengas, M. S. & Lowe, S. W. Apoptosis and melanoma chemoresistance. Oncogene 22,
3138-3151 (2003).
93 126. Judge, A. D. et al. Sequence-dependent stimulation of the mammalian innate immune
response by synthetic siRNA. Nat. Biotechnol. 23, 457-462 (2005).
127. Morrissey, D. V. et al. Potent and persistent in vivo anti-HBV activity of chemically
modified siRNAs. Nat. Biotechnol. 23, 1002-1007 (2005).
128. Huwyler, J., Wu, D. & Pardridge, W. M. Brain drug delivery of small molecules using
immunoliposomes. Proc. Natl. Acad. Sci. U. S. A. 93, 14164-14169 (1996).
129. Di Paolo, D. et al. Liposome-mediated therapy of neuroblastoma. Methods Enzymol. 465,
225-249 (2009).
130. Gutierrez-Puente, Y. et al. Safety, pharmacokinetics, and tissue distribution of liposomal Pethoxy antisense oligonucleotides targeted to Bcl-2. J. Pharmacol. Exp. Ther. 291, 865869 (1999).
131. Mui, B., Raney, S. G., Semple, S. C. & Hope, M. J. Immune stimulation by a CpGcontaining oligodeoxynucleotide is enhanced when encapsulated and delivered in lipid
particles. J. Pharmacol. Exp. Ther. 298, 1185-1192 (2001).
132. Tamm, I., Dorken, B. & Hartmann, G. Antisense therapy in oncology: new hope for an old
idea? Lancet 358, 489-497 (2001).
133. Meerum Terwogt, J. M. et al. Phase I and pharmacokinetic study of SPI-77, a liposomal
encapsulated dosage form of cisplatin. Cancer Chemother. Pharmacol. 49, 201-210
(2002).
134. Cabanes, A., Briggs, K. E., Gokhale, P. C., Treat, J. A. & Rahman, A. Comparative in vivo
studies with paclitaxel and liposome-encapsulated paclitaxel. Int. J. Oncol. 12, 1035-1040
(1998).
135. Cardoso, A. L. et al. Tf-lipoplexes for neuronal siRNA delivery: a promising system to
mediate gene silencing in the CNS. J. Control. Release 132, 113-123 (2008).
136. Zheng, X. et al. A novel in vivo siRNA delivery system specifically targeting dendritic cells
and silencing CD40 genes for immunomodulation. Blood 113, 2646-2654 (2009).
137. Jimbow, K. et al. Melanin pigments and melanosomal proteins as differentiation markers
unique to normal and neoplastic melanocytes. J. Invest. Dermatol. 100, 259S-268S
(1993).
138. Pelayo, B. A., Fu, Y. M. & Meadows, G. G. Decreased tissue plasminogen activator and
increased plasminogen activator inhibitors and increased activator protein-1 and specific
promoter 1 are associated with inhibition of invasion in human A375 melanoma deprived
of tyrosine and phenylalanine. Int. J. Oncol. 18, 877-883 (2001).
139. Pelayo, B. A., Fu, Y. M. & Meadows, G. G. Inhibition of B16BL6 melanoma invasion by
tyrosine and phenylalanine deprivation is associated with decreased secretion of
plasminogen activators and increased plasminogen activator inhibitors. Clin. Exp.
Metastasis 17, 841-848 (1999).
140. Fu, Y. M., Yu, Z. X., Pelayo, B. A., Ferrans, V. J. & Meadows, G. G. Focal adhesion
kinase-dependent apoptosis of melanoma induced by tyrosine and phenylalanine
deficiency. Cancer Res. 59, 758-765 (1999).
141. Fu, Y. M., Li, Y. Q. & Meadows, G. G. Influence of tyrosine and phenylalanine limitation
of cytotoxicity of chimeric TGF-alpha toxins on B16BL6 murine melanoma in vitro.
Nutr. Cancer 31, 1-7 (1998).
142. Fu, Y. M., Yu, Z. X., Ferrans, V. J. & Meadows, G. G. Tyrosine and phenylalanine
restriction induces G0/G1 cell cycle arrest in murine melanoma in vitro and in vivo. Nutr.
Cancer 29, 104-113 (1997).
94 143. Uhlenkott, C. E., Huijzer, J. C., Cardeiro, D. J., Elstad, C. A. & Meadows, G. G.
Attachment, invasion, chemotaxis, and proteinase expression of B16-BL6 melanoma
cells exhibiting a low metastatic phenotype after exposure to dietary restriction of
tyrosine and phenylalanine. Clin. Exp. Metastasis 14, 125-137 (1996).
144. Hadley, M. E. & Dorr, R. T. Melanocortin peptide therapeutics: historical milestones,
clinical studies and commercialization. Peptides 27, 921-930 (2006).
145. Hruby, V. J. et al. Design, synthesis, and conformation of superpotent and prolonged acting
melanotropins. Ann. N. Y. Acad. Sci. 680, 51-63 (1993).
146. Salazar-Onfray, F. et al. Tissue distribution and differential expression of melanocortin 1
receptor, a malignant melanoma marker. Br. J. Cancer 87, 414-422 (2002).
147. Dippold, W. G., Dienes, H. P., Knuth, A. & Meyer zum Buschenfelde, K. H.
Immunohistochemical localization of ganglioside GD3 in human malignant melanoma,
epithelial tumors, and normal tissues. Cancer Res. 45, 3699-3705 (1985).
148. Ruan, S., Raj, B. K. & Lloyd, K. O. Relationship of glycosyltransferases and mRNA levels
to ganglioside expression in neuroblastoma and melanoma cells. J. Neurochem. 72, 514521 (1999).
149. Lee, F. T. et al. Specific localization, gamma camera imaging, and intracellular trafficking
of radiolabelled chimeric anti-G(D3) ganglioside monoclonal antibody KM871 in SKMEL-28 melanoma xenografts. Cancer Res. 61, 4474-4482 (2001).
150. Pastorino, F. et al. Doxorubicin-loaded Fab' fragments of anti-disialoganglioside
immunoliposomes selectively inhibit the growth and dissemination of human
neuroblastoma in nude mice. Cancer Res. 63, 86-92 (2003).
151. Frost, J. D. et al. A phase I/IB trial of murine monoclonal anti-GD2 antibody 14.G2a plus
interleukin-2 in children with refractory neuroblastoma: a report of the Children's Cancer
Group. Cancer 80, 317-333 (1997).
152. Uttenreuther-Fischer, M. M., Huang, C. S., Reisfeld, R. A. & Yu, A. L. Pharmacokinetics of
anti-ganglioside GD2 mAb 14G2a in a phase I trial in pediatric cancer patients. Cancer
Immunol. Immunother. 41, 29-36 (1995).
153. Pukel, C. S. et al. GD3, a prominent ganglioside of human melanoma. Detection and
characterisation by mouse monoclonal antibody. J. Exp. Med. 155, 1133-1147 (1982).
154. Seyfried, T. N. & Yu, R. K. Ganglioside GD3: structure, cellular distribution, and possible
function. Mol. Cell. Biochem. 68, 3-10 (1985).
155. Welt, S., Carswell, E. A., Vogel, C. W., Oettgen, H. F. & Old, L. J. Immune and
nonimmune effector functions of IgG3 mouse monoclonal antibody R24 detecting the
disialoganglioside GD3 on the surface of melanoma cells. Clin. Immunol. Immunopathol.
45, 214-229 (1987).
156. Boussiotis, V. A. et al. R24 anti-GD3 ganglioside antibody can induce costimulation and
prevent the induction of alloantigen-specific T cell clonal anergy. Eur. J. Immunol. 26,
2149-2154 (1996).
157. Siddiqui, B., Buehler, J., DeGregorio, M. W. & Macher, B. A. Differential expression of
ganglioside GD3 by human leukocytes and leukemia cells. Cancer Res. 44, 5262-5265
(1984).
158. Urmacher, C., Cordon-Cardo, C. & Houghton, A. N. Tissue distribution of GD3 ganglioside
detected by mouse monoclonal antibody R24. Am. J. Dermatopathol. 11, 577-581 (1989).
95 159. Albino, A. P. et al. Class II histocompatibility antigen expression in human melanocytes
transformed by Harvey murine sarcoma virus (Ha-MSV) and Kirsten MSV retroviruses.
J. Exp. Med. 164, 1710-1722 (1986).
160. Nakano, J., Raj, B. K., Asagami, C. & Lloyd, K. O. Human melanoma cell lines deficient in
GD3 ganglioside expression exhibit altered growth and tumorigenic characteristics. J.
Invest. Dermatol. 107, 543-548 (1996).
161. Houghton, A. N. et al. Mouse monoclonal IgG3 antibody detecting GD3 ganglioside: a
phase I trial in patients with malignant melanoma. Proc. Natl. Acad. Sci. U. S. A. 82,
1242-1246 (1985).
162. Dippold, W. G. et al. Cell surface antigens of human malignant melanoma: definition of six
antigenic systems with mouse monoclonal antibodies. Proc. Natl. Acad. Sci. U. S. A. 77,
6114-6118 (1980).
163. Graus, F., Cordon-Cardo, C., Houghton, A. N., Melamed, M. R. & Old, L. J. Distribution of
the ganglioside GD3 in the human nervous system detected by R24 mouse monoclonal
antibody. Brain Res. 324, 190-194 (1984).
164. Pagnan, G. et al. Delivery of c-myb antisense oligodeoxynucleotides to human
neuroblastoma cells via disialoganglioside GD(2)-targeted immunoliposomes: antitumor
effects. J. Natl. Cancer Inst. 92, 253-261 (2000).
165. Uster, P. S. et al. Insertion of poly(ethylene glycol) derivatized phospholipid into preformed liposomes results in prolonged in vivo circulation time. FEBS Lett. 386, 243-246
(1996).
166. Leonetti, J. P., Machy, P., Degols, G., Lebleu, B. & Leserman, L. Antibody-targeted
liposomes containing oligodeoxyribonucleotides complementary to viral RNA selectively
inhibit viral replication. Proc. Natl. Acad. Sci. U. S. A. 87, 2448-2451 (1990).
167. Thierry, A. R. & Dritschilo, A. Liposomal delivery of antisense oligodeoxynucleotides.
Application to the inhibition of the multidrug resistance in cancer cells. Ann. N. Y. Acad.
Sci. 660, 300-302 (1992).
168. Ropert, C., Lavignon, M., Dubernet, C., Couvreur, P. & Malvy, C. Oligonucleotides
encapsulated in pH sensitive liposomes are efficient toward Friend retrovirus. Biochem.
Biophys. Res. Commun. 183, 879-885 (1992).
169. Thierry, A. R., Rahman, A. & Dritschilo, A. Overcoming multidrug resistance in human
tumor cells using free and liposomally encapsulated antisense oligodeoxynucleotides.
Biochem. Biophys. Res. Commun. 190, 952-960 (1993).
170. Richards, R. L. et al. Influence of vesicle size on complement-dependent immune damage
to liposomes. Biochim. Biophys. Acta 855, 223-230 (1986).
171. Harashima, H., Hiraiwa, T., Ochi, Y. & Kiwada, H. Size dependent liposome degradation in
blood: in vivo/in vitro correlation by kinetic modeling. J. Drug Target. 3, 253-261 (1995).
172. Liu, D., Mori, A. & Huang, L. Role of liposome size and RES blockade in controlling
biodistribution and tumor uptake of GM1-containing liposomes. Biochim. Biophys. Acta
1104, 95-101 (1992).
173. Maruyama, K. et al. Prolonged circulation time in vivo of large unilamellar liposomes
composed of distearoyl phosphatidylcholine and cholesterol containing amphipathic
poly(ethylene glycol). Biochim. Biophys. Acta 1128, 44-49 (1992).
174. Moore, N. M., Sheppard, C. L., Barbour, T. R. & Sakiyama-Elbert, S. E. The effect of
endosomal escape peptides on in vitro gene delivery of polyethylene glycol-based
vehicles. J. Gene Med. 10, 1134-1149 (2008).
96 175. Schiffelers, R. M. et al. Transporting silence: design of carriers for siRNA to angiogenic
endothelium. J. Control. Release 109, 5-14 (2005).
176. Iden, D. L. & Allen, T. M. In vitro and in vivo comparison of immunoliposomes made by
conventional coupling techniques with those made by a new post-insertion approach.
Biochim. Biophys. Acta 1513, 207-216 (2001).
177. Hansen, C. B., Kao, G. Y., Moase, E. H., Zalipsky, S. & Allen, T. M. Attachment of
antibodies to sterically stabilized liposomes: evaluation, comparison and optimization of
coupling procedures. Biochim. Biophys. Acta 1239, 133-144 (1995).
178. Moreira, J. N., Ishida, T., Gaspar, R. & Allen, T. M. Use of the post-insertion technique to
insert peptide ligands into pre-formed stealth liposomes with retention of binding activity
and cytotoxicity. Pharm. Res. 19, 265-269 (2002).
179. Uster, P. S. et al. Insertion of poly(ethylene glycol) derivatized phospholipid into preformed liposomes results in prolonged in vivo circulation time. FEBS Lett. 386, 243-246
(1996).
180. Patten, R. M., Shuman, W. P. & Teefey, S. Subcutaneous metastases from malignant
melanoma: prevalence and findings on CT. AJR Am. J. Roentgenol. 152, 1009-1012
(1989).
181. Zheng, X. et al. A novel in vivo siRNA delivery system specifically targeting dendritic cells
and silencing CD40 genes for immunomodulation. Blood 113, 2646-2654 (2009).
182. Judge, A., McClintock, K., Phelps, J. R. & Maclachlan, I. Hypersensitivity and loss of
disease site targeting caused by antibody responses to PEGylated liposomes. Mol. Ther.
13, 328-337 (2006).
183. Pastorino, F. et al. Enhanced antitumor efficacy of clinical-grade vasculature-targeted
liposomal doxorubicin. Clin. Cancer Res. 14, 7320-7329 (2008).
184. Khan, A. et al. Sustained polymeric delivery of gene silencing antisense ODNs, siRNA,
DNAzymes and ribozymes: in vitro and in vivo studies. J. Drug Target. 12, 393-404
(2004).
185. Cheng, W. W. & Allen, T. M. Targeted delivery of anti-CD19 liposomal doxorubicin in Bcell lymphoma: a comparison of whole monoclonal antibody, Fab' fragments and single
chain Fv. J. Control. Release 126, 50-58 (2008).
186. Heyes, J., Palmer, L., Bremner, K. & MacLachlan, I. Cationic lipid saturation influences
intracellular delivery of encapsulated nucleic acids. J. Control. Release 107, 276-287
(2005).
187. Litzinger, D. C. & Huang, L. Phosphatidylethanolamine liposomes: drug delivery, gene
transfer and immunodiagnostic applications. Biochim. Biophys. Acta 1113, 201-227
(1992).
188. Brgles, M., Habjanec, L., Halassy, B. & Tomasic, J. Liposome fusogenicity and entrapment
efficiency of antigen determine the Th1/Th2 bias of antigen-specific immune response.
Vaccine 27, 5435-5442 (2009).
97