Human pluripotent stem cell derived cellular vehicles for cancer gene therapy

1.1.1.6.1 Gene therapy to inhibit pro-angiogenic pathways 12 1.1.1.6.2 Gene therapy to stimulate anti-angiogenic pathways 12 1.2 Stem cells- unique candidates for therapeutic platform 19

HUMAN PLURIPOTENT STEM CELL-DERIVED CELLULAR VEHICLES FOR CANCER GENE

THERAPY

DANG HOANG LAM

NATIONAL UNIVERSITY OF SINGAPORE

2012

HUMAN PLURIPOTENT STEM CELL-DERIVED CELLULAR VEHICLES FOR CANCER GENE

THERAPY DANG HOANG LAM

(B.Sc., University of Natural Science, Vietnam)

A THESIS SUBMITTED

FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF BIOLOGICAL SCIENCES NATIONAL UNIVERSITY OF SINGAPORE

&

INSTITUTE OF BIOENGINEERING AND NANOTECHNOLOGY

2012

ACKNOWLEDGMENTS

First and foremost, I would like to express my sincere thankfulness to my scientific supervisor, Dr Wang Shu for his constant and invaluable guidance throughout my PhD study This thesis would not have been completed without his support, understanding and motivation

I am so grateful to have opportunities to cooperate with Xiao Ying Bak, Zhao Ying, Yang Jing and Yukti Choudhury It was unforgettable experience working with them and together we discovered meaningful findings which will be presented in this dissertation Thank to all lab members Esther Lee, Ghayathri Balasundaram, Jiakai Lin, Seong Loong

Lo, Chrishan Julian Alles Ramachandra, Mohammad Shahbazi, Chunxiao Wu, Kai Ye, Jieming Zeng and Detu Zhu for the help and the laugh they brought in my difficult time of research Thank to Yovita Ida Purwanti and Timothy Kwang for accompanying me during lunchtime

Special thank to Thomas Ngo for his help during thesis writing process

I am also thankful to my friends including Thuc Nha, Quang Vinh, and Thanh Duong, who always encouraged and supported me during my stay in Singapore

Last but most important, this thesis is dedicated to my parents and brother who unconditionally loved me and stood by me in any time of troubles

1.1.1 Gene therapy strategies for cancer treatment 2

1.1.1.1.1 Herpes Simplex Virus Thymidine Kinase/Ganciclovir

1.1.1.6.1 Gene therapy to inhibit pro-angiogenic pathways 12

1.1.1.6.2 Gene therapy to stimulate anti-angiogenic pathways 12

1.2 Stem cells- unique candidates for therapeutic platform 19

1.2.1.1.2 Induced pluripotent stem cells 21

1.2.2 Clinical relevance of stem cell vectors for cancer gene therapy 24

1.2.3 New reliable sources for therapeutic stem cells 26

2.2 Mouse strains 44

2.3 Viral vector preparation and cell transduction 44

2.10 Intracranial xenografting of human cells 50

Chapter 3 Results: Human Embryonic Stem Cell-derived Mesenchymal

and Neural Stem Cells as Cellular Vehicles for Suicide Gene Therapy of

3.2 hESC-MSCs deliver suicide gene in HSV-tk/GCV system to treat

3.2.1 In vitro and in vivo glioma tropism of hESC-MSCs 59

3.2.2 Transgene expression in hESC-MSCs following baculoviral and

3.2.3 Intratumor transplantation of HSVtk-expressing hESC-MSCs

transduced by baculovirus for anti-glioma therapy 61

3.2.4 Migratory HSVtk-expressing hESC-MSCs transduced by lentivirus

3.3 hESC-NSCs deliver suicide gene in HSV-tk/GCV system to treat

3.3.1 Generation and characterization of NSCs derived from hESCs in

3.3.2 Glioma-specific migration property of NSC1 cells 68

3.3.3 Transgene expression in NSC1 cells using baculoviral vector 69

3.3.4 Tumor killing efficacy of HSVtk-expressing NSC1 cells in glioma

3.4 Immunomodulation effects of MSC and NSC on tumor growth 76

Chapter 4 Results: Human Induced Pluripotent Stem Cell-derived

Neural Stem Cells as Cellular Vehicles for Suicide Gene Therapy

of Invasive Glioma and Metastatic Breast Cancer 80

4.2 Human Induced Pluripotent Stem Cell-derived Neural Stem Cells

as Cellular Vehicles for Suicide Gene Therapy of Invasive Glioma 85

4.2.1 Establishment of highly metastatic and invasive glioblastoma model

in mouse using experimental lung metastasis (ELM) assay 85

4.2.2 In vitro and in vivo migration of iPS-NSCs towards invasive 2M1 88

4.2.3 2M1 glioma growth inhibition by engineered iPS-NSCs 89

4.3 Gene therapy application of intravenously injected human iPS-NSCs

4.3.1 In vivo distribution of human iPS-NSCs in normal mice without tumors 96

4.3.2 In vivo distribution of human iPS-NSCs in 4T1 breast tumor-bearing

mice 98 4.3.3 Systemic injection of HSVtk-expressing iPSNSCs inhibits 4T1 breast

References 110

5.1 hESCs as source for MSCs and NSCs in suicide gene therapy 113

5.1.1 Baculoviruses as gene therapy vectors for human cancer 113

5.1.2 HESC- derived MSCs and NSCs as gene delivery vehicles for

5.1.3 NSCs may be a better option for glioma therapy 118

5.2 iPS cells as source for NSCs in treating invasive glioblastoma and

5.2.1 2M1 glioma cell line expresses similarities with human GBM 121

5.2.2 iPS-NSCs – promising gene carriers for highly metastatic cancers 122

Chapter 6 Conclusion 131

Cancer, also known as malignant neoplasm, is one of the leading causes of morbidity and mortality throughout the world Hitherto cancer therapies that have been widely accepted and practiced are skillful surgery or microsurgery, radiotherapy and chemotherapy or combination Those treatments, however, are too invasive for normal and healthy tissues and lead to further patient weakening and shortened survival By conventional methods, cancer is still a deadly disease with poor prognosis and the outcome is variable and depends on many factors such as age, gender, tumor type and stage Hence, more efficacious system for cancer treatment is desirable

Of all candidates, gene therapy to achieve therapeutic effects through transgene expression is a promising treatment Unfortunately, although gene therapy using viral vectors produced encouraging results in various animal models, the same accomplishments have not been observed in clinical trials owing to low gene expression and more importantly poor distribution of vectors in patients Stem cells have recently become an emerging solution to serve as fascinating vehicles for gene delivery in cancer therapy because they show favorable migration towards local and metastatic tumors This allows more specific targeting of transgenes to tumor sites, thus increasing efficacy

However, one major issue with stem cell based gene therapy is the use of adult stem cells which causes limited cell sources and variable quality of cells from donors To overcome these obstacles, alternative large scale production of therapeutic stem cells must be identified Pluripotent stem cells including embryonic stem cells and induced pluripotent stem cells may hold the answer to this question

From that point of view, we conducted investigations to assess the feasibility of using human embryonic stem cell (hESC)-derived neural stem cells (NSCs) and mesenchymal stem cells (MSCs) for glioma gene therapy We observed that human glioma bearing mice showed noticeable reduction in tumor size and improvement in survival rate when baculoviral and lentiviral herpes simplex virus-thymidine kinase (HSV-tk) gene was transferred by hESC-derived MSCs and NSCs in combination with the systemic administration of ganciclovir (GCV)

Employing an in-house invasive glioblastoma model, we further proved that induced pluripotent stem cell (iPS cell)-derived NSCs also displayed a robust migration towards brain tumor This resulted in a significant tumor regression when cells were coupled with HSVtk/GCV system

Interestingly, using a dual-color whole body imaging technology, we discovered that NSCs possessed a strong migration capability to established orthotopic 4T1 mouse mammary tumors after tail vein injection Transduced iPS-NSCs with baculoviral vector containing HSVtk gene inhibited the growth of breast tumor and the metastatic spread of the cancer cells in the presence of GCV

iPS-Therefore, our study provides evidence supporting the use of stem cells as gene carriers for anti cancer therapy and proposes pluripotent stem cells as potential alternatives to derive therapeutic cells for this purpose

LIST OF PUBLICATIONS 1.Tumor tropism of intravenously injected human-induced pluripotent stem cell- derived neural stem cells and their gene therapy application in a metastatic breast cancer model Stem cells (Dayton, Ohio), Volume 30 , Issue 5 (2012)

Yang J ** , Lam D H **, Goh S S , Lee E X , Zhao Y , Tay F C , Chen C , Du S , Balasundaram G , Shahbazi M , Tham C K , Ng W H , Toh H C , Wang S

2.Targeted suicide gene therapy for glioma using human embryonic stem derived neural stem cells genetically modified by baculoviral vectors Gene therapy , Volume 19 , Issue 2 (2012)

cell-Zhao Y , Lam D H , Yang J , Lin J , Tham C K , Ng W H , Wang S

3.Glioma gene therapy using induced pluripotent stem cell derived neural stem cells Molecular pharmaceutics, Volume 8, Issue 5 (2011)

Lee E X , Lam D H , Wu C , Yang J , Tham C K , Ng W H , Wang S

4.Human embryonic stem cell-derived mesenchymal stem cells as cellular delivery vehicles for prodrug gene therapy of glioblastoma Human gene therapy , Volume 22 , Issue 11 (2011)

Bak X Y ** , Lam D H ** , Yang J , Ye K , Wei E L , Lim S K , Wang S

5.Combinatorial control of suicide gene expression by tissue-specific promoter and microRNA regulation for cancer therapy Molecular therapy: the journal of the American Society of Gene Therapy , Volume 17 , Issue 12 (2009)

Wu C , Lin J , Hong M , Choudhury Y , Balani P , Leung D , Lam D H , Zhao Y , Zeng J ,

Wang S

**

The two authors contributed equally to the work

LIST OF TABLES

TABLES PAGE

Table 4.1 qPCR for human specific HLA-A STR#54 to quantify iPS-NSCs 109

LIST OF FIGURES

FIGURES PAGE

Figure 1.1 Percentage of each gene therapy strategy for clinical trials 13

Figure 3.1 In vitro and in vivo migration of hESC-MSCs towards glioma tumor 63

Figure 3.3 In vivo glioma therapy using hESC-MSCs transduced with

Figure 3.4 In vivo glioma therapy using hESC-MSCs transduced with lentiviral

vectors 66

Figure 3.6 In vitro and in vivo migration of hESC-NSCs towards glioma cells 73

Figure 3.8 Anti-glioma effect by contralaterally injection of lentiviral and

Figure 3.9 Immunomodulation of hESC derived NSCs and MSCs promotes

Figure 4.1 ELM assay to select highly invasive U87 cells 90

Figure 4.3 Gene expression difference between 2M1 and U87 92

Figure 4.4 In vitro migration of iPS-NSCs towards 2M1 and U87 93

Figure 4.5 Distribution of iPS-NSCs in 2M1 tumors 94

Figure 4.7 In vivo distribution of iPS-NSCs in normal and tumor bearing NSG

Figure 4.13 Therapeutic effects of intravenously transplanted iPS-NSCs in 4T1

CRAds Conditionally replicating adenoviruses

EGFP Enhanced green fluorescent protein

element

EPR Enhanced permeability and retention

GCV Ganciclovir

GDEPT Gene directed enzyme prodrug therapy

GFAP Glial fibrillary acidic protein

GM-CSF Granulocyte macrophage colony-stimulating factor

GPAT Gene prodrug activation therapy

HIV Human immunodeficiency virus

HSVtk Herpes simplex virus thymidine kinase

IL Interleukin

iPS cell Induced pluripotent stem cell

LV Lentivirus

miRNA microRNA

NSG NOD.Cg-Prkdc scid Il2rg tm1Wjl/SzJ

RISC RNA inducible silencing complex

TMZ Temozolomide

VDEPT Virus directed enzyme prodrug therapy

WPRE Woodchuck hepatitis virus post-transcriptional regulatory

CHAPTER 1

INTRODUCTION

1.1 Gene therapy in cancer treatment

Rapid advances in molecular technology over the last two decades have enabled the development of gene therapy, which introduces genetic material into cells for therapeutic purposes Originally, gene therapy was proposed to correct genetic disorders by transferring a normal copy of the defective gene to prevent or reverse the course of disease, as exemplified by clear progress in the treatment of blindness and inherited immune deficiencies It became more apparent that the power of gene therapy

is not limited to regenerative medicine, but could also be used to treat acquired diseases such as cancer

1.1.1 Gene therapy strategies for cancer treatment

A large number of potential cancer gene therapies have been proposed and tested as reviewed below

1.1.1.1 Suicide gene therapy

Conceptually, suicide gene therapy has a major advantage over classical chemotherapy since it involves the delivery of a gene encoding enzyme which is not toxic per se, but can convert an inert substance or prodrug to highly toxic metabolite Thus, if applicable, this method could selectively abolish malignant cells without damaging normal and healthy tissue Although prodrug treatments for cancer in animal models with high level

of endogenous enzymes have gained certain achievements (Connors and Whisson, 1966), clinical trials have failed to deliver any benefits due to the low level of activating enzymes in human cancers (Connors, 1995) These results led to the development of the gene transferring system known as suicide gene therapy to enhance enzyme levels

at the target and design more efficient enzyme/prodrug couplings

In the literature, suicide gene therapy (SGT) is also defined as: gene directed enzyme prodrug therapy (GDEPT) (Marais et al., 1996); virus directed enzyme prodrug therapy (VDEPT) (Huber and Richards, 1996); and gene prodrug activation therapy (GPAT) (Eaton et al., 2001) All terms are to describe a two-step treatment (Niculescu-Duvaz and Springer, 2005) In the first step, a gene encoding for a foreign enzyme (viral, bacterial or yeast origin) is introduced by various methods into the tumor region In the second step, a harmless prodrug when systematically administered is converted to its deadly counterpart by the expressed enzyme It is crucial that the foreign enzyme needs

to be intensively expressed in tumor cells in comparison with normal tissue and blood Furthermore, since it is impossible for foreign enzyme to affect all tumor cells, a bystander cytotoxic effect, which was initially described by Moolten (1986), is required for any SGT system (Greco and Dachs, 2001) Moolten’s explanation for the bystander phenomenon wherein the active drug kills not only the reachable tumor cells but also the neighboring untransfected tumor cells holds the key to understanding how nearly complete tumor eradication can be accomplished with only 5-10 % of the target cells Several SGT systems have been developed over the last 25 years since their inception Among those systems, two that have been studied extensively and tested in clinical trial are the herpes simplex virus thymidine kinase/ganciclovir (HSV-tk/GCV) prodrug system and the cytosine deaminase-5- fluorocytosine (CD-5-FC) prodrug system

1.1.1.1.1 Herpes Simplex Virus Thymidine Kinase/Ganciclovir (HSV-tk/GCV) Prodrug System

After phosphorylation by the enzyme HSV-tk, GCV, a nontoxic purine analogue, is phosphorylated by endogenous kinases to become GCV-triphosphate, which interrupts DNA synthesis and consequently causes cell death through apoptosis (Moolten and Wells, 1990) The precise mechanisms of cell death induced by HSV-tk, however, are still not fully understood

During the 1990s, more than 400 papers have shown increasing evidence to support the potential of HSV-tk/GCV (Greco and Dachs, 2001) Encouragingly, studies utilizing adeno- and retroviral vectors were conducted in many animal models to successfully erase liver metastases (Caruso et al., 1993), hepatocellular carcinomas (Kuriyama et al., 1999), head and neck carcinomas (O’Malley, 1995), and glioblastomas (Takamiya et al., 1992)

Despite its popularity in cancer research, the HSV-tk/GCV system has several drawbacks which include nonspecific toxicity, slow killing effects, incomplete tumor killing and a weaker bystander effect compared to other suicide gene systems (Kuriyama et al., 1999) To overcome these limitations, efforts have been made to modify the established HSV-tk to improve the kinetics of activation for GCV (Black et al.,

2001) TK.007 variant was designed (Preuss et al., 2011) with superior killing efficiency

and a bystander effect that had a significant improvement in anti-tumor activity compared with conventional HSV-tk

1.1.1.1.2 Cytosine Deaminase-5- Fluorocytosine (CD-5-FC) Prodrug System

The principle of the CD-5-FC system is that the enzyme cytosine deaminase from bacteria or yeast, but not mammalian cells transforms non-toxic 5-FC into active 5-

fluorouracil (5-FU) (O'Malley et al., 1995), which is further modified to three potent metabolites (5-FdUMP, 5-FdUTP, and 5-FUTP) The resultant products induce cell death by thymidylate synthase inhibition and the formation of (5-FU) RNA and (5-FU) DNA complexes (Springer and Niculescu-Duvaz, 2000)

anti-A strong bystander effect is one of the advantages of the CD-5-FC system anti-A significant improvement in the survival rate has been demonstrated in human colorectal carcinomas in nude mice when only 4% of the inoculated cells expressed CD (Trinh et al., 1995)

1.1.1.2 Mutant gene correction/Tumor suppressor gene replacement

Cancer is a consequence of complex genetic and epigenetic alterations involving tumor suppressor genes and oncogenes A potential approach to treat cancer is to restore the expression of tumor suppressor genes or to inhibit oncogene expression Among the candidates, p53 activities have been intensively investigated

p53 was first described as transcription factor that positively or negatively regulates the expression of a diverse array of target genes responsible for various pro-apoptotic Bcl-2 proteins (Bax, Bak, Puma, and Noxa), caspases, Apaf-1, PIDD, death receptors (Fas/CD95, DR4, DR5), DNA repair proteins or the cell cycle inhibitor p21 As a result, p53 seems to have a dictatorial effect on cell stress responses such as cell cycle arrest, apoptosis, and senescence (Essmann and Schulze-Osthoff, 2012)

In healthy unstressed cells, p53 has a short half life and its expression remains low owing to constant proteasomal degradation (Bossi and Sacchi, 2007) When cellular stresses such as DNA damage, hypoxia, oncogene overexpression, or viral infection

responses to prevent cellular transformation and tumorigenesis (Bossi and Sacchi, 2007)

In approximately 50% of human cancers, however, mutations in p53 are present and malfunctions of the p53 pathway are observed in the remaining half (Soussi et al., 2000) Mouse models without functional p53 develop spontaneous tumors (Bossi and Sacchi, 2007) Hence, a straightforward gene therapy to fight cancer is the restoration

of wild- type (wt) p53 function in human cancer

In the earliest phase I study, direct intratumoral injection of a retrovirus vector expressing the wt-p53 gene was tested in patients with non-small cell lung carcinoma (Roth et al., 1996) There was evidence of increased apoptosis and tumor regression

To avoid the risk of tumorigenesis by a retrovirus, adenoviral vectors have been used as alternatives for several subsequent clinical trials for lung, bladder, ovarian, and breast cancers (Essmann and Schulze-Osthoff, 2012)

Despite many supportive examples of therapeutic efficacy achieved by wt-p53 delivery, the Food and Drug Administration (FDA) did not approve Advexin, a wt p53-encoding adenovirus developed by Introgen Therapeutics The arguments against the approval were based on the low transfection efficiency of adenoviral vectors and the lack of strong bystander effects

Nonetheless, adenovirus-p53 has been approved in China for head and neck cancer treatments in combination with radiotherapy (Shi and Zheng, 2009; Yang et al., 2010)

1.1.1.3 Cancer immunotherapy

An emerging and striking hallmark of cancer is its ability to avoid constant monitoring by the various arms of immune surveillance and inhibit immunological killing, thereby

evading its own eradication (Hanahan and Weinberg, 2011) Hence, it is beneficial for cancer treatment if a gene transferred to tumor cells can make them more highly immunogenic Transferring genes that encode tumor-specific antigens and inflammatory cytokines have been investigated (Touchefeu et al., 2010)

Tumor-specific antigens, which are normally not unique among types of cancer, can originate from oncogenic viruses such as Epstein-Barr virus and human papilloma virus

or can be self-antigens (Touchefeu et al., 2010) The mechanisms of how self proteins

in normal cells become tumor antigens in malignant cells are still unclear (Disis et al., 2009) Possibilities include gene mutation or post-translational modification to make them more immunogenic, such as the case with P53 Over-expressed proteins such as HER2/neu in breast, ovarian and prostate cancers and EGFR in non-small cell lung cancer can also serve as tumor-antigens (Disis et al., 2009) Incorrectly glycosylated proteins such as MUC1 (mucin) are immunogenic in several types of malignant diseases (Disis et al., 2009) Additional examples of tumor antigens include: gp100, tyrosinase, and the MAGE, GAGE, and BAGE families Currently, the introduction of self-antigens by means of peptides (subdominant or cryptic epitopes), full proteins, skin

or muscle cells transfected with antigen plasmid DNA, and dendritic cells presenting self-antigens can activate the T-cell response and produce a better outcome in cancer patients One future direction is the use of gene therapy to deliver those tumor-antigens

In addition to tumor antigens, immunotherapy also utilizes vectors expressing inflammatory cytokines such as interleukin-2 (IL-2), interleukin-12 (IL-12), tumor necrosis factor (TNF)-α, interferon-γ, and granulocyte macrophage colony-stimulating factor (GM-CSF) (Vachani et al., 2010)

IL-12 showed a superior anti-tumor capability in comparison with other cytokines Recombinant IL-12 or gene therapy of IL-12, when introduced into mouse models, induced an anti-tumor immune response against melanoma, sarcoma, and mammary carcinoma (Hallaj-Nezhadi et al., 2010)

Likewise, GM-CSF is one of the most promising cytokines for immunotherapy In experiments using a B16 melanoma model to compare efficacy of 10 different cytokines (IL-1RA, IL-2, IL-4, IL-5, IL-6, GM-CSF, IFN γ, TNF and CD2), the incorporation of GM-CSF into vaccinating tumor cells produced the highest degree of systemic antitumor immunity via the CD4 and CD8 pathways (Dranoff et al., 1993)

1.1.1.4 Oncolytic viruses/ Virotherapy

Traditionally, the viruses used to express therapeutic genes, such as retroviruses, adenoviruses and adeno-associated viruses, are non-replicating viruses (Kyritsis et al., 2009)

In contrast, oncolytic viruses are replication-competent viruses that have been selected

or modified to preferentially replicate in tumor cells with minimal toxicity to normal tissue This anti-cancer approach is based on tumor-cell lysis induced by oncolytic vectors after their infection and replication in malignant cells New viruses are also released to infect surrounding tumor cells Moreover, oncolytic cell death can act as an immunogenic factor to further control tumor growth (Pesonen et al., 2011)

As an example for oncolytic viruses, adenoviruses have been genetically engineered to construct conditionally replicating adenoviruses (CRAds) (Kyritsis et al., 2009) There are two types of CRAds developed Type 1 CRAds feature loss-of-function mutations in

the virus genome, which allows virus replication to occur only in cancer cells (Kyritsis et al., 2009)

Adenovirus dl520 ONYX-015® (Onyx Pharmaceuticals Inc., CA, USA), the first oncolytic virus to undergo clinical trial in 1998, carries two mutations in the viral E1B-55kDa gene (Kirn, 2001) The E1b-55kDa protein is responsible for binding and inactivating p53 in infected cells and required for effective virus replication The mutations in E1b-55kDa make ONYX-015 viruses able to replicate in tumor cells lacking functional p53 but not in normal cells (Ries, 2005)

Another CRAd is the Delta-24 adenovirus which contains a 24-bp deletion in the pRb binding site in the E1A region The binding of pRb to the E1A region is necessary for E2F release, which activates the E2 viral promoter and initiates viral replication in normal cells Thus, Delta-24 adenoviruses are unable to replicate in healthy normal cells In tumor cells where E2F is abundant, however, E1A-pRb binding is unnecessary (Fueyo et al., 2000)

In type-2 CRAds, viral promoters are replaced by tumor-specific promoters such as prostate-specific antigen, carsinoembryonic antigen, alpha-fetoprotein, tyrosinase, human telomerase (hTERT), and cyclooxygenase-2 (Cox-2) (Pesonen et al., 2011) Besides oncolytic adenoviruses, several viruses such as oncolytic herpes simplex virus (HSV), vaccinia virus, reovirus, Newcastle disease virus, Seneca Valley virus and measles virus have been developed and tested in cancer patients (Kanai et al., 2010)

In addition to tumor-lysis activities, oncolytic viruses have been designed to carry cancer transgene for therapeutic effect enhancement The combination between double suicide gene therapy with cytosine kinase – HSV-1 thymidine kinase and oncolytic

anti-adenoviruses showed an enhanced therapeutic effect (Rogulski et al., 2000) Vaccinia viruses expressing cytokine granulocyte-macrophage colony stimulating factor (GMCSF) have also progressed to phase 3 studies (Merrick et al., 2009)

A recognized shortcoming from using oncolytic viruses is that the antiviral immune response neutralizes viral activities and reduces the transduction efficiency, thus limiting efficacy of the treatment (Pesonen et al., 2011) It is not clear whether the immunogenicity of oncolytic viruses is an advantage or disadvantage for cancer treatment as clearing viral vectors can prevent their spreading into normal tissues and inflammatory response can provide additive antitumor effect To avoid an antiviral immune response, carrier cells or coating agents have been developed (Pesonen et al., 2011) This aspect will be discussed extensively in the coming parts of Introduction chapter

1.1.1.5 RNA interference

RNA interference (RNAi) is a new therapeutic for many human genetic diseases including amyotrophic lateral sclerosis, Huntington’s disease, Alzheimer’s disease, Parkinson’s disease, spinocerebellar ataxia, dominant muscular dystrophies, and cancer (Seyhan, 2011)

Cancer is a disease of genetic and epigenetic heterogeneity in a tumor mass, in distant sites from the same patient, and among different individuals Hence, attempts to inhibit tumor growth by establishing standard systemic therapeutic approaches were of little avail An alternative is to develop personalized therapy programs in which RNAi with its specificity may serve as a great candidate (Wang et al., 2011)

RNAi, discovered in 1998 (Fire et al., 1998), is defined as gene-silencing produced by small RNAs including microRNA (miRNA), small interfering RNA (siRNA), or short hairpin RNA (shRNA) through either cleavage-dependent or cleavage-independent RNA inducible silencing complex (RISC) effector processes (Wang et al., 2011)

siRNA and shRNA are synthetic RNA introduced into cells to induce sequence-specific gene-silencing Various preclinical studies using siRNA and shRNA have indicated better outcomes for cell growth, metastasis, angiogenesis and chemo-resistance (Wang

et al., 2011) For example, an overexpression of the oncogene pancreatic duodenal homeobox-1 (PDX-1) has been observed in pancreatic adenocarcinoma Three injections of a plasmid encoding shRNA to knockdown PDX-1 have shown noticeable reduction of tumor size and PDX-1 expression (Liu et al., 2008)

miRNA is endogenous RNA that correlates with malignant processes Depending on the difference in the expression levels between normal and malignant cells, the upregulation or downregulation of miRNA cause tumor suppression (Wang et al., 2011) For example, miR-26a is highly expressed in normal liver tissue but is maintained at low levels in liver tumors Patients with low levels of miR-26a have a shorter overall survival (Ji et al., 2009) Similarly, miR-34a is often lost in human cancers such as lung and prostate cancers Systemic delivery of miR-34a displayed a reduction in tumor volume

in a mouse model of non-small cell lung cancer (Trang et al., 2011)

1.1.1.6 Anti-angiogenic cancer gene therapy

Another potential strategy for cancer treatment is anti-angiogenic therapy The most important and common mediators of angiogenesis are vascular endothelial growth

factor (VEGF) and its tyrosine kinase receptor (or VEGFR) (Ferrara, 2002; Ferrara et al., 2003)

1.1.1.6.1 Gene therapy to inhibit pro-angiogenic pathways

VEGF and VEGFR signaling pathways are main targets to inhibit pro-angiogenesis The delivery of soluble forms of VEGFRs (sVEGFRs), which have no transmembrane domain to activate intracellular signal cascades, can block intracellular signals by forming dominant-negative heterodimers with transmembrane VEGFRs (Samaranayake

et al., 2010) sVEGFR-1, the first sVEGFR version, demonstrated a high affinity for VEGF thus secluding it from binding with transmembrane receptors (Samaranayake et al., 2010) Exploiting sVEGFR-1 in cancer models did provide inhibition of tumor angiogenesis, metastasis, and growth Subsequently, sVEGFR-2 and -3 were also developed (Samaranayake et al., 2010)

VEGF inhibition was also achieved with anti-sense VEGF constructs (Saleh et al., 1996) and VEGF siRNA delivery (Im et al., 1999)

1.1.1.6.2 Gene therapy to stimulate anti-angiogenic pathways

Endostatin is a VEGF and bFGF inhibitor that is employed to treat several types of cancer with both viral and non viral vectors (Samaranayake et al., 2010)

Angiostatin, a plasmin degradation product, binds to and inhibits the proliferation and migration of endothelial cells (Samaranayake et al., 2010) The overexpression of angiostatin in liver cancer cells suppressed tumor growth but did not inhibit tumor angiogenesis in mouse models (Ishikawa et al., 2003; Schmidt et al., 2006)

Further attempts have been made to introduce fusion proteins of endostatin and angiostatin In some studies, the fusion gene approach has shown superior effects

compared to single gene expression or even the injection of two genes (Samaranayake

1.1.2 Gene carriers for cancer therapy

Successful gene therapy relies largely on gene delivery efficiency To accomplish this task, numerous viral and nonviral vectors have been developed Each vector possesses various properties of DNA carrying capacity, gene transfer efficiency, and safety for healthy tissue and patients, indicating that no vector is universal for all types of cancer

1.1.2.1 Viral vectors

Adenoviruses are nonenveloped and medium-sized (90-100 nm) viruses, which contain

a double-stranded DNA genome (Chen et al., 2010) The wild-type viruses, which cause benign respiratory and eye infections in humans, comprise four early (E1, E2, E3 and E4), four intermediate, and one late transcriptional unit (Liu et al., 2010) To modify adenoviruses for gene transfer, the viral E1 gene, which is required for reproduction, must be deleted and replaced by a therapeutic gene Additional gene alterations in the E2 gene and PEG modification or removal of the coding sequence in the viral genome provides a better DNA carrying capacity (Crowther et al., 2008; Engelhardt et al., 1994) Subgroup C and serotype 5 adenoviruses (Ad5) often serve as vectors for clinical application as their pathology is well understood (Pesonen et al., 2011)

Adenoviruses are widely used for cancer gene therapy (25% of clinical trials) (Edelstein

et al., 2007) This is due to several advantageous features of adenoviruses: relative ease in high-titer production, high transduction efficiency, high gene expression, capabilities to infect both dividing and non-dividing cells, no risk of insertional mutagenesis as the viral genome stays episomal in target cell, and a large DNA load (Pesonen et al., 2011) Moreover, the safety profile of adenoviral vectors in humans has been excellent

However, transfection by adenoviruses is transient, and thus repetitive administration of adenovirus vector is needed (Liu et al., 2010) This is also an issue of using adenoviruses since their administration induces both innate and acquired immune response involving cytokine surge, viral neutralizing antibodies, and cytotoxic lymphocytes All of these effects reduce systemic viral particles available for treatment

Retrovirus and Lentivirus

Retroviruses are small, single-stranded RNA viruses with a DNA intermediate Genes encoding viral proteins such as gag, pol and env need to be removed for the development of the retroviral vector (Liu et al., 2010) The main advantage of this vector

is its ability to achieve long-term and stable expression of therapeutic genes as the DNA intermediate is able to integrate into the host genome Furthermore, a smaller immune response against the viral antigens occurs because their structural proteins are not expressed (Morse and Lyerly, 2002) Retroviral vectors were encapsulated with genes responsible for IL-12, IL-23, IL-27 and HSV-tk (Fillat et al., 2003; Liu et al., 2009; Liu et al., 2008; Wang et al., 2007) However, multiple gene delivery is inapplicable for the retroviral vector owing to its limited DNA load (Liu et al., 2010) Additionally, retroviruses are unable to infect non-dividing cells (Liu et al., 2010) Although it seems safe for normal tissue, tumors also contain non-dividing cells in the resting phase G0 Consequently, those cells can evade therapy (Liu et al., 2010)

Lentiviruses belong to the retrovirus family and include the human immunodeficiency virus (HIV) Lentiviruses have an advantage over retroviruses as they can infect non dividing cells and provide a higher transfection efficiency compared to retroviruses (Chen et al., 2010)

Lentiviruses are more complex than retroviruses, and there is a possible risk of insertional mutagenesis by lentiviral vectors Hence, the development of lentiviral vectors is slow and cautious (Liu et al., 2010)

Baculovirus

As most viruses used in biotherapy are mammalian viruses, there is the possibility of pre-existing human immune memory that can hinder their application for gene therapy

(Rivera-Gonzalez et al., 2011) Moreover, mammalian viral proteins may be expressed

in patients and lead to diseases with cancer as an example Therefore, baculovirus

Autographa californica multiple nucelopolyhedrovirus (AcMNPV), insect viruses with an

inactive viral promoter in human cells, may represent a novel and exotic vector for cancer gene therapy

There are certain benefits of using baculoviruses as gene carriers First, baculoviruses transduce both dividing and quiescent mammalian cells with a high level of gene expression (Rivera-Gonzalez et al., 2011) Second, no record has shown baculovirus replication in mammalian cells, or cytotoxicity in human cells even at high virus particle per cell ratios and no host genome integration that lead to malignant development as compared to retroviruses (Laakkonen et al., 2008) Third, the construction and high titer production of baculoviruses can be performed easily In addition, there was evidence indicating th at baculovirus particles penetrate complex 3D tissues in a prostate spheroid culture, while adenoviruses only infected the outermost cell layers (Rivera-Gonzalez et al., 2011; Maitland et al., 2010)

Other viral vectors also generated interest, e.g adeno-asssociated virus, the herpes simplex virus, the vaccinia virus, the vesicular stomatitis virus and the measles virus

1.1.2.2 Nonviral vectors

Although viral vectors possess superiority in efficient gene delivery, their clinical application is limited by their safety issues with unexpected adverse effects Therefore, several non viral strategies have been developed recently

Naked DNA

The direct transfer of naked or plasmid DNA into certain tissues, particularly muscle, is the simplest nonviral gene delivery system that indeed shows high levels of gene expression Surprisingly, intra-arterial injection of a naked plasmid encoding endostatin inhibited systemic angiogenesis and delayed tumor growth (Barnett et al., 2004) These results, along with the simplicity of the procedure, proved the potential of using naked DNA for cancer gene therapy

However, free DNA is rapidly degraded and has low cellular uptake (Touchefeu et al., 2010) As a result, major applications of naked DNA are vaccination and immunotherapy Plasmid DNA also suffers from low transfection efficiency Electroporation increased the transfer efficiency of DNA into cells, though it is still transient, and the safety of this method needs to be investigated (Prud'homme et al., 2006; Favard et al., 2007)

Synthetic vectors

Cationic liposomes are artificial vesicles of phospholipids and cholesterol (Chen et al., 2010) Liposomes have been used to deliver various molecules such as small drug molecules, proteins, nucleotides, and plasmids Liposomes can be modified to become more hydrophilic to increase their circulation time in the bloodstream when introduced intravenously (Chen et al., 2010) They can also be conjugated with antibodies, proteins and peptides to enhance tumor targeting The complex generated by the incubation of DNA with a liposome is called a lipoplex (Touchefeu et al., 2010) DNA delivery in the form of a lipoplex prevents DNA from degrading and increases expression levels

Over the last few decades, advances in nanotechnology have introduced nanoparticles

as possible gene therapy vectors Nanoparticles are nanometer-scale vectors that can

be lipoplexes, other polymers, or metal- and ceramic-based (Touchefeu et al., 2010) Historically, the first generation nanovectors developed accumulated passively at tumor sites based on the permeability of the tumor-associated neovasculature by a mechanism named enhanced permeability and retention (EPR) effect (Maeda and Matsumura, 2011) The second generation marked the improvement in remote activation, environmentally sensitive components and targeting moieties (Torchilin, 2005; Debbage, 2009) Currently, the third generation nanovectors with concept called multistage nanovectors are under investigation (Sengupta et al., 2005; Wong et al., 2011)

Bacteria

The notion to employ bacteria as gene therapy vectors was derived from reports that showed the correlation between bacterial infection and tumor regression in the 1950s and nineteenth century (Svoboda MG et al., 1999) There are two approaches for the use of bacteria as gene vectors: tumor-specific bacterial replication and intracellular plasmid transfer (bactofection) describing a bacterial system that carries therapeutic plasmid DNA (Baban et al., 2010)

Tumor-specific replication of bacteria or bacterial colonization in tumor can be explained

by hypoxic nature of tumors This typical hypoxic tumor microenvironment may promote the growth of anaerobic bacteria (Al-Mariri et al., 2002) Moreover, necrotic regions in tumors provide nutrients and support bacterial colonization Because neovasculature is highly disorganized, it allows bacteria circulate to enter tumor tissue Although wound healing tissue has a similar vasculature to tumor, bacteria is cleared rapidly This suggests that the immune suppression of the tumor protects bacteria from being

eradicated Secreted factors from tumors also attract bacterial movement (Baban et al., 2010) Hence, bacterial tumor colonization shows great value for developing bacteria as gene therapy vectors Bacteria can be modified for oncolytic vectors and anti-angiogenic agents In addition, fluorescent and luminescent reporter systems have been transferred to bacteria to locate tumors where bacteria colonize (Cronin et al., 2008;

Riedel et al., 2007) E.Coli (Baban et al., 2010)and S.typhimurium (Soghomonyan et al.,

2005) were engineered to express HSV-tk

The uniqueness of the bacterial system is that bacteria can be controlled after administration by antibiotics (Baban et al., 2010)

1.2 Stem cells as unique candidates for therapeutic platform

To recapitulate the previous section, although constant attempts have been made to design and investigate novel vectors for cancer therapy, the dynamics and complexity of tumor biology create numerous barriers to overcome Each type of vector, on the one hand, provides attractive solutions for anti-cancer treatment, but on the other hand may expose some shortcomings that generate hesitation For instance, viral vectors deliver superior gene expression, but produce the risk of viral protein production and random integration that leads to malignancy Likewise, one should not underestimate viral immunogenicity that can decrease transgene expression Nonviral vectors such as naked DNA, liposomes or nanoparticles may induce low immune response without the risk of mutations, and be easily modified for better targeting, but have such low efficiency performance and are quickly degraded Both types reveal other issues of poor tumor penetration and uneven distribution throughout tumor mass, especially if tumors are densely packed Hence, it is crucial to make further modifications in existent vectors

to minimize drawbacks or propose other alternatives Stem cell vectors have recently received more attention and may give us practical options for cancer gene therapy

1.2.1 What are stem cells

Unique properties for any stem cells include their capability to renew and maintain themselves in an undifferentiated stage for a long time through mitotic cell division and

to differentiate into specialized cell types The latter one is defined as the potency of

stem cells

1.2.1.1 Pluripotent stem cells

Except for totipotent stem cells consisting of zygotes and their offspring cells from the first few divisions, pluripotent stem cells are the most powerful cells that are capable of giving rise to all cell progenies in an adult body

1.2.1.1.1 Embryonic stem cells

Embryonic stem cells (ESCs) were first derived from the inner cell mass (ICM) layer of

surplus blastocysts from in vitro fertilization (Philonenko et al., 2011) Other sources

such as morulae, late-arrested embryos or blastomeres also have been used to create new ESC lines (Philonenko et al., 2011) Under special cell culture conditions, ESCs maintain their pluripotency for an extended time When ESCs are cultured on a feeder layer of mitotically inactivated mouse embryonic fibroblasts (MEFs), a small portion of the cells spontaneously differentiate With new method to culture on Matrigel in mTESR1 medium, fewer differentiated cells are observed Up to now, approximately 59 ESC lines have been characterized with defined markers: surface antigens SSEA3, SSEA4, TRA-1-60, TRA-1-81, GCTM2, GCT343, CD9, and Thy1 and pluripotent genes

Nanog, Oct4, TDGF1, GDF3, DNMT3B, and GABRB3 (International Stem Cell Initiative

et al., 2007)

Currently, protocols to differentiate ESCs into more than 50 specialized cell types have been developed (Deb and Sarda, 2008) The first clinical trial using human ESC derivatives was approved by the FDA in 2009 (Philonenko et al., 2011) The purpose of this study was to demonstrate the efficiency and safety of injecting oligodendrocyte progenitors derived from human ESCs In 2010, approval of the second clinical trial for human ESC derivatives was achieved with pigmented epithelium cells to treat Stargardt’s Macular Dystrophy and age-related macular degeneration (Philonenko et al., 2011)

However, more investigation and optimization need to be done to ensure that differentiation methods are based on defined culture conditions, that large-scale production of desired cells is feasible, and that contaminated cells are completely removed

1.2.1.1.2 Induced pluripotent stem cells

Induced pluripotent stem cells (iPSCs) were first generated in the Takahashi and Yamanaka lab in 2006 by retroviral delivery of four transcription factors Oct4, Sox2, Klf4, and c-Myc (Takahashi and Yamanaka, 2006) The ability to restore pluripotency from fully differentiated cells provides opportunities for personalized therapies and immunogenicity-free strategies because therapeutic cells can be derived from the same patient As of today, more than 300 iPSC lines have been obtained from keratinocytes, neural cells, stomach and liver cells, and lymphocytes (Philonenko et al., 2011)

Moreover, reprogramming provides significant improvement for studying the molecular and cellular basis of diseases by creating iPSC lines from patients with hereditary diseases such as Parkinson’s disease, Huntington’s disease, Down syndrome, thalassemia, and sickle-cell anemia (Philonenko et al., 2011)

One technical issue with iPSC production is the low reprogramming efficiency, approximately 0.02% for Takahashi and Yamanaka (Takahashi and Yamanaka, 2006) and 0.1% for mean effectiveness (Philonenko et al., 2011) Numerous studies have focused on increasing this efficiency It was found that various cell sources for reprogramming can generate different percentage of iPSCs (Philonenko et al., 2011) The higher the cells are in hierarchy of potency, the better outcome is achieved For example, 13% of iPSCs are obtained from hematopoietic stem cells, whereas only 0.02% from differentiated B and T cells (Eminli et al., 2009) Small molecules (valproic acid, BIX-01294, Bayk8644, and sodium butyrate) increase the efficiency or replace the delivery of some required transcription factors (Philonenko et al., 2011) Reprogramming in hypoxic environment or with inhibition of apoptosis and p53 may improve effectiveness (Philonenko et al., 2011) To reduce the risk of random integration by viruses, efforts were also made to produce iPSC lines with plasmid vectors, but efficiency was extremely low (0.001%) (Philonenko et al., 2011)

Although ESCs and iPSCs share similar pluripotency and characteristics, there are a number of differences in the genetic and epigenetic profile Small differences in the efficiency of neuronal and hematopoietic differentiation from ESCs and iPSCs have been observed (Philonenko et al., 2011) These dissimilarities need to be considered for