CHAPTER 107 Gene- and Viral-Based Therapies for Gliomas
Advances in understanding gene structure/function and in the capacity to manipulate their expression have set the stage to alter genetic material for fighting or preventing disease. As gene products and their role within the cellular environment become increasingly understood, the potential for treating disease by using DNA as a drug is being realized. All proteins are coded for by DNA, and most neoplastic diseases ultimately result from the expression or lack thereof of one or more proteins (e.g., oncogenes or tumor suppressor genes, respectively). In theory, therefore, diseases could be treated by expression of the appropriate protein in the affected cells. Gene therapy is an experimental treatment that involves introducing genetic material (DNA or RNA) into cells, and it has made important advances over the past decade. Within this short time span it has moved from the conceptual laboratory research stage to clinical translational trials for a variety of diseases. Among the diseases being studied are primary genetic disorders1 and acquired diseases (e.g., cancer, neurodegenerative diseases), with brain tumors being among the first human malignancies to be targeted by gene therapy. The most efficient approaches for gene delivery are based on viral vectors, which have proved relatively safe in the central nervous system (CNS) despite occasional morbidity and death in non-neurosurgical trials. Most recently, results from the first clinical trial to test gene therapy for a type of inherited blindness showed that the experimental treatment is safe and can improve sight.2,3 These new developments are a landmark for gene therapy technology and paint a promising future. Although applying gene therapy to inherited CNS disorders remains an extremely difficult challenge, far more promise has been shown in the area of acquired CNS disorders such as brain tumors, spinal cord injury, stroke, and degenerative diseases (Parkinson’s and Alzheimer’s).4 This chapter reviews the basic principles and applications of gene therapy for brain tumors.
What is Gene Therapy?
Two fundamental considerations in gene therapy relate to (1) what gene should be delivered/expressed and (2) how to deliver it. In its simplest form, gene therapy is the process by which nucleic acids are transferred into cells to generate a therapeutic effect. This can be achieved by either replacing defective or missing genes or introducing new functions to the host’s cells. For this purpose, the genetic material is coupled to additional regulatory sequences (promoters, enhancers, and regulatory elements) and packaged inside a gene delivery vehicle to enable transfer and expression of the intended gene product inside the cell (Fig. 107-1).
How is Gene Therapy Carried Out?
Although conceptually straightforward, efficient expression of foreign genes is the most critical aspect for the success of in vivo gene therapy. The first step in gene therapy involves gene delivery to facilitate expression of the therapeutic gene in the interior of a cell. The simplest method is direct introduction of therapeutic DNA into target cells by physical (i.e., electroporation) or chemical (i.e., lipofection) techniques.5 This approach still remains limited in application because it is relatively inefficient, can be used only with certain tissues, and requires large amounts of DNA. Furthermore, among the several barriers to successful gene delivery, foreign genes or the vectors used to deliver them, or both, can trigger a range of immune responses. However, sometimes these immune responses have been harnessed behind the concept of using gene therapy as a vaccine (i.e., DNA vaccines).6 The rationale behind this approach is that the immune response against the vector or introduced DNA, or both, could be exploited to provide a vaccine effect. This type of approach is thus used in some gene therapies against cancer or infectious disease.
The next difficulty for the foreign genetic material is that once within the cell, it must escape intracellular degradation and enter the nucleus so that it can be expressed (Fig. 107-1).7,8 Therefore, gene delivery systems (vectors) were designed to protect the genetic material. An ideal vector needs to meet three criteria: (1) it should protect the transgene against degradation by nucleases in the extracellular matrix, (2) it should bring the transgene across the plasma membrane into the nucleus of target cells, and (3) it should have no detrimental effects. Currently, such vectors for gene transfer can be classified into two categories: viral and nonviral.9,10
Nonviral Vectors
There are several methods for nonviral gene transfer. Physical approaches, including needle injection,11 electroporation,12,13 gene gun,14,15 ultrasound,16 and hydrodynamic delivery,17,18 use a physical force that permeates the cell membrane to facilitate intracellular gene transfer. However, these methods have been relatively inefficient in their capacity to transfer genes to a sufficiently elevated number of cells. Other nonviral carriers use cationic lipids, polymers, ceramic-based nanomaterial, carbon nanotubes, metal nanorods, and silica-based nanoparticles19,20 to deliver genes. Chemical approaches use synthetic or naturally occurring compounds as carriers to deliver the genetic material into cells.21 This approach involves the creation of an artificial lipid sphere with an aqueous core.22–24 This liposome, which carries the therapeutic DNA, is capable of passing the DNA through the target cell’s membrane.25,26 Synthetic and naturally occurring polymers represent another category of DNA carriers that have been used widely for gene delivery.27 In addition, lipid-polymer hybrid systems can be used.28,29 There is also new research on its way in which a 47th artificial human chromosome is introduced into target cells. This chromosome would exist along with the standard 46 without affecting their actions or causing any mutations, and it would be a large vector carrying a substantial amount of genes. A problem with this potential method is the difficulty delivering such a large molecule to the nucleus of a target cell.30,31 Although significant progress has been made in the application of various nonviral gene delivery systems, the majority of nonviral approaches remain much less efficient than viral vectors, especially for in vivo gene delivery.
Viral Vectors
Thus far, the most effective means of transferring DNA into somatic cells remains the use of a viral-based vector. Vectors for brain tumor therapy can be divided into two general categories: (1) replication-defective vectors (which we will designate vector from here on) and (2) replication-competent (replication-conditional, oncolytic viruses), which we will designate oncolytic viruses (OVs) from here on. In the first instance, the vector is derived from a virus from which all or most of the viral genes have been removed to minimize virus-mediated toxicity. In the second instance, selected viral genes are deleted or mutated so that viral targeting or replication, or both, can occur selectively in tumor versus endogenous neural cells. Up to now, the replication-defective vectors used in gene therapy trials for brain tumors have been based on retrovirus (RV) and adenovirus (AV). In terms of replication-competent (oncolytic) viruses, the ones used in clinical trials of brain tumors have been herpes simplex virus (HSV), AV, reovirus, and Newcastle disease virus (NDV) (Table 107-1). However, experimentally, almost any type of virus has been used either in a replication-defective or replication-competent fashion.
TABLE 107-1 Comparison of Viral Vectors and Oncolytic Viruses Used for Clinical Trials
| VECTOR OR OV | ADVANTAGES | DISADVANTAGES |
|---|---|---|
| Retrovirus (vector) |
FDA, Food and Drug Administration; MEK, mitogen-activated protein kinase/extracellular signal–regulated kinase; OV, oncolytic virus.
Retroviruses
RVs are a class of enveloped viruses containing a single-stranded RNA molecule as the genome. After infection, the viral RNA genome is reverse-transcribed into double-stranded DNA, which can be integrated into the chromosomes of host cells and expressed as proteins.32,33 They have been used primarily as vectors, although there are preclinical studies and a probable future attempt at a clinical trial involving the use of an RV-based OV.34 The advantages of RV vectors are that they are relatively easy to manipulate for gene therapy purposes and have been used widely.35 The available long-term experience and low toxicity to normal brain tissue make this vector a safe candidate for CNS gene therapy.36,37 One of the problems of RV vectors is that the viral genome can be inserted randomly in the genome of the host. If the insertion happens to be in the middle of one of the host genes, this gene will be disrupted (insertional mutagenesis). If the disrupted host gene is involved in regulating cell division, uncontrolled cell division (i.e., cancer) can occur.38,39 Other disadvantages of RV vectors are low titers, instability of the viral particles, and low transgene capacity (the maximal amount of DNA that can be packaged into an RV is just 7.5 kilobases [kb] of foreign DNA). Another drawback of RV vectors is the requirement that the target cell be dividing for integration and expression of viral genes. This restricts gene therapy solely to proliferating cells.40,41 In fact, although RV vectors were used in the initial gene therapy trials, their use for more recent gene therapy trials for cancer has been greatly reduced.
Adenoviral Vectors
AVs are nonenveloped viruses that contain a linear double-stranded DNA genome and cause respiratory and eye infections in humans. For CNS gene therapy, the commonly used AV vectors are derived from a subgroup that can be manipulated to produce replication-deficient vectors and have an insertion capacity of 10 kb of foreign DNA. AV vectors have become the mainstay of gene therapy and, in fact, have become a common tool in the kit of the molecular biology laboratory whenever a gene needs to be expressed in a mammalian cell. The vector can be extensively modified to target it away from its usual receptor, present in only some cells,42 to other receptors present in a desired target cell.43 Once bound to its receptor, the virus enters the cell in endosomal vesicles that fuse to lysosomes.44 Herein, the virus is able to liberate itself and escape to the cell nucleus. Viral DNA does not generally integrate and survives as an extrachromosomal element, from which gene expression derives. Extensive animal studies have revealed that gene expression is usually transient, unless the immune response can be minimized. In fact, long-term gene expression in animal neurons has been achieved by eliminating exposure of gene products/proteins to non-CNS areas45 and using AV vectors completely devoid of all viral genes (gutless vectors).46 In terms of gene delivery to brain tumors, the vector has been used to deliver almost any type of anticancer gene. For glioma clinical trials, however, AV vector–mediated delivery of the genes for HSV thymidine kinase (HSVtk),47 p53,48 and human interferon-β have been published thus far.
In summary, AVs have been used widely as vectors in gene therapy trials, and their advantages are their relative ease of manipulation and ability to be produced at high titer. Furthermore, AV vectors are very efficient at transducing a wide variety of cells, both dividing and nondividing, in vitro and in vivo. AVs do not usually integrate their genetic material into the host genome; rather, they replicate as episomal elements in the nucleus of the host cell and induce transient gene expression, and consequently there is no risk for insertional mutagenesis.49 The limitations of AV vectors are their short-term gene expression and the fact that they can produce toxic acute-phase responses. As clinical trials have begun to progress, it has become apparent that AV vectors may cause a significant innate immune response.50
Oncolytic Adenovirus
In this paradigm, engineered or naturally occurring strains of virus are created or discovered that appear to replicate better in tumor than in normal cells. This selectivity can occur at three general levels. First, every time a virus infects a cell, a robust antiviral response takes place inside the cell that consists of numerous “stress” or “danger” signals (Fig. 107-2). The overall effect of these signals is to limit the ability of the infecting virus to replicate to high levels and thus limit the number of viral progeny generated that could eventually infect neighboring cells. These “stress” signals consist of genes involved in the interferon, nuclear factor (NF)-κB, Toll-like receptor, PKR (double-stranded RNA-dependent protein kinase), and other pathways.51 In some cases, tumor cells have disabled some of these responses because they tend to be proapoptotic and antiproliferative. Therefore, tumor cells that have such disabled responses provide better targets for viral replication than do normal cells with intact antiviral responses. Second, in tumor cells, several genes involved in cell cycle regulation and apoptosis signaling are disrupted. These disruptions in tumor cell factors can be exploited to rationally engineer viral mutants that cannot replicate well in normal cells with intact cell cycle/apoptosis controls but will replicate in tumor cells that harbor these disruptions. Finally, viral mutants can be re-engineered so that they will target cell surface receptors present on tumor as opposed to normal cells. Thus far, clinical trials of brain tumors with OVs have involved mutants that use primarily the first two mechanisms of selective tumor targeting/lysis.
For instance, an engineered oncolytic AV (named ONYX-015) lacks a viral gene (E1B) that encodes for a protein that inactivates the cellular tumor suppressor protein p53. Initially, this mutant AV was thought to replicate in and lyse p53-deficient human tumor cells, but not cells with functional p53.52,53 Based on gene mutations, this vector could be of use in up to 50% of human glioblastomas carrying a p53 mutation.54 It may possess anticancer effects against the remaining tumors because p53 function is modulated by p14ARF protein, whose gene (CDKN2A) is frequently deleted in glioblastoma multiforme (GBM).55,56 However, subsequent experiments with this virus have revealed that the mechanism of tumor selectivity is not based on the lack of p53 tumor suppressor function but rather on a more complicated mechanism related to the nuclear export of viral mRNA into tumor cells.57 Other oncolytic AVs include those that target defects in the p16INK4a tumor suppressor pathway.58 The p16INK4a protein shares the CDKN2A gene region with p14ARF, thus making it one of the most commonly mutated sites in malignant brain tumors, as discussed in detail in another chapter.
Herpes Simplex Virus Type 1
HSV-1 is a human neurotropic virus, which led to initial interest focused on using HSV-1 as a vector for gene transfer to the nervous system but has not yet resulted in clinical trials. HSV-1 is an enveloped DNA virus whose genome spans 152 kb and encodes for more than 80 genes. Wild-type HSV-1 is able to proceed into a lytic life cycle after infection or persist as an intranuclear episome in a latent state. Latently infected neurons function normally and are not rejected by the immune system. Although the latent virus is almost silent transcriptionally, it does possess neuron-specific promoters that are capable of functioning during latency.59 Interest in the use of HSV-1 as a cancer-killing agent intensified in the 1990s after reports showed that a genetically engineered form was oncolytic in brain tumor models.60 Since then, numerous reports have detailed various types of genetically engineered mutants for glioma therapy. The two most commonly used types have revolved around deletions of the viral gene (UL39 or UL40) that encodes for the viral protein ICP6 (Fig. 107-3). This protein possesses the function of a ribonucleotide reductase, and recent reports have linked this defect to an ability to replicate more selectively in cells with a defect in the p16 tumor suppressor gene.61 The other type has focused on mutation of the viral gene that encodes for the viral protein ICP34.5. This viral gene disables/counteracts a host cell response (based on the enzyme PKR) that promotes cellular apoptosis during infection.62 Several studies have investigated the mechanisms of tumor selectivity of this mutant virus. In one case, the authors linked this viral defect to enhanced replication in tumor cells with elevated levels of Ras activity.63 However, others have disputed this claim and have linked mutant OV replication to activity of the MEK (mitogen-activated protein kinase/extracellular signal–regulated kinase) pathway, which is present in tumor more than normal cells.64,65 In addition, it is also possible that tumor cells may have lost some PKR function, thus enabling this mutant virus to replicate.66
